| { | |
| "architectures": [ | |
| "MixtralForCausalLM" | |
| ], | |
| "attention_dropout": 0.0, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "head_dim": null, | |
| "hidden_act": "silu", | |
| "hidden_size": 6144, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 16384, | |
| "max_position_embeddings": 32768, | |
| "model_type": "mixtral", | |
| "num_attention_heads": 48, | |
| "num_experts_per_tok": 2, | |
| "num_hidden_layers": 56, | |
| "num_key_value_heads": 8, | |
| "num_local_experts": 2, | |
| "output_router_logits": false, | |
| "quantization": { | |
| "group_size": 64, | |
| "bits": 8, | |
| "model.embed_tokens": true, | |
| "model.layers.0.self_attn.q_proj": true, | |
| "model.layers.0.self_attn.k_proj": true, | |
| "model.layers.0.self_attn.v_proj": true, | |
| "model.layers.0.self_attn.o_proj": true, | |
| "model.layers.0.self_attn.rope": false, | |
| "model.layers.0.block_sparse_moe.gate": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.0.input_layernorm": false, | |
| "model.layers.0.post_attention_layernorm": false, | |
| "model.layers.1.self_attn.q_proj": true, | |
| "model.layers.1.self_attn.k_proj": true, | |
| "model.layers.1.self_attn.v_proj": true, | |
| "model.layers.1.self_attn.o_proj": true, | |
| "model.layers.1.self_attn.rope": false, | |
| "model.layers.1.block_sparse_moe.gate": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.1.input_layernorm": false, | |
| "model.layers.1.post_attention_layernorm": false, | |
| "model.layers.2.self_attn.q_proj": true, | |
| "model.layers.2.self_attn.k_proj": true, | |
| "model.layers.2.self_attn.v_proj": true, | |
| "model.layers.2.self_attn.o_proj": true, | |
| "model.layers.2.self_attn.rope": false, | |
| "model.layers.2.block_sparse_moe.gate": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.2.input_layernorm": false, | |
| "model.layers.2.post_attention_layernorm": false, | |
| "model.layers.3.self_attn.q_proj": true, | |
| "model.layers.3.self_attn.k_proj": true, | |
| "model.layers.3.self_attn.v_proj": true, | |
| "model.layers.3.self_attn.o_proj": true, | |
| "model.layers.3.self_attn.rope": false, | |
| "model.layers.3.block_sparse_moe.gate": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.3.input_layernorm": false, | |
| "model.layers.3.post_attention_layernorm": false, | |
| "model.layers.4.self_attn.q_proj": true, | |
| "model.layers.4.self_attn.k_proj": true, | |
| "model.layers.4.self_attn.v_proj": true, | |
| "model.layers.4.self_attn.o_proj": true, | |
| "model.layers.4.self_attn.rope": false, | |
| "model.layers.4.block_sparse_moe.gate": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.4.input_layernorm": false, | |
| "model.layers.4.post_attention_layernorm": false, | |
| "model.layers.5.self_attn.q_proj": true, | |
| "model.layers.5.self_attn.k_proj": true, | |
| "model.layers.5.self_attn.v_proj": true, | |
| "model.layers.5.self_attn.o_proj": true, | |
| "model.layers.5.self_attn.rope": false, | |
| "model.layers.5.block_sparse_moe.gate": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.5.input_layernorm": false, | |
| "model.layers.5.post_attention_layernorm": false, | |
| "model.layers.6.self_attn.q_proj": true, | |
| "model.layers.6.self_attn.k_proj": true, | |
| "model.layers.6.self_attn.v_proj": true, | |
| "model.layers.6.self_attn.o_proj": true, | |
| "model.layers.6.self_attn.rope": false, | |
| "model.layers.6.block_sparse_moe.gate": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.6.input_layernorm": false, | |
| "model.layers.6.post_attention_layernorm": false, | |
| "model.layers.7.self_attn.q_proj": true, | |
| "model.layers.7.self_attn.k_proj": true, | |
| "model.layers.7.self_attn.v_proj": true, | |
| "model.layers.7.self_attn.o_proj": true, | |
| "model.layers.7.self_attn.rope": false, | |
| "model.layers.7.block_sparse_moe.gate": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.7.input_layernorm": false, | |
| "model.layers.7.post_attention_layernorm": false, | |
| "model.layers.8.self_attn.q_proj": true, | |
| "model.layers.8.self_attn.k_proj": true, | |
| "model.layers.8.self_attn.v_proj": true, | |
| "model.layers.8.self_attn.o_proj": true, | |
| "model.layers.8.self_attn.rope": false, | |
| "model.layers.8.block_sparse_moe.gate": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.8.input_layernorm": false, | |
| "model.layers.8.post_attention_layernorm": false, | |
| "model.layers.9.self_attn.q_proj": true, | |
| "model.layers.9.self_attn.k_proj": true, | |
| "model.layers.9.self_attn.v_proj": true, | |
| "model.layers.9.self_attn.o_proj": true, | |
| "model.layers.9.self_attn.rope": false, | |
| "model.layers.9.block_sparse_moe.gate": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.9.input_layernorm": false, | |
| "model.layers.9.post_attention_layernorm": false, | |
| "model.layers.10.self_attn.q_proj": true, | |
| "model.layers.10.self_attn.k_proj": true, | |
| "model.layers.10.self_attn.v_proj": true, | |
| "model.layers.10.self_attn.o_proj": true, | |
| "model.layers.10.self_attn.rope": false, | |
| "model.layers.10.block_sparse_moe.gate": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.10.input_layernorm": false, | |
| "model.layers.10.post_attention_layernorm": false, | |
| "model.layers.11.self_attn.q_proj": true, | |
| "model.layers.11.self_attn.k_proj": true, | |
| "model.layers.11.self_attn.v_proj": true, | |
| "model.layers.11.self_attn.o_proj": true, | |
| "model.layers.11.self_attn.rope": false, | |
| "model.layers.11.block_sparse_moe.gate": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.11.input_layernorm": false, | |
| "model.layers.11.post_attention_layernorm": false, | |
| "model.layers.12.self_attn.q_proj": true, | |
| "model.layers.12.self_attn.k_proj": true, | |
| "model.layers.12.self_attn.v_proj": true, | |
| "model.layers.12.self_attn.o_proj": true, | |
| "model.layers.12.self_attn.rope": false, | |
| "model.layers.12.block_sparse_moe.gate": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.12.input_layernorm": false, | |
| "model.layers.12.post_attention_layernorm": false, | |
| "model.layers.13.self_attn.q_proj": true, | |
| "model.layers.13.self_attn.k_proj": true, | |
| "model.layers.13.self_attn.v_proj": true, | |
| "model.layers.13.self_attn.o_proj": true, | |
| "model.layers.13.self_attn.rope": false, | |
| "model.layers.13.block_sparse_moe.gate": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.13.input_layernorm": false, | |
| "model.layers.13.post_attention_layernorm": false, | |
| "model.layers.14.self_attn.q_proj": true, | |
| "model.layers.14.self_attn.k_proj": true, | |
| "model.layers.14.self_attn.v_proj": true, | |
| "model.layers.14.self_attn.o_proj": true, | |
| "model.layers.14.self_attn.rope": false, | |
| "model.layers.14.block_sparse_moe.gate": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.14.input_layernorm": false, | |
| "model.layers.14.post_attention_layernorm": false, | |
| "model.layers.15.self_attn.q_proj": true, | |
| "model.layers.15.self_attn.k_proj": true, | |
| "model.layers.15.self_attn.v_proj": true, | |
| "model.layers.15.self_attn.o_proj": true, | |
| "model.layers.15.self_attn.rope": false, | |
| "model.layers.15.block_sparse_moe.gate": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.15.input_layernorm": false, | |
| "model.layers.15.post_attention_layernorm": false, | |
| "model.layers.16.self_attn.q_proj": true, | |
| "model.layers.16.self_attn.k_proj": true, | |
| "model.layers.16.self_attn.v_proj": true, | |
| "model.layers.16.self_attn.o_proj": true, | |
| "model.layers.16.self_attn.rope": false, | |
| "model.layers.16.block_sparse_moe.gate": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.16.input_layernorm": false, | |
| "model.layers.16.post_attention_layernorm": false, | |
| "model.layers.17.self_attn.q_proj": true, | |
| "model.layers.17.self_attn.k_proj": true, | |
| "model.layers.17.self_attn.v_proj": true, | |
| "model.layers.17.self_attn.o_proj": true, | |
| "model.layers.17.self_attn.rope": false, | |
| "model.layers.17.block_sparse_moe.gate": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.17.input_layernorm": false, | |
| "model.layers.17.post_attention_layernorm": false, | |
| "model.layers.18.self_attn.q_proj": true, | |
| "model.layers.18.self_attn.k_proj": true, | |
| "model.layers.18.self_attn.v_proj": true, | |
| "model.layers.18.self_attn.o_proj": true, | |
| "model.layers.18.self_attn.rope": false, | |
| "model.layers.18.block_sparse_moe.gate": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.18.input_layernorm": false, | |
| "model.layers.18.post_attention_layernorm": false, | |
| "model.layers.19.self_attn.q_proj": true, | |
| "model.layers.19.self_attn.k_proj": true, | |
| "model.layers.19.self_attn.v_proj": true, | |
| "model.layers.19.self_attn.o_proj": true, | |
| "model.layers.19.self_attn.rope": false, | |
| "model.layers.19.block_sparse_moe.gate": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.19.input_layernorm": false, | |
| "model.layers.19.post_attention_layernorm": false, | |
| "model.layers.20.self_attn.q_proj": true, | |
| "model.layers.20.self_attn.k_proj": true, | |
| "model.layers.20.self_attn.v_proj": true, | |
| "model.layers.20.self_attn.o_proj": true, | |
| "model.layers.20.self_attn.rope": false, | |
| "model.layers.20.block_sparse_moe.gate": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.20.input_layernorm": false, | |
| "model.layers.20.post_attention_layernorm": false, | |
| "model.layers.21.self_attn.q_proj": true, | |
| "model.layers.21.self_attn.k_proj": true, | |
| "model.layers.21.self_attn.v_proj": true, | |
| "model.layers.21.self_attn.o_proj": true, | |
| "model.layers.21.self_attn.rope": false, | |
| "model.layers.21.block_sparse_moe.gate": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.21.input_layernorm": false, | |
| "model.layers.21.post_attention_layernorm": false, | |
| "model.layers.22.self_attn.q_proj": true, | |
| "model.layers.22.self_attn.k_proj": true, | |
| "model.layers.22.self_attn.v_proj": true, | |
| "model.layers.22.self_attn.o_proj": true, | |
| "model.layers.22.self_attn.rope": false, | |
| "model.layers.22.block_sparse_moe.gate": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.22.input_layernorm": false, | |
| "model.layers.22.post_attention_layernorm": false, | |
| "model.layers.23.self_attn.q_proj": true, | |
| "model.layers.23.self_attn.k_proj": true, | |
| "model.layers.23.self_attn.v_proj": true, | |
| "model.layers.23.self_attn.o_proj": true, | |
| "model.layers.23.self_attn.rope": false, | |
| "model.layers.23.block_sparse_moe.gate": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.23.input_layernorm": false, | |
| "model.layers.23.post_attention_layernorm": false, | |
| "model.layers.24.self_attn.q_proj": true, | |
| "model.layers.24.self_attn.k_proj": true, | |
| "model.layers.24.self_attn.v_proj": true, | |
| "model.layers.24.self_attn.o_proj": true, | |
| "model.layers.24.self_attn.rope": false, | |
| "model.layers.24.block_sparse_moe.gate": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.24.input_layernorm": false, | |
| "model.layers.24.post_attention_layernorm": false, | |
| "model.layers.25.self_attn.q_proj": true, | |
| "model.layers.25.self_attn.k_proj": true, | |
| "model.layers.25.self_attn.v_proj": true, | |
| "model.layers.25.self_attn.o_proj": true, | |
| "model.layers.25.self_attn.rope": false, | |
| "model.layers.25.block_sparse_moe.gate": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.25.input_layernorm": false, | |
| "model.layers.25.post_attention_layernorm": false, | |
| "model.layers.26.self_attn.q_proj": true, | |
| "model.layers.26.self_attn.k_proj": true, | |
| "model.layers.26.self_attn.v_proj": true, | |
| "model.layers.26.self_attn.o_proj": true, | |
| "model.layers.26.self_attn.rope": false, | |
| "model.layers.26.block_sparse_moe.gate": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.26.input_layernorm": false, | |
| "model.layers.26.post_attention_layernorm": false, | |
| "model.layers.27.self_attn.q_proj": true, | |
| "model.layers.27.self_attn.k_proj": true, | |
| "model.layers.27.self_attn.v_proj": true, | |
| "model.layers.27.self_attn.o_proj": true, | |
| "model.layers.27.self_attn.rope": false, | |
| "model.layers.27.block_sparse_moe.gate": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.27.input_layernorm": false, | |
| "model.layers.27.post_attention_layernorm": false, | |
| "model.layers.28.self_attn.q_proj": true, | |
| "model.layers.28.self_attn.k_proj": true, | |
| "model.layers.28.self_attn.v_proj": true, | |
| "model.layers.28.self_attn.o_proj": true, | |
| "model.layers.28.self_attn.rope": false, | |
| "model.layers.28.block_sparse_moe.gate": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.28.input_layernorm": false, | |
| "model.layers.28.post_attention_layernorm": false, | |
| "model.layers.29.self_attn.q_proj": true, | |
| "model.layers.29.self_attn.k_proj": true, | |
| "model.layers.29.self_attn.v_proj": true, | |
| "model.layers.29.self_attn.o_proj": true, | |
| "model.layers.29.self_attn.rope": false, | |
| "model.layers.29.block_sparse_moe.gate": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.29.input_layernorm": false, | |
| "model.layers.29.post_attention_layernorm": false, | |
| "model.layers.30.self_attn.q_proj": true, | |
| "model.layers.30.self_attn.k_proj": true, | |
| "model.layers.30.self_attn.v_proj": true, | |
| "model.layers.30.self_attn.o_proj": true, | |
| "model.layers.30.self_attn.rope": false, | |
| "model.layers.30.block_sparse_moe.gate": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.30.input_layernorm": false, | |
| "model.layers.30.post_attention_layernorm": false, | |
| "model.layers.31.self_attn.q_proj": true, | |
| "model.layers.31.self_attn.k_proj": true, | |
| "model.layers.31.self_attn.v_proj": true, | |
| "model.layers.31.self_attn.o_proj": true, | |
| "model.layers.31.self_attn.rope": false, | |
| "model.layers.31.block_sparse_moe.gate": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.31.input_layernorm": false, | |
| "model.layers.31.post_attention_layernorm": false, | |
| "model.layers.32.self_attn.q_proj": true, | |
| "model.layers.32.self_attn.k_proj": true, | |
| "model.layers.32.self_attn.v_proj": true, | |
| "model.layers.32.self_attn.o_proj": true, | |
| "model.layers.32.self_attn.rope": false, | |
| "model.layers.32.block_sparse_moe.gate": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.32.input_layernorm": false, | |
| "model.layers.32.post_attention_layernorm": false, | |
| "model.layers.33.self_attn.q_proj": true, | |
| "model.layers.33.self_attn.k_proj": true, | |
| "model.layers.33.self_attn.v_proj": true, | |
| "model.layers.33.self_attn.o_proj": true, | |
| "model.layers.33.self_attn.rope": false, | |
| "model.layers.33.block_sparse_moe.gate": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.33.input_layernorm": false, | |
| "model.layers.33.post_attention_layernorm": false, | |
| "model.layers.34.self_attn.q_proj": true, | |
| "model.layers.34.self_attn.k_proj": true, | |
| "model.layers.34.self_attn.v_proj": true, | |
| "model.layers.34.self_attn.o_proj": true, | |
| "model.layers.34.self_attn.rope": false, | |
| "model.layers.34.block_sparse_moe.gate": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.34.input_layernorm": false, | |
| "model.layers.34.post_attention_layernorm": false, | |
| "model.layers.35.self_attn.q_proj": true, | |
| "model.layers.35.self_attn.k_proj": true, | |
| "model.layers.35.self_attn.v_proj": true, | |
| "model.layers.35.self_attn.o_proj": true, | |
| "model.layers.35.self_attn.rope": false, | |
| "model.layers.35.block_sparse_moe.gate": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.35.input_layernorm": false, | |
| "model.layers.35.post_attention_layernorm": false, | |
| "model.layers.36.self_attn.q_proj": true, | |
| "model.layers.36.self_attn.k_proj": true, | |
| "model.layers.36.self_attn.v_proj": true, | |
| "model.layers.36.self_attn.o_proj": true, | |
| "model.layers.36.self_attn.rope": false, | |
| "model.layers.36.block_sparse_moe.gate": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.36.input_layernorm": false, | |
| "model.layers.36.post_attention_layernorm": false, | |
| "model.layers.37.self_attn.q_proj": true, | |
| "model.layers.37.self_attn.k_proj": true, | |
| "model.layers.37.self_attn.v_proj": true, | |
| "model.layers.37.self_attn.o_proj": true, | |
| "model.layers.37.self_attn.rope": false, | |
| "model.layers.37.block_sparse_moe.gate": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.37.input_layernorm": false, | |
| "model.layers.37.post_attention_layernorm": false, | |
| "model.layers.38.self_attn.q_proj": true, | |
| "model.layers.38.self_attn.k_proj": true, | |
| "model.layers.38.self_attn.v_proj": true, | |
| "model.layers.38.self_attn.o_proj": true, | |
| "model.layers.38.self_attn.rope": false, | |
| "model.layers.38.block_sparse_moe.gate": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.38.input_layernorm": false, | |
| "model.layers.38.post_attention_layernorm": false, | |
| "model.layers.39.self_attn.q_proj": true, | |
| "model.layers.39.self_attn.k_proj": true, | |
| "model.layers.39.self_attn.v_proj": true, | |
| "model.layers.39.self_attn.o_proj": true, | |
| "model.layers.39.self_attn.rope": false, | |
| "model.layers.39.block_sparse_moe.gate": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.39.input_layernorm": false, | |
| "model.layers.39.post_attention_layernorm": false, | |
| "model.layers.40.self_attn.q_proj": true, | |
| "model.layers.40.self_attn.k_proj": true, | |
| "model.layers.40.self_attn.v_proj": true, | |
| "model.layers.40.self_attn.o_proj": true, | |
| "model.layers.40.self_attn.rope": false, | |
| "model.layers.40.block_sparse_moe.gate": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.40.input_layernorm": false, | |
| "model.layers.40.post_attention_layernorm": false, | |
| "model.layers.41.self_attn.q_proj": true, | |
| "model.layers.41.self_attn.k_proj": true, | |
| "model.layers.41.self_attn.v_proj": true, | |
| "model.layers.41.self_attn.o_proj": true, | |
| "model.layers.41.self_attn.rope": false, | |
| "model.layers.41.block_sparse_moe.gate": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.41.input_layernorm": false, | |
| "model.layers.41.post_attention_layernorm": false, | |
| "model.layers.42.self_attn.q_proj": true, | |
| "model.layers.42.self_attn.k_proj": true, | |
| "model.layers.42.self_attn.v_proj": true, | |
| "model.layers.42.self_attn.o_proj": true, | |
| "model.layers.42.self_attn.rope": false, | |
| "model.layers.42.block_sparse_moe.gate": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.42.input_layernorm": false, | |
| "model.layers.42.post_attention_layernorm": false, | |
| "model.layers.43.self_attn.q_proj": true, | |
| "model.layers.43.self_attn.k_proj": true, | |
| "model.layers.43.self_attn.v_proj": true, | |
| "model.layers.43.self_attn.o_proj": true, | |
| "model.layers.43.self_attn.rope": false, | |
| "model.layers.43.block_sparse_moe.gate": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.43.input_layernorm": false, | |
| "model.layers.43.post_attention_layernorm": false, | |
| "model.layers.44.self_attn.q_proj": true, | |
| "model.layers.44.self_attn.k_proj": true, | |
| "model.layers.44.self_attn.v_proj": true, | |
| "model.layers.44.self_attn.o_proj": true, | |
| "model.layers.44.self_attn.rope": false, | |
| "model.layers.44.block_sparse_moe.gate": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.44.input_layernorm": false, | |
| "model.layers.44.post_attention_layernorm": false, | |
| "model.layers.45.self_attn.q_proj": true, | |
| "model.layers.45.self_attn.k_proj": true, | |
| "model.layers.45.self_attn.v_proj": true, | |
| "model.layers.45.self_attn.o_proj": true, | |
| "model.layers.45.self_attn.rope": false, | |
| "model.layers.45.block_sparse_moe.gate": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.45.input_layernorm": false, | |
| "model.layers.45.post_attention_layernorm": false, | |
| "model.layers.46.self_attn.q_proj": true, | |
| "model.layers.46.self_attn.k_proj": true, | |
| "model.layers.46.self_attn.v_proj": true, | |
| "model.layers.46.self_attn.o_proj": true, | |
| "model.layers.46.self_attn.rope": false, | |
| "model.layers.46.block_sparse_moe.gate": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.46.input_layernorm": false, | |
| "model.layers.46.post_attention_layernorm": false, | |
| "model.layers.47.self_attn.q_proj": true, | |
| "model.layers.47.self_attn.k_proj": true, | |
| "model.layers.47.self_attn.v_proj": true, | |
| "model.layers.47.self_attn.o_proj": true, | |
| "model.layers.47.self_attn.rope": false, | |
| "model.layers.47.block_sparse_moe.gate": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.47.input_layernorm": false, | |
| "model.layers.47.post_attention_layernorm": false, | |
| "model.layers.48.self_attn.q_proj": true, | |
| "model.layers.48.self_attn.k_proj": true, | |
| "model.layers.48.self_attn.v_proj": true, | |
| "model.layers.48.self_attn.o_proj": true, | |
| "model.layers.48.self_attn.rope": false, | |
| "model.layers.48.block_sparse_moe.gate": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.48.input_layernorm": false, | |
| "model.layers.48.post_attention_layernorm": false, | |
| "model.layers.49.self_attn.q_proj": true, | |
| "model.layers.49.self_attn.k_proj": true, | |
| "model.layers.49.self_attn.v_proj": true, | |
| "model.layers.49.self_attn.o_proj": true, | |
| "model.layers.49.self_attn.rope": false, | |
| "model.layers.49.block_sparse_moe.gate": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.49.input_layernorm": false, | |
| "model.layers.49.post_attention_layernorm": false, | |
| "model.layers.50.self_attn.q_proj": true, | |
| "model.layers.50.self_attn.k_proj": true, | |
| "model.layers.50.self_attn.v_proj": true, | |
| "model.layers.50.self_attn.o_proj": true, | |
| "model.layers.50.self_attn.rope": false, | |
| "model.layers.50.block_sparse_moe.gate": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.50.input_layernorm": false, | |
| "model.layers.50.post_attention_layernorm": false, | |
| "model.layers.51.self_attn.q_proj": true, | |
| "model.layers.51.self_attn.k_proj": true, | |
| "model.layers.51.self_attn.v_proj": true, | |
| "model.layers.51.self_attn.o_proj": true, | |
| "model.layers.51.self_attn.rope": false, | |
| "model.layers.51.block_sparse_moe.gate": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.51.input_layernorm": false, | |
| "model.layers.51.post_attention_layernorm": false, | |
| "model.layers.52.self_attn.q_proj": true, | |
| "model.layers.52.self_attn.k_proj": true, | |
| "model.layers.52.self_attn.v_proj": true, | |
| "model.layers.52.self_attn.o_proj": true, | |
| "model.layers.52.self_attn.rope": false, | |
| "model.layers.52.block_sparse_moe.gate": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.52.input_layernorm": false, | |
| "model.layers.52.post_attention_layernorm": false, | |
| "model.layers.53.self_attn.q_proj": true, | |
| "model.layers.53.self_attn.k_proj": true, | |
| "model.layers.53.self_attn.v_proj": true, | |
| "model.layers.53.self_attn.o_proj": true, | |
| "model.layers.53.self_attn.rope": false, | |
| "model.layers.53.block_sparse_moe.gate": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.53.input_layernorm": false, | |
| "model.layers.53.post_attention_layernorm": false, | |
| "model.layers.54.self_attn.q_proj": true, | |
| "model.layers.54.self_attn.k_proj": true, | |
| "model.layers.54.self_attn.v_proj": true, | |
| "model.layers.54.self_attn.o_proj": true, | |
| "model.layers.54.self_attn.rope": false, | |
| "model.layers.54.block_sparse_moe.gate": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.54.input_layernorm": false, | |
| "model.layers.54.post_attention_layernorm": false, | |
| "model.layers.55.self_attn.q_proj": true, | |
| "model.layers.55.self_attn.k_proj": true, | |
| "model.layers.55.self_attn.v_proj": true, | |
| "model.layers.55.self_attn.o_proj": true, | |
| "model.layers.55.self_attn.rope": false, | |
| "model.layers.55.block_sparse_moe.gate": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.55.input_layernorm": false, | |
| "model.layers.55.post_attention_layernorm": false, | |
| "model.norm": false, | |
| "lm_head": true | |
| }, | |
| "quantization_config": { | |
| "group_size": 64, | |
| "bits": 8, | |
| "model.embed_tokens": true, | |
| "model.layers.0.self_attn.q_proj": true, | |
| "model.layers.0.self_attn.k_proj": true, | |
| "model.layers.0.self_attn.v_proj": true, | |
| "model.layers.0.self_attn.o_proj": true, | |
| "model.layers.0.self_attn.rope": false, | |
| "model.layers.0.block_sparse_moe.gate": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.0.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.0.input_layernorm": false, | |
| "model.layers.0.post_attention_layernorm": false, | |
| "model.layers.1.self_attn.q_proj": true, | |
| "model.layers.1.self_attn.k_proj": true, | |
| "model.layers.1.self_attn.v_proj": true, | |
| "model.layers.1.self_attn.o_proj": true, | |
| "model.layers.1.self_attn.rope": false, | |
| "model.layers.1.block_sparse_moe.gate": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.1.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.1.input_layernorm": false, | |
| "model.layers.1.post_attention_layernorm": false, | |
| "model.layers.2.self_attn.q_proj": true, | |
| "model.layers.2.self_attn.k_proj": true, | |
| "model.layers.2.self_attn.v_proj": true, | |
| "model.layers.2.self_attn.o_proj": true, | |
| "model.layers.2.self_attn.rope": false, | |
| "model.layers.2.block_sparse_moe.gate": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.2.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.2.input_layernorm": false, | |
| "model.layers.2.post_attention_layernorm": false, | |
| "model.layers.3.self_attn.q_proj": true, | |
| "model.layers.3.self_attn.k_proj": true, | |
| "model.layers.3.self_attn.v_proj": true, | |
| "model.layers.3.self_attn.o_proj": true, | |
| "model.layers.3.self_attn.rope": false, | |
| "model.layers.3.block_sparse_moe.gate": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.3.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.3.input_layernorm": false, | |
| "model.layers.3.post_attention_layernorm": false, | |
| "model.layers.4.self_attn.q_proj": true, | |
| "model.layers.4.self_attn.k_proj": true, | |
| "model.layers.4.self_attn.v_proj": true, | |
| "model.layers.4.self_attn.o_proj": true, | |
| "model.layers.4.self_attn.rope": false, | |
| "model.layers.4.block_sparse_moe.gate": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.4.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.4.input_layernorm": false, | |
| "model.layers.4.post_attention_layernorm": false, | |
| "model.layers.5.self_attn.q_proj": true, | |
| "model.layers.5.self_attn.k_proj": true, | |
| "model.layers.5.self_attn.v_proj": true, | |
| "model.layers.5.self_attn.o_proj": true, | |
| "model.layers.5.self_attn.rope": false, | |
| "model.layers.5.block_sparse_moe.gate": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.5.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.5.input_layernorm": false, | |
| "model.layers.5.post_attention_layernorm": false, | |
| "model.layers.6.self_attn.q_proj": true, | |
| "model.layers.6.self_attn.k_proj": true, | |
| "model.layers.6.self_attn.v_proj": true, | |
| "model.layers.6.self_attn.o_proj": true, | |
| "model.layers.6.self_attn.rope": false, | |
| "model.layers.6.block_sparse_moe.gate": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.6.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.6.input_layernorm": false, | |
| "model.layers.6.post_attention_layernorm": false, | |
| "model.layers.7.self_attn.q_proj": true, | |
| "model.layers.7.self_attn.k_proj": true, | |
| "model.layers.7.self_attn.v_proj": true, | |
| "model.layers.7.self_attn.o_proj": true, | |
| "model.layers.7.self_attn.rope": false, | |
| "model.layers.7.block_sparse_moe.gate": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.7.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.7.input_layernorm": false, | |
| "model.layers.7.post_attention_layernorm": false, | |
| "model.layers.8.self_attn.q_proj": true, | |
| "model.layers.8.self_attn.k_proj": true, | |
| "model.layers.8.self_attn.v_proj": true, | |
| "model.layers.8.self_attn.o_proj": true, | |
| "model.layers.8.self_attn.rope": false, | |
| "model.layers.8.block_sparse_moe.gate": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.8.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.8.input_layernorm": false, | |
| "model.layers.8.post_attention_layernorm": false, | |
| "model.layers.9.self_attn.q_proj": true, | |
| "model.layers.9.self_attn.k_proj": true, | |
| "model.layers.9.self_attn.v_proj": true, | |
| "model.layers.9.self_attn.o_proj": true, | |
| "model.layers.9.self_attn.rope": false, | |
| "model.layers.9.block_sparse_moe.gate": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.9.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.9.input_layernorm": false, | |
| "model.layers.9.post_attention_layernorm": false, | |
| "model.layers.10.self_attn.q_proj": true, | |
| "model.layers.10.self_attn.k_proj": true, | |
| "model.layers.10.self_attn.v_proj": true, | |
| "model.layers.10.self_attn.o_proj": true, | |
| "model.layers.10.self_attn.rope": false, | |
| "model.layers.10.block_sparse_moe.gate": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.10.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.10.input_layernorm": false, | |
| "model.layers.10.post_attention_layernorm": false, | |
| "model.layers.11.self_attn.q_proj": true, | |
| "model.layers.11.self_attn.k_proj": true, | |
| "model.layers.11.self_attn.v_proj": true, | |
| "model.layers.11.self_attn.o_proj": true, | |
| "model.layers.11.self_attn.rope": false, | |
| "model.layers.11.block_sparse_moe.gate": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.11.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.11.input_layernorm": false, | |
| "model.layers.11.post_attention_layernorm": false, | |
| "model.layers.12.self_attn.q_proj": true, | |
| "model.layers.12.self_attn.k_proj": true, | |
| "model.layers.12.self_attn.v_proj": true, | |
| "model.layers.12.self_attn.o_proj": true, | |
| "model.layers.12.self_attn.rope": false, | |
| "model.layers.12.block_sparse_moe.gate": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.12.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.12.input_layernorm": false, | |
| "model.layers.12.post_attention_layernorm": false, | |
| "model.layers.13.self_attn.q_proj": true, | |
| "model.layers.13.self_attn.k_proj": true, | |
| "model.layers.13.self_attn.v_proj": true, | |
| "model.layers.13.self_attn.o_proj": true, | |
| "model.layers.13.self_attn.rope": false, | |
| "model.layers.13.block_sparse_moe.gate": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.13.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.13.input_layernorm": false, | |
| "model.layers.13.post_attention_layernorm": false, | |
| "model.layers.14.self_attn.q_proj": true, | |
| "model.layers.14.self_attn.k_proj": true, | |
| "model.layers.14.self_attn.v_proj": true, | |
| "model.layers.14.self_attn.o_proj": true, | |
| "model.layers.14.self_attn.rope": false, | |
| "model.layers.14.block_sparse_moe.gate": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.14.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.14.input_layernorm": false, | |
| "model.layers.14.post_attention_layernorm": false, | |
| "model.layers.15.self_attn.q_proj": true, | |
| "model.layers.15.self_attn.k_proj": true, | |
| "model.layers.15.self_attn.v_proj": true, | |
| "model.layers.15.self_attn.o_proj": true, | |
| "model.layers.15.self_attn.rope": false, | |
| "model.layers.15.block_sparse_moe.gate": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.15.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.15.input_layernorm": false, | |
| "model.layers.15.post_attention_layernorm": false, | |
| "model.layers.16.self_attn.q_proj": true, | |
| "model.layers.16.self_attn.k_proj": true, | |
| "model.layers.16.self_attn.v_proj": true, | |
| "model.layers.16.self_attn.o_proj": true, | |
| "model.layers.16.self_attn.rope": false, | |
| "model.layers.16.block_sparse_moe.gate": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.16.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.16.input_layernorm": false, | |
| "model.layers.16.post_attention_layernorm": false, | |
| "model.layers.17.self_attn.q_proj": true, | |
| "model.layers.17.self_attn.k_proj": true, | |
| "model.layers.17.self_attn.v_proj": true, | |
| "model.layers.17.self_attn.o_proj": true, | |
| "model.layers.17.self_attn.rope": false, | |
| "model.layers.17.block_sparse_moe.gate": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.17.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.17.input_layernorm": false, | |
| "model.layers.17.post_attention_layernorm": false, | |
| "model.layers.18.self_attn.q_proj": true, | |
| "model.layers.18.self_attn.k_proj": true, | |
| "model.layers.18.self_attn.v_proj": true, | |
| "model.layers.18.self_attn.o_proj": true, | |
| "model.layers.18.self_attn.rope": false, | |
| "model.layers.18.block_sparse_moe.gate": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.18.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.18.input_layernorm": false, | |
| "model.layers.18.post_attention_layernorm": false, | |
| "model.layers.19.self_attn.q_proj": true, | |
| "model.layers.19.self_attn.k_proj": true, | |
| "model.layers.19.self_attn.v_proj": true, | |
| "model.layers.19.self_attn.o_proj": true, | |
| "model.layers.19.self_attn.rope": false, | |
| "model.layers.19.block_sparse_moe.gate": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.19.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.19.input_layernorm": false, | |
| "model.layers.19.post_attention_layernorm": false, | |
| "model.layers.20.self_attn.q_proj": true, | |
| "model.layers.20.self_attn.k_proj": true, | |
| "model.layers.20.self_attn.v_proj": true, | |
| "model.layers.20.self_attn.o_proj": true, | |
| "model.layers.20.self_attn.rope": false, | |
| "model.layers.20.block_sparse_moe.gate": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.20.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.20.input_layernorm": false, | |
| "model.layers.20.post_attention_layernorm": false, | |
| "model.layers.21.self_attn.q_proj": true, | |
| "model.layers.21.self_attn.k_proj": true, | |
| "model.layers.21.self_attn.v_proj": true, | |
| "model.layers.21.self_attn.o_proj": true, | |
| "model.layers.21.self_attn.rope": false, | |
| "model.layers.21.block_sparse_moe.gate": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.21.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.21.input_layernorm": false, | |
| "model.layers.21.post_attention_layernorm": false, | |
| "model.layers.22.self_attn.q_proj": true, | |
| "model.layers.22.self_attn.k_proj": true, | |
| "model.layers.22.self_attn.v_proj": true, | |
| "model.layers.22.self_attn.o_proj": true, | |
| "model.layers.22.self_attn.rope": false, | |
| "model.layers.22.block_sparse_moe.gate": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.22.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.22.input_layernorm": false, | |
| "model.layers.22.post_attention_layernorm": false, | |
| "model.layers.23.self_attn.q_proj": true, | |
| "model.layers.23.self_attn.k_proj": true, | |
| "model.layers.23.self_attn.v_proj": true, | |
| "model.layers.23.self_attn.o_proj": true, | |
| "model.layers.23.self_attn.rope": false, | |
| "model.layers.23.block_sparse_moe.gate": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.23.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.23.input_layernorm": false, | |
| "model.layers.23.post_attention_layernorm": false, | |
| "model.layers.24.self_attn.q_proj": true, | |
| "model.layers.24.self_attn.k_proj": true, | |
| "model.layers.24.self_attn.v_proj": true, | |
| "model.layers.24.self_attn.o_proj": true, | |
| "model.layers.24.self_attn.rope": false, | |
| "model.layers.24.block_sparse_moe.gate": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.24.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.24.input_layernorm": false, | |
| "model.layers.24.post_attention_layernorm": false, | |
| "model.layers.25.self_attn.q_proj": true, | |
| "model.layers.25.self_attn.k_proj": true, | |
| "model.layers.25.self_attn.v_proj": true, | |
| "model.layers.25.self_attn.o_proj": true, | |
| "model.layers.25.self_attn.rope": false, | |
| "model.layers.25.block_sparse_moe.gate": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.25.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.25.input_layernorm": false, | |
| "model.layers.25.post_attention_layernorm": false, | |
| "model.layers.26.self_attn.q_proj": true, | |
| "model.layers.26.self_attn.k_proj": true, | |
| "model.layers.26.self_attn.v_proj": true, | |
| "model.layers.26.self_attn.o_proj": true, | |
| "model.layers.26.self_attn.rope": false, | |
| "model.layers.26.block_sparse_moe.gate": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.26.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.26.input_layernorm": false, | |
| "model.layers.26.post_attention_layernorm": false, | |
| "model.layers.27.self_attn.q_proj": true, | |
| "model.layers.27.self_attn.k_proj": true, | |
| "model.layers.27.self_attn.v_proj": true, | |
| "model.layers.27.self_attn.o_proj": true, | |
| "model.layers.27.self_attn.rope": false, | |
| "model.layers.27.block_sparse_moe.gate": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.27.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.27.input_layernorm": false, | |
| "model.layers.27.post_attention_layernorm": false, | |
| "model.layers.28.self_attn.q_proj": true, | |
| "model.layers.28.self_attn.k_proj": true, | |
| "model.layers.28.self_attn.v_proj": true, | |
| "model.layers.28.self_attn.o_proj": true, | |
| "model.layers.28.self_attn.rope": false, | |
| "model.layers.28.block_sparse_moe.gate": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.28.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.28.input_layernorm": false, | |
| "model.layers.28.post_attention_layernorm": false, | |
| "model.layers.29.self_attn.q_proj": true, | |
| "model.layers.29.self_attn.k_proj": true, | |
| "model.layers.29.self_attn.v_proj": true, | |
| "model.layers.29.self_attn.o_proj": true, | |
| "model.layers.29.self_attn.rope": false, | |
| "model.layers.29.block_sparse_moe.gate": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.29.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.29.input_layernorm": false, | |
| "model.layers.29.post_attention_layernorm": false, | |
| "model.layers.30.self_attn.q_proj": true, | |
| "model.layers.30.self_attn.k_proj": true, | |
| "model.layers.30.self_attn.v_proj": true, | |
| "model.layers.30.self_attn.o_proj": true, | |
| "model.layers.30.self_attn.rope": false, | |
| "model.layers.30.block_sparse_moe.gate": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.30.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.30.input_layernorm": false, | |
| "model.layers.30.post_attention_layernorm": false, | |
| "model.layers.31.self_attn.q_proj": true, | |
| "model.layers.31.self_attn.k_proj": true, | |
| "model.layers.31.self_attn.v_proj": true, | |
| "model.layers.31.self_attn.o_proj": true, | |
| "model.layers.31.self_attn.rope": false, | |
| "model.layers.31.block_sparse_moe.gate": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.31.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.31.input_layernorm": false, | |
| "model.layers.31.post_attention_layernorm": false, | |
| "model.layers.32.self_attn.q_proj": true, | |
| "model.layers.32.self_attn.k_proj": true, | |
| "model.layers.32.self_attn.v_proj": true, | |
| "model.layers.32.self_attn.o_proj": true, | |
| "model.layers.32.self_attn.rope": false, | |
| "model.layers.32.block_sparse_moe.gate": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.32.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.32.input_layernorm": false, | |
| "model.layers.32.post_attention_layernorm": false, | |
| "model.layers.33.self_attn.q_proj": true, | |
| "model.layers.33.self_attn.k_proj": true, | |
| "model.layers.33.self_attn.v_proj": true, | |
| "model.layers.33.self_attn.o_proj": true, | |
| "model.layers.33.self_attn.rope": false, | |
| "model.layers.33.block_sparse_moe.gate": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.33.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.33.input_layernorm": false, | |
| "model.layers.33.post_attention_layernorm": false, | |
| "model.layers.34.self_attn.q_proj": true, | |
| "model.layers.34.self_attn.k_proj": true, | |
| "model.layers.34.self_attn.v_proj": true, | |
| "model.layers.34.self_attn.o_proj": true, | |
| "model.layers.34.self_attn.rope": false, | |
| "model.layers.34.block_sparse_moe.gate": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.34.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.34.input_layernorm": false, | |
| "model.layers.34.post_attention_layernorm": false, | |
| "model.layers.35.self_attn.q_proj": true, | |
| "model.layers.35.self_attn.k_proj": true, | |
| "model.layers.35.self_attn.v_proj": true, | |
| "model.layers.35.self_attn.o_proj": true, | |
| "model.layers.35.self_attn.rope": false, | |
| "model.layers.35.block_sparse_moe.gate": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.35.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.35.input_layernorm": false, | |
| "model.layers.35.post_attention_layernorm": false, | |
| "model.layers.36.self_attn.q_proj": true, | |
| "model.layers.36.self_attn.k_proj": true, | |
| "model.layers.36.self_attn.v_proj": true, | |
| "model.layers.36.self_attn.o_proj": true, | |
| "model.layers.36.self_attn.rope": false, | |
| "model.layers.36.block_sparse_moe.gate": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.36.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.36.input_layernorm": false, | |
| "model.layers.36.post_attention_layernorm": false, | |
| "model.layers.37.self_attn.q_proj": true, | |
| "model.layers.37.self_attn.k_proj": true, | |
| "model.layers.37.self_attn.v_proj": true, | |
| "model.layers.37.self_attn.o_proj": true, | |
| "model.layers.37.self_attn.rope": false, | |
| "model.layers.37.block_sparse_moe.gate": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.37.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.37.input_layernorm": false, | |
| "model.layers.37.post_attention_layernorm": false, | |
| "model.layers.38.self_attn.q_proj": true, | |
| "model.layers.38.self_attn.k_proj": true, | |
| "model.layers.38.self_attn.v_proj": true, | |
| "model.layers.38.self_attn.o_proj": true, | |
| "model.layers.38.self_attn.rope": false, | |
| "model.layers.38.block_sparse_moe.gate": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.38.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.38.input_layernorm": false, | |
| "model.layers.38.post_attention_layernorm": false, | |
| "model.layers.39.self_attn.q_proj": true, | |
| "model.layers.39.self_attn.k_proj": true, | |
| "model.layers.39.self_attn.v_proj": true, | |
| "model.layers.39.self_attn.o_proj": true, | |
| "model.layers.39.self_attn.rope": false, | |
| "model.layers.39.block_sparse_moe.gate": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.39.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.39.input_layernorm": false, | |
| "model.layers.39.post_attention_layernorm": false, | |
| "model.layers.40.self_attn.q_proj": true, | |
| "model.layers.40.self_attn.k_proj": true, | |
| "model.layers.40.self_attn.v_proj": true, | |
| "model.layers.40.self_attn.o_proj": true, | |
| "model.layers.40.self_attn.rope": false, | |
| "model.layers.40.block_sparse_moe.gate": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.40.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.40.input_layernorm": false, | |
| "model.layers.40.post_attention_layernorm": false, | |
| "model.layers.41.self_attn.q_proj": true, | |
| "model.layers.41.self_attn.k_proj": true, | |
| "model.layers.41.self_attn.v_proj": true, | |
| "model.layers.41.self_attn.o_proj": true, | |
| "model.layers.41.self_attn.rope": false, | |
| "model.layers.41.block_sparse_moe.gate": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.41.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.41.input_layernorm": false, | |
| "model.layers.41.post_attention_layernorm": false, | |
| "model.layers.42.self_attn.q_proj": true, | |
| "model.layers.42.self_attn.k_proj": true, | |
| "model.layers.42.self_attn.v_proj": true, | |
| "model.layers.42.self_attn.o_proj": true, | |
| "model.layers.42.self_attn.rope": false, | |
| "model.layers.42.block_sparse_moe.gate": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.42.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.42.input_layernorm": false, | |
| "model.layers.42.post_attention_layernorm": false, | |
| "model.layers.43.self_attn.q_proj": true, | |
| "model.layers.43.self_attn.k_proj": true, | |
| "model.layers.43.self_attn.v_proj": true, | |
| "model.layers.43.self_attn.o_proj": true, | |
| "model.layers.43.self_attn.rope": false, | |
| "model.layers.43.block_sparse_moe.gate": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.43.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.43.input_layernorm": false, | |
| "model.layers.43.post_attention_layernorm": false, | |
| "model.layers.44.self_attn.q_proj": true, | |
| "model.layers.44.self_attn.k_proj": true, | |
| "model.layers.44.self_attn.v_proj": true, | |
| "model.layers.44.self_attn.o_proj": true, | |
| "model.layers.44.self_attn.rope": false, | |
| "model.layers.44.block_sparse_moe.gate": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.44.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.44.input_layernorm": false, | |
| "model.layers.44.post_attention_layernorm": false, | |
| "model.layers.45.self_attn.q_proj": true, | |
| "model.layers.45.self_attn.k_proj": true, | |
| "model.layers.45.self_attn.v_proj": true, | |
| "model.layers.45.self_attn.o_proj": true, | |
| "model.layers.45.self_attn.rope": false, | |
| "model.layers.45.block_sparse_moe.gate": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.45.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.45.input_layernorm": false, | |
| "model.layers.45.post_attention_layernorm": false, | |
| "model.layers.46.self_attn.q_proj": true, | |
| "model.layers.46.self_attn.k_proj": true, | |
| "model.layers.46.self_attn.v_proj": true, | |
| "model.layers.46.self_attn.o_proj": true, | |
| "model.layers.46.self_attn.rope": false, | |
| "model.layers.46.block_sparse_moe.gate": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.46.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.46.input_layernorm": false, | |
| "model.layers.46.post_attention_layernorm": false, | |
| "model.layers.47.self_attn.q_proj": true, | |
| "model.layers.47.self_attn.k_proj": true, | |
| "model.layers.47.self_attn.v_proj": true, | |
| "model.layers.47.self_attn.o_proj": true, | |
| "model.layers.47.self_attn.rope": false, | |
| "model.layers.47.block_sparse_moe.gate": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.47.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.47.input_layernorm": false, | |
| "model.layers.47.post_attention_layernorm": false, | |
| "model.layers.48.self_attn.q_proj": true, | |
| "model.layers.48.self_attn.k_proj": true, | |
| "model.layers.48.self_attn.v_proj": true, | |
| "model.layers.48.self_attn.o_proj": true, | |
| "model.layers.48.self_attn.rope": false, | |
| "model.layers.48.block_sparse_moe.gate": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.48.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.48.input_layernorm": false, | |
| "model.layers.48.post_attention_layernorm": false, | |
| "model.layers.49.self_attn.q_proj": true, | |
| "model.layers.49.self_attn.k_proj": true, | |
| "model.layers.49.self_attn.v_proj": true, | |
| "model.layers.49.self_attn.o_proj": true, | |
| "model.layers.49.self_attn.rope": false, | |
| "model.layers.49.block_sparse_moe.gate": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.49.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.49.input_layernorm": false, | |
| "model.layers.49.post_attention_layernorm": false, | |
| "model.layers.50.self_attn.q_proj": true, | |
| "model.layers.50.self_attn.k_proj": true, | |
| "model.layers.50.self_attn.v_proj": true, | |
| "model.layers.50.self_attn.o_proj": true, | |
| "model.layers.50.self_attn.rope": false, | |
| "model.layers.50.block_sparse_moe.gate": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.50.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.50.input_layernorm": false, | |
| "model.layers.50.post_attention_layernorm": false, | |
| "model.layers.51.self_attn.q_proj": true, | |
| "model.layers.51.self_attn.k_proj": true, | |
| "model.layers.51.self_attn.v_proj": true, | |
| "model.layers.51.self_attn.o_proj": true, | |
| "model.layers.51.self_attn.rope": false, | |
| "model.layers.51.block_sparse_moe.gate": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.51.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.51.input_layernorm": false, | |
| "model.layers.51.post_attention_layernorm": false, | |
| "model.layers.52.self_attn.q_proj": true, | |
| "model.layers.52.self_attn.k_proj": true, | |
| "model.layers.52.self_attn.v_proj": true, | |
| "model.layers.52.self_attn.o_proj": true, | |
| "model.layers.52.self_attn.rope": false, | |
| "model.layers.52.block_sparse_moe.gate": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.52.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.52.input_layernorm": false, | |
| "model.layers.52.post_attention_layernorm": false, | |
| "model.layers.53.self_attn.q_proj": true, | |
| "model.layers.53.self_attn.k_proj": true, | |
| "model.layers.53.self_attn.v_proj": true, | |
| "model.layers.53.self_attn.o_proj": true, | |
| "model.layers.53.self_attn.rope": false, | |
| "model.layers.53.block_sparse_moe.gate": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.53.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.53.input_layernorm": false, | |
| "model.layers.53.post_attention_layernorm": false, | |
| "model.layers.54.self_attn.q_proj": true, | |
| "model.layers.54.self_attn.k_proj": true, | |
| "model.layers.54.self_attn.v_proj": true, | |
| "model.layers.54.self_attn.o_proj": true, | |
| "model.layers.54.self_attn.rope": false, | |
| "model.layers.54.block_sparse_moe.gate": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.54.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.54.input_layernorm": false, | |
| "model.layers.54.post_attention_layernorm": false, | |
| "model.layers.55.self_attn.q_proj": true, | |
| "model.layers.55.self_attn.k_proj": true, | |
| "model.layers.55.self_attn.v_proj": true, | |
| "model.layers.55.self_attn.o_proj": true, | |
| "model.layers.55.self_attn.rope": false, | |
| "model.layers.55.block_sparse_moe.gate": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.gate_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.up_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.down_proj": true, | |
| "model.layers.55.block_sparse_moe.switch_mlp.activation": false, | |
| "model.layers.55.input_layernorm": false, | |
| "model.layers.55.post_attention_layernorm": false, | |
| "model.norm": false, | |
| "lm_head": true | |
| }, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 1000000.0, | |
| "router_aux_loss_coef": 0.001, | |
| "router_jitter_noise": 0.0, | |
| "sliding_window": null, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.52.0.dev0", | |
| "use_cache": true, | |
| "vocab_size": 32768 | |
| } |