{ "quant_method": "exl3", "version": "0.0.2", "bits": 6.0, "head_bits": 6, "calibration": { "rows": 100, "cols": 2048 }, "out_scales": "auto", "tensor_storage": { "model.embed_tokens": { "stored_tensors": { "model.embed_tokens.weight": { "shape": [ 32000, 4096 ], "n_bytes": 262144000, "dtype": "torch.float16" } } }, "model.layers.0.input_layernorm": { "stored_tensors": { "model.layers.0.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.0.self_attn.q_proj": { "stored_tensors": { "model.layers.0.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.self_attn.k_proj": { "stored_tensors": { "model.layers.0.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.0.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.self_attn.v_proj": { "stored_tensors": { "model.layers.0.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.0.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.self_attn.o_proj": { "stored_tensors": { "model.layers.0.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.post_attention_layernorm": { "stored_tensors": { "model.layers.0.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.0.block_sparse_moe.gate": { "stored_tensors": { "model.layers.0.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.0.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.0.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.0.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.0.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.input_layernorm": { "stored_tensors": { "model.layers.1.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.1.self_attn.q_proj": { "stored_tensors": { "model.layers.1.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.self_attn.k_proj": { "stored_tensors": { "model.layers.1.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.1.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.self_attn.v_proj": { "stored_tensors": { "model.layers.1.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.1.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.self_attn.o_proj": { "stored_tensors": { "model.layers.1.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.post_attention_layernorm": { "stored_tensors": { "model.layers.1.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.1.block_sparse_moe.gate": { "stored_tensors": { "model.layers.1.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.1.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.1.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.1.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.1.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.input_layernorm": { "stored_tensors": { "model.layers.2.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.2.self_attn.q_proj": { "stored_tensors": { "model.layers.2.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.self_attn.k_proj": { "stored_tensors": { "model.layers.2.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.2.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.self_attn.v_proj": { "stored_tensors": { "model.layers.2.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.2.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.self_attn.o_proj": { "stored_tensors": { "model.layers.2.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.post_attention_layernorm": { "stored_tensors": { "model.layers.2.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.2.block_sparse_moe.gate": { "stored_tensors": { "model.layers.2.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.2.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.2.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.2.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.2.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.input_layernorm": { "stored_tensors": { "model.layers.3.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.3.self_attn.q_proj": { "stored_tensors": { "model.layers.3.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.self_attn.k_proj": { "stored_tensors": { "model.layers.3.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.3.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.self_attn.v_proj": { "stored_tensors": { "model.layers.3.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.3.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.self_attn.o_proj": { "stored_tensors": { "model.layers.3.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.post_attention_layernorm": { "stored_tensors": { "model.layers.3.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.3.block_sparse_moe.gate": { "stored_tensors": { "model.layers.3.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.3.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.3.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.3.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.3.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.input_layernorm": { "stored_tensors": { "model.layers.4.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.4.self_attn.q_proj": { "stored_tensors": { "model.layers.4.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.self_attn.k_proj": { "stored_tensors": { "model.layers.4.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.4.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.self_attn.v_proj": { "stored_tensors": { "model.layers.4.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.4.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.self_attn.o_proj": { "stored_tensors": { "model.layers.4.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.post_attention_layernorm": { "stored_tensors": { "model.layers.4.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.4.block_sparse_moe.gate": { "stored_tensors": { "model.layers.4.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.4.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.4.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.4.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.4.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.input_layernorm": { "stored_tensors": { "model.layers.5.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.5.self_attn.q_proj": { "stored_tensors": { "model.layers.5.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.self_attn.k_proj": { "stored_tensors": { "model.layers.5.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.5.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.self_attn.v_proj": { "stored_tensors": { "model.layers.5.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.5.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.self_attn.o_proj": { "stored_tensors": { "model.layers.5.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.post_attention_layernorm": { "stored_tensors": { "model.layers.5.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.5.block_sparse_moe.gate": { "stored_tensors": { "model.layers.5.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.5.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.5.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.5.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.5.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.input_layernorm": { "stored_tensors": { "model.layers.6.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.6.self_attn.q_proj": { "stored_tensors": { "model.layers.6.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.self_attn.k_proj": { "stored_tensors": { "model.layers.6.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.6.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.self_attn.v_proj": { "stored_tensors": { "model.layers.6.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.6.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.self_attn.o_proj": { "stored_tensors": { "model.layers.6.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.post_attention_layernorm": { "stored_tensors": { "model.layers.6.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.6.block_sparse_moe.gate": { "stored_tensors": { "model.layers.6.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.6.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.6.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.6.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.6.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.input_layernorm": { "stored_tensors": { "model.layers.7.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.7.self_attn.q_proj": { "stored_tensors": { "model.layers.7.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.self_attn.k_proj": { "stored_tensors": { "model.layers.7.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.7.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.self_attn.v_proj": { "stored_tensors": { "model.layers.7.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.7.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.self_attn.o_proj": { "stored_tensors": { "model.layers.7.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.post_attention_layernorm": { "stored_tensors": { "model.layers.7.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.7.block_sparse_moe.gate": { "stored_tensors": { "model.layers.7.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.7.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.7.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.7.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.7.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.input_layernorm": { "stored_tensors": { "model.layers.8.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.8.self_attn.q_proj": { "stored_tensors": { "model.layers.8.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.self_attn.k_proj": { "stored_tensors": { "model.layers.8.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.8.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.self_attn.v_proj": { "stored_tensors": { "model.layers.8.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.8.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.self_attn.o_proj": { "stored_tensors": { "model.layers.8.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.post_attention_layernorm": { "stored_tensors": { "model.layers.8.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.8.block_sparse_moe.gate": { "stored_tensors": { "model.layers.8.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.8.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.8.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.8.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.8.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.input_layernorm": { "stored_tensors": { "model.layers.9.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.9.self_attn.q_proj": { "stored_tensors": { "model.layers.9.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.self_attn.k_proj": { "stored_tensors": { "model.layers.9.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.9.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.self_attn.v_proj": { "stored_tensors": { "model.layers.9.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.9.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.self_attn.o_proj": { "stored_tensors": { "model.layers.9.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.post_attention_layernorm": { "stored_tensors": { "model.layers.9.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.9.block_sparse_moe.gate": { "stored_tensors": { "model.layers.9.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.9.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.9.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.9.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.9.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.input_layernorm": { "stored_tensors": { "model.layers.10.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.10.self_attn.q_proj": { "stored_tensors": { "model.layers.10.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.self_attn.k_proj": { "stored_tensors": { "model.layers.10.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.10.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.self_attn.v_proj": { "stored_tensors": { "model.layers.10.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.10.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.self_attn.o_proj": { "stored_tensors": { "model.layers.10.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.post_attention_layernorm": { "stored_tensors": { "model.layers.10.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.10.block_sparse_moe.gate": { "stored_tensors": { "model.layers.10.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.10.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.10.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.10.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.10.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.input_layernorm": { "stored_tensors": { "model.layers.11.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.11.self_attn.q_proj": { "stored_tensors": { "model.layers.11.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.self_attn.k_proj": { "stored_tensors": { "model.layers.11.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.11.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.self_attn.v_proj": { "stored_tensors": { "model.layers.11.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.11.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.self_attn.o_proj": { "stored_tensors": { "model.layers.11.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.post_attention_layernorm": { "stored_tensors": { "model.layers.11.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.11.block_sparse_moe.gate": { "stored_tensors": { "model.layers.11.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.11.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.11.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.11.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.11.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.input_layernorm": { "stored_tensors": { "model.layers.12.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.12.self_attn.q_proj": { "stored_tensors": { "model.layers.12.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.self_attn.k_proj": { "stored_tensors": { "model.layers.12.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.12.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.self_attn.v_proj": { "stored_tensors": { "model.layers.12.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.12.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.self_attn.o_proj": { "stored_tensors": { "model.layers.12.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.post_attention_layernorm": { "stored_tensors": { "model.layers.12.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.12.block_sparse_moe.gate": { "stored_tensors": { "model.layers.12.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.12.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.12.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.12.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.12.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.input_layernorm": { "stored_tensors": { "model.layers.13.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.13.self_attn.q_proj": { "stored_tensors": { "model.layers.13.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.self_attn.k_proj": { "stored_tensors": { "model.layers.13.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.13.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.self_attn.v_proj": { "stored_tensors": { "model.layers.13.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.13.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.self_attn.o_proj": { "stored_tensors": { "model.layers.13.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.post_attention_layernorm": { "stored_tensors": { "model.layers.13.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.13.block_sparse_moe.gate": { "stored_tensors": { "model.layers.13.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.13.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.13.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.13.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.13.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.input_layernorm": { "stored_tensors": { "model.layers.14.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.14.self_attn.q_proj": { "stored_tensors": { "model.layers.14.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.self_attn.k_proj": { "stored_tensors": { "model.layers.14.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.14.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.self_attn.v_proj": { "stored_tensors": { "model.layers.14.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.14.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.self_attn.o_proj": { "stored_tensors": { "model.layers.14.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.post_attention_layernorm": { "stored_tensors": { "model.layers.14.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.14.block_sparse_moe.gate": { "stored_tensors": { "model.layers.14.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.14.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.14.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.14.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.14.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.input_layernorm": { "stored_tensors": { "model.layers.15.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.15.self_attn.q_proj": { "stored_tensors": { "model.layers.15.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.self_attn.k_proj": { "stored_tensors": { "model.layers.15.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.15.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.self_attn.v_proj": { "stored_tensors": { "model.layers.15.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.15.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.self_attn.o_proj": { "stored_tensors": { "model.layers.15.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.post_attention_layernorm": { "stored_tensors": { "model.layers.15.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.15.block_sparse_moe.gate": { "stored_tensors": { "model.layers.15.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.15.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.15.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.15.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.15.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.input_layernorm": { "stored_tensors": { "model.layers.16.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.16.self_attn.q_proj": { "stored_tensors": { "model.layers.16.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.self_attn.k_proj": { "stored_tensors": { "model.layers.16.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.16.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.self_attn.v_proj": { "stored_tensors": { "model.layers.16.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.16.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.self_attn.o_proj": { "stored_tensors": { "model.layers.16.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.post_attention_layernorm": { "stored_tensors": { "model.layers.16.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.16.block_sparse_moe.gate": { "stored_tensors": { "model.layers.16.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.16.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.16.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.16.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.16.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.input_layernorm": { "stored_tensors": { "model.layers.17.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.17.self_attn.q_proj": { "stored_tensors": { "model.layers.17.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.self_attn.k_proj": { "stored_tensors": { "model.layers.17.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.17.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.self_attn.v_proj": { "stored_tensors": { "model.layers.17.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.17.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.self_attn.o_proj": { "stored_tensors": { "model.layers.17.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.post_attention_layernorm": { "stored_tensors": { "model.layers.17.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.17.block_sparse_moe.gate": { "stored_tensors": { "model.layers.17.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.17.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.17.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.17.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.17.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.input_layernorm": { "stored_tensors": { "model.layers.18.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.18.self_attn.q_proj": { "stored_tensors": { "model.layers.18.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.self_attn.k_proj": { "stored_tensors": { "model.layers.18.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.18.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.self_attn.v_proj": { "stored_tensors": { "model.layers.18.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.18.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.self_attn.o_proj": { "stored_tensors": { "model.layers.18.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.post_attention_layernorm": { "stored_tensors": { "model.layers.18.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.18.block_sparse_moe.gate": { "stored_tensors": { "model.layers.18.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.18.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.18.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.18.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.18.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.input_layernorm": { "stored_tensors": { "model.layers.19.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.19.self_attn.q_proj": { "stored_tensors": { "model.layers.19.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.self_attn.k_proj": { "stored_tensors": { "model.layers.19.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.19.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.self_attn.v_proj": { "stored_tensors": { "model.layers.19.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.19.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.self_attn.o_proj": { "stored_tensors": { "model.layers.19.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.post_attention_layernorm": { "stored_tensors": { "model.layers.19.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.19.block_sparse_moe.gate": { "stored_tensors": { "model.layers.19.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.19.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.19.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.19.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.19.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.input_layernorm": { "stored_tensors": { "model.layers.20.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.20.self_attn.q_proj": { "stored_tensors": { "model.layers.20.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.self_attn.k_proj": { "stored_tensors": { "model.layers.20.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.20.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.self_attn.v_proj": { "stored_tensors": { "model.layers.20.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.20.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.self_attn.o_proj": { "stored_tensors": { "model.layers.20.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.post_attention_layernorm": { "stored_tensors": { "model.layers.20.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.20.block_sparse_moe.gate": { "stored_tensors": { "model.layers.20.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.20.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.20.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.20.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.20.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.input_layernorm": { "stored_tensors": { "model.layers.21.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.21.self_attn.q_proj": { "stored_tensors": { "model.layers.21.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.self_attn.k_proj": { "stored_tensors": { "model.layers.21.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.21.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.self_attn.v_proj": { "stored_tensors": { "model.layers.21.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.21.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.self_attn.o_proj": { "stored_tensors": { "model.layers.21.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.post_attention_layernorm": { "stored_tensors": { "model.layers.21.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.21.block_sparse_moe.gate": { "stored_tensors": { "model.layers.21.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.21.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.21.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.21.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.21.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.input_layernorm": { "stored_tensors": { "model.layers.22.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.22.self_attn.q_proj": { "stored_tensors": { "model.layers.22.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.self_attn.k_proj": { "stored_tensors": { "model.layers.22.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.22.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.self_attn.v_proj": { "stored_tensors": { "model.layers.22.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.22.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.self_attn.o_proj": { "stored_tensors": { "model.layers.22.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.post_attention_layernorm": { "stored_tensors": { "model.layers.22.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.22.block_sparse_moe.gate": { "stored_tensors": { "model.layers.22.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.22.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.22.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.22.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.22.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.input_layernorm": { "stored_tensors": { "model.layers.23.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.23.self_attn.q_proj": { "stored_tensors": { "model.layers.23.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.self_attn.k_proj": { "stored_tensors": { "model.layers.23.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.23.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.self_attn.v_proj": { "stored_tensors": { "model.layers.23.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.23.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.self_attn.o_proj": { "stored_tensors": { "model.layers.23.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.post_attention_layernorm": { "stored_tensors": { "model.layers.23.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.23.block_sparse_moe.gate": { "stored_tensors": { "model.layers.23.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.23.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.23.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.23.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.23.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.input_layernorm": { "stored_tensors": { "model.layers.24.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.24.self_attn.q_proj": { "stored_tensors": { "model.layers.24.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.self_attn.k_proj": { "stored_tensors": { "model.layers.24.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.24.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.self_attn.v_proj": { "stored_tensors": { "model.layers.24.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.24.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.self_attn.o_proj": { "stored_tensors": { "model.layers.24.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.post_attention_layernorm": { "stored_tensors": { "model.layers.24.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.24.block_sparse_moe.gate": { "stored_tensors": { "model.layers.24.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.24.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.24.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.24.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.24.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.input_layernorm": { "stored_tensors": { "model.layers.25.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.25.self_attn.q_proj": { "stored_tensors": { "model.layers.25.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.self_attn.k_proj": { "stored_tensors": { "model.layers.25.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.25.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.self_attn.v_proj": { "stored_tensors": { "model.layers.25.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.25.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.self_attn.o_proj": { "stored_tensors": { "model.layers.25.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.post_attention_layernorm": { "stored_tensors": { "model.layers.25.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.25.block_sparse_moe.gate": { "stored_tensors": { "model.layers.25.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.25.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.25.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.25.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.25.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.input_layernorm": { "stored_tensors": { "model.layers.26.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.26.self_attn.q_proj": { "stored_tensors": { "model.layers.26.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.self_attn.k_proj": { "stored_tensors": { "model.layers.26.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.26.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.self_attn.v_proj": { "stored_tensors": { "model.layers.26.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.26.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.self_attn.o_proj": { "stored_tensors": { "model.layers.26.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.post_attention_layernorm": { "stored_tensors": { "model.layers.26.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.26.block_sparse_moe.gate": { "stored_tensors": { "model.layers.26.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.26.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.26.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.26.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.26.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.input_layernorm": { "stored_tensors": { "model.layers.27.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.27.self_attn.q_proj": { "stored_tensors": { "model.layers.27.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.self_attn.k_proj": { "stored_tensors": { "model.layers.27.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.27.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.self_attn.v_proj": { "stored_tensors": { "model.layers.27.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.27.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.self_attn.o_proj": { "stored_tensors": { "model.layers.27.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.post_attention_layernorm": { "stored_tensors": { "model.layers.27.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.27.block_sparse_moe.gate": { "stored_tensors": { "model.layers.27.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.27.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.27.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.27.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.27.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.input_layernorm": { "stored_tensors": { "model.layers.28.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.28.self_attn.q_proj": { "stored_tensors": { "model.layers.28.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.self_attn.k_proj": { "stored_tensors": { "model.layers.28.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.28.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.self_attn.v_proj": { "stored_tensors": { "model.layers.28.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.28.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.self_attn.o_proj": { "stored_tensors": { "model.layers.28.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.post_attention_layernorm": { "stored_tensors": { "model.layers.28.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.28.block_sparse_moe.gate": { "stored_tensors": { "model.layers.28.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.28.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.28.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.28.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.28.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.input_layernorm": { "stored_tensors": { "model.layers.29.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.29.self_attn.q_proj": { "stored_tensors": { "model.layers.29.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.self_attn.k_proj": { "stored_tensors": { "model.layers.29.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.29.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.self_attn.v_proj": { "stored_tensors": { "model.layers.29.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.29.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.self_attn.o_proj": { "stored_tensors": { "model.layers.29.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.post_attention_layernorm": { "stored_tensors": { "model.layers.29.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.29.block_sparse_moe.gate": { "stored_tensors": { "model.layers.29.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.29.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.29.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.29.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.29.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.input_layernorm": { "stored_tensors": { "model.layers.30.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.30.self_attn.q_proj": { "stored_tensors": { "model.layers.30.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.self_attn.k_proj": { "stored_tensors": { "model.layers.30.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.30.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.self_attn.v_proj": { "stored_tensors": { "model.layers.30.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.30.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.self_attn.o_proj": { "stored_tensors": { "model.layers.30.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.post_attention_layernorm": { "stored_tensors": { "model.layers.30.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.30.block_sparse_moe.gate": { "stored_tensors": { "model.layers.30.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.30.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.30.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.30.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.30.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.input_layernorm": { "stored_tensors": { "model.layers.31.input_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.31.self_attn.q_proj": { "stored_tensors": { "model.layers.31.self_attn.q_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.q_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.q_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.self_attn.k_proj": { "stored_tensors": { "model.layers.31.self_attn.k_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.k_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.31.self_attn.k_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.self_attn.v_proj": { "stored_tensors": { "model.layers.31.self_attn.v_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.v_proj.svh": { "shape": [ 1024 ], "n_bytes": 2048, "dtype": "torch.float16" }, "model.layers.31.self_attn.v_proj.trellis": { "shape": [ 256, 64, 96 ], "n_bytes": 3145728, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.self_attn.o_proj": { "stored_tensors": { "model.layers.31.self_attn.o_proj.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.o_proj.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.self_attn.o_proj.trellis": { "shape": [ 256, 256, 96 ], "n_bytes": 12582912, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.post_attention_layernorm": { "stored_tensors": { "model.layers.31.post_attention_layernorm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "model.layers.31.block_sparse_moe.gate": { "stored_tensors": { "model.layers.31.block_sparse_moe.gate.weight": { "shape": [ 8, 4096 ], "n_bytes": 65536, "dtype": "torch.float16" } } }, "model.layers.31.block_sparse_moe.experts.0.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.0.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.0.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.0.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.0.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.0.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.0.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.1.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.1.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.1.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.1.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.1.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.1.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.1.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.2.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.2.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.2.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.2.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.2.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.2.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.2.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.3.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.3.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.3.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.3.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.3.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.3.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.3.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.4.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.4.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.4.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.4.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.4.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.4.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.4.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.5.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.5.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.5.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.5.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.5.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.5.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.5.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.6.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.6.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.6.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.6.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.6.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.6.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.6.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.7.w3": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.7.w3.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w3.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w3.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.7.w1": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.7.w1.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w1.svh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w1.trellis": { "shape": [ 256, 896, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.layers.31.block_sparse_moe.experts.7.w2": { "stored_tensors": { "model.layers.31.block_sparse_moe.experts.7.w2.suh": { "shape": [ 14336 ], "n_bytes": 28672, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w2.svh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "model.layers.31.block_sparse_moe.experts.7.w2.trellis": { "shape": [ 896, 256, 96 ], "n_bytes": 44040192, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 }, "model.norm": { "stored_tensors": { "model.norm.weight": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" } } }, "lm_head": { "stored_tensors": { "lm_head.suh": { "shape": [ 4096 ], "n_bytes": 8192, "dtype": "torch.float16" }, "lm_head.svh": { "shape": [ 32000 ], "n_bytes": 64000, "dtype": "torch.float16" }, "lm_head.trellis": { "shape": [ 256, 2000, 96 ], "n_bytes": 98304000, "dtype": "torch.int16" } }, "quant_format": "exl3", "bits_per_weight": 6 } } }