Upload folder using huggingface_hub

README.md

@@ -7,31 +7,3 @@ base_model: Qwen/Qwen3-14B
 tags:
 - mlx
 ---
-
-# mlx-community/Qwen3-14B-4bit-AWQ
-
-This model [mlx-community/Qwen3-14B-4bit-AWQ](https://huggingface.co/mlx-community/Qwen3-14B-AWQ-MLX) was
-converted to MLX format from [Qwen/Qwen3-14B](https://huggingface.co/Qwen/Qwen3-14B)
-using mlx-lm version **0.24.0**.
-
-## Use with mlx
-
-```bash
-pip install mlx-lm
-```
-
-```python
-from mlx_lm import load, generate
-
-model, tokenizer = load("mlx-community/Qwen3-14B-4bit-AWQ")
-
-prompt = "hello"
-
-if tokenizer.chat_template is not None:
-    messages = [{"role": "user", "content": prompt}]
-    prompt = tokenizer.apply_chat_template(
-        messages, add_generation_prompt=True
-    )
-
-response = generate(model, tokenizer, prompt=prompt, verbose=True)
-```

config.json

@@ -25,1321 +25,1321 @@
             "bits": 4
         },
         "model.layers.0.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.self_attn.q_norm": false,
         "model.layers.0.self_attn.k_norm": false,
         "model.layers.0.self_attn.rope": false,
         "model.layers.0.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.0.input_layernorm": false,
         "model.layers.0.post_attention_layernorm": false,
         "model.layers.1.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.self_attn.q_norm": false,
         "model.layers.1.self_attn.k_norm": false,
         "model.layers.1.self_attn.rope": false,
         "model.layers.1.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.1.input_layernorm": false,
         "model.layers.1.post_attention_layernorm": false,
         "model.layers.2.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.self_attn.q_norm": false,
         "model.layers.2.self_attn.k_norm": false,
         "model.layers.2.self_attn.rope": false,
         "model.layers.2.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.2.input_layernorm": false,
         "model.layers.2.post_attention_layernorm": false,
         "model.layers.3.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.self_attn.q_norm": false,
         "model.layers.3.self_attn.k_norm": false,
         "model.layers.3.self_attn.rope": false,
         "model.layers.3.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.3.input_layernorm": false,
         "model.layers.3.post_attention_layernorm": false,
         "model.layers.4.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.self_attn.q_norm": false,
         "model.layers.4.self_attn.k_norm": false,
         "model.layers.4.self_attn.rope": false,
         "model.layers.4.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.4.input_layernorm": false,
         "model.layers.4.post_attention_layernorm": false,
         "model.layers.5.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.self_attn.q_norm": false,
         "model.layers.5.self_attn.k_norm": false,
         "model.layers.5.self_attn.rope": false,
         "model.layers.5.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.5.input_layernorm": false,
         "model.layers.5.post_attention_layernorm": false,
         "model.layers.6.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.self_attn.q_norm": false,
         "model.layers.6.self_attn.k_norm": false,
         "model.layers.6.self_attn.rope": false,
         "model.layers.6.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.6.input_layernorm": false,
         "model.layers.6.post_attention_layernorm": false,
         "model.layers.7.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.self_attn.q_norm": false,
         "model.layers.7.self_attn.k_norm": false,
         "model.layers.7.self_attn.rope": false,
         "model.layers.7.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.7.input_layernorm": false,
         "model.layers.7.post_attention_layernorm": false,
         "model.layers.8.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.self_attn.q_norm": false,
         "model.layers.8.self_attn.k_norm": false,
         "model.layers.8.self_attn.rope": false,
         "model.layers.8.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.8.input_layernorm": false,
         "model.layers.8.post_attention_layernorm": false,
         "model.layers.9.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.self_attn.q_norm": false,
         "model.layers.9.self_attn.k_norm": false,
         "model.layers.9.self_attn.rope": false,
         "model.layers.9.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.9.input_layernorm": false,
         "model.layers.9.post_attention_layernorm": false,
         "model.layers.10.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.self_attn.q_norm": false,
         "model.layers.10.self_attn.k_norm": false,
         "model.layers.10.self_attn.rope": false,
         "model.layers.10.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.10.input_layernorm": false,
         "model.layers.10.post_attention_layernorm": false,
         "model.layers.11.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.self_attn.q_norm": false,
         "model.layers.11.self_attn.k_norm": false,
         "model.layers.11.self_attn.rope": false,
         "model.layers.11.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.11.input_layernorm": false,
         "model.layers.11.post_attention_layernorm": false,
         "model.layers.12.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.self_attn.q_norm": false,
         "model.layers.12.self_attn.k_norm": false,
         "model.layers.12.self_attn.rope": false,
         "model.layers.12.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.12.input_layernorm": false,
         "model.layers.12.post_attention_layernorm": false,
         "model.layers.13.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.self_attn.q_norm": false,
         "model.layers.13.self_attn.k_norm": false,
         "model.layers.13.self_attn.rope": false,
         "model.layers.13.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.13.input_layernorm": false,
         "model.layers.13.post_attention_layernorm": false,
         "model.layers.14.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.self_attn.q_norm": false,
         "model.layers.14.self_attn.k_norm": false,
         "model.layers.14.self_attn.rope": false,
         "model.layers.14.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.14.input_layernorm": false,
         "model.layers.14.post_attention_layernorm": false,
         "model.layers.15.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.self_attn.q_norm": false,
         "model.layers.15.self_attn.k_norm": false,
         "model.layers.15.self_attn.rope": false,
         "model.layers.15.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.15.input_layernorm": false,
         "model.layers.15.post_attention_layernorm": false,
         "model.layers.16.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.self_attn.q_norm": false,
         "model.layers.16.self_attn.k_norm": false,
         "model.layers.16.self_attn.rope": false,
         "model.layers.16.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.16.input_layernorm": false,
         "model.layers.16.post_attention_layernorm": false,
         "model.layers.17.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.self_attn.q_norm": false,
         "model.layers.17.self_attn.k_norm": false,
         "model.layers.17.self_attn.rope": false,
         "model.layers.17.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.17.input_layernorm": false,
         "model.layers.17.post_attention_layernorm": false,
         "model.layers.18.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.self_attn.q_norm": false,
         "model.layers.18.self_attn.k_norm": false,
         "model.layers.18.self_attn.rope": false,
         "model.layers.18.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.18.input_layernorm": false,
         "model.layers.18.post_attention_layernorm": false,
         "model.layers.19.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.self_attn.q_norm": false,
         "model.layers.19.self_attn.k_norm": false,
         "model.layers.19.self_attn.rope": false,
         "model.layers.19.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.19.input_layernorm": false,
         "model.layers.19.post_attention_layernorm": false,
         "model.layers.20.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.self_attn.q_norm": false,
         "model.layers.20.self_attn.k_norm": false,
         "model.layers.20.self_attn.rope": false,
         "model.layers.20.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.20.input_layernorm": false,
         "model.layers.20.post_attention_layernorm": false,
         "model.layers.21.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.self_attn.q_norm": false,
         "model.layers.21.self_attn.k_norm": false,
         "model.layers.21.self_attn.rope": false,
         "model.layers.21.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.21.input_layernorm": false,
         "model.layers.21.post_attention_layernorm": false,
         "model.layers.22.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.self_attn.q_norm": false,
         "model.layers.22.self_attn.k_norm": false,
         "model.layers.22.self_attn.rope": false,
         "model.layers.22.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.22.input_layernorm": false,
         "model.layers.22.post_attention_layernorm": false,
         "model.layers.23.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.self_attn.q_norm": false,
         "model.layers.23.self_attn.k_norm": false,
         "model.layers.23.self_attn.rope": false,
         "model.layers.23.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.23.input_layernorm": false,
         "model.layers.23.post_attention_layernorm": false,
         "model.layers.24.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.self_attn.q_norm": false,
         "model.layers.24.self_attn.k_norm": false,
         "model.layers.24.self_attn.rope": false,
         "model.layers.24.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.24.input_layernorm": false,
         "model.layers.24.post_attention_layernorm": false,
         "model.layers.25.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.self_attn.q_norm": false,
         "model.layers.25.self_attn.k_norm": false,
         "model.layers.25.self_attn.rope": false,
         "model.layers.25.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.25.input_layernorm": false,
         "model.layers.25.post_attention_layernorm": false,
         "model.layers.26.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.self_attn.q_norm": false,
         "model.layers.26.self_attn.k_norm": false,
         "model.layers.26.self_attn.rope": false,
         "model.layers.26.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.26.input_layernorm": false,
         "model.layers.26.post_attention_layernorm": false,
         "model.layers.27.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.self_attn.q_norm": false,
         "model.layers.27.self_attn.k_norm": false,
         "model.layers.27.self_attn.rope": false,
         "model.layers.27.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.27.input_layernorm": false,
         "model.layers.27.post_attention_layernorm": false,
         "model.layers.28.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.self_attn.q_norm": false,
         "model.layers.28.self_attn.k_norm": false,
         "model.layers.28.self_attn.rope": false,
         "model.layers.28.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.28.input_layernorm": false,
         "model.layers.28.post_attention_layernorm": false,
         "model.layers.29.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.self_attn.q_norm": false,
         "model.layers.29.self_attn.k_norm": false,
         "model.layers.29.self_attn.rope": false,
         "model.layers.29.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.29.input_layernorm": false,
         "model.layers.29.post_attention_layernorm": false,
         "model.layers.30.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.self_attn.q_norm": false,
         "model.layers.30.self_attn.k_norm": false,
         "model.layers.30.self_attn.rope": false,
         "model.layers.30.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.30.input_layernorm": false,
         "model.layers.30.post_attention_layernorm": false,
         "model.layers.31.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.self_attn.q_norm": false,
         "model.layers.31.self_attn.k_norm": false,
         "model.layers.31.self_attn.rope": false,
         "model.layers.31.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.31.input_layernorm": false,
         "model.layers.31.post_attention_layernorm": false,
         "model.layers.32.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.self_attn.q_norm": false,
         "model.layers.32.self_attn.k_norm": false,
         "model.layers.32.self_attn.rope": false,
         "model.layers.32.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.32.input_layernorm": false,
         "model.layers.32.post_attention_layernorm": false,
         "model.layers.33.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.self_attn.q_norm": false,
         "model.layers.33.self_attn.k_norm": false,
         "model.layers.33.self_attn.rope": false,
         "model.layers.33.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.33.input_layernorm": false,
         "model.layers.33.post_attention_layernorm": false,
         "model.layers.34.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.self_attn.q_norm": false,
         "model.layers.34.self_attn.k_norm": false,
         "model.layers.34.self_attn.rope": false,
         "model.layers.34.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.34.input_layernorm": false,
         "model.layers.34.post_attention_layernorm": false,
         "model.layers.35.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.self_attn.q_norm": false,
         "model.layers.35.self_attn.k_norm": false,
         "model.layers.35.self_attn.rope": false,
         "model.layers.35.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.35.input_layernorm": false,
         "model.layers.35.post_attention_layernorm": false,
         "model.layers.36.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.self_attn.q_norm": false,
         "model.layers.36.self_attn.k_norm": false,
         "model.layers.36.self_attn.rope": false,
         "model.layers.36.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.36.input_layernorm": false,
         "model.layers.36.post_attention_layernorm": false,
         "model.layers.37.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.self_attn.q_norm": false,
         "model.layers.37.self_attn.k_norm": false,
         "model.layers.37.self_attn.rope": false,
         "model.layers.37.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.37.input_layernorm": false,
         "model.layers.37.post_attention_layernorm": false,
         "model.layers.38.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.self_attn.q_norm": false,
         "model.layers.38.self_attn.k_norm": false,
         "model.layers.38.self_attn.rope": false,
         "model.layers.38.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.38.input_layernorm": false,
         "model.layers.38.post_attention_layernorm": false,
         "model.layers.39.self_attn.q_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.self_attn.k_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.self_attn.v_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.self_attn.o_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.self_attn.q_norm": false,
         "model.layers.39.self_attn.k_norm": false,
         "model.layers.39.self_attn.rope": false,
         "model.layers.39.mlp.gate_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.mlp.down_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.mlp.up_proj": {
-            "group_size": 128,
+            "group_size": 64,
             "bits": 4
         },
         "model.layers.39.input_layernorm": false,

model.safetensors.index.json

@@ -1,6 +1,6 @@
 {
     "metadata": {
-        "total_size": 7992145920
+        "total_size": 8405022720
     },
     "weight_map": {
         "lm_head.biases": "model-00002-of-00002.safetensors",
@@ -484,17 +484,17 @@
         "model.layers.25.self_attn.v_proj.biases": "model-00001-of-00002.safetensors",
         "model.layers.25.self_attn.v_proj.scales": "model-00001-of-00002.safetensors",
         "model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.down_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.down_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.gate_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.gate_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.up_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.up_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
+        "model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.down_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.down_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.gate_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.gate_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.up_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.up_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.26.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
         "model.layers.26.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
         "model.layers.26.self_attn.k_proj.biases": "model-00001-of-00002.safetensors",
         "model.layers.26.self_attn.k_proj.scales": "model-00001-of-00002.safetensors",
@@ -510,30 +510,30 @@
         "model.layers.26.self_attn.v_proj.scales": "model-00001-of-00002.safetensors",
         "model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
         "model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
-        "model.layers.27.mlp.down_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.mlp.down_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.mlp.gate_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.mlp.gate_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
+        "model.layers.27.mlp.down_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.mlp.down_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.mlp.gate_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.mlp.gate_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
         "model.layers.27.mlp.up_proj.biases": "model-00002-of-00002.safetensors",
         "model.layers.27.mlp.up_proj.scales": "model-00002-of-00002.safetensors",
         "model.layers.27.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
         "model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
-        "model.layers.27.self_attn.k_norm.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.k_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.k_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.o_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.o_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.q_norm.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.q_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.q_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.v_proj.biases": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.v_proj.scales": "model-00001-of-00002.safetensors",
-        "model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
+        "model.layers.27.self_attn.k_norm.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.k_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.k_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.o_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.o_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.q_norm.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.q_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.q_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.v_proj.biases": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.v_proj.scales": "model-00002-of-00002.safetensors",
+        "model.layers.27.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
         "model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
         "model.layers.28.mlp.down_proj.biases": "model-00002-of-00002.safetensors",
         "model.layers.28.mlp.down_proj.scales": "model-00002-of-00002.safetensors",

tokenizer_config.json

@@ -227,7 +227,9 @@
     "<|video_pad|>"
   ],
   "bos_token": null,
-  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- set tool_start = \"<tool_response>\" %}\n    {%- set tool_start_length = tool_start|length %}\n    {%- set start_of_message = message.content[:tool_start_length] %}\n    {%- set tool_end = \"</tool_response>\" %}\n    {%- set tool_end_length = tool_end|length %}\n    {%- set start_pos = (message.content|length) - tool_end_length %}\n    {%- if start_pos < 0 %}\n        {%- set start_pos = 0 %}\n    {%- endif %}\n    {%- set end_of_message = message.content[start_pos:] %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and not(start_of_message == tool_start and end_of_message == tool_end) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in message.content %}\n                {%- set content = (message.content.split('</think>')|last).lstrip('\\n') %}\n        {%- set reasoning_content = (message.content.split('</think>')|first).rstrip('\\n') %}\n        {%- set reasoning_content = (reasoning_content.split('<think>')|last).lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",  "clean_up_tokenization_spaces": false,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0].role == 'system' %}\n        {{- messages[0].content + '\\n\\n' }}\n    {%- endif %}\n    {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0].role == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for forward_message in messages %}\n    {%- set index = (messages|length - 1) - loop.index0 %}\n    {%- set message = messages[index] %}\n    {%- set current_content = message.content if message.content is not none else '' %}\n    {%- set tool_start = '<tool_response>' %}\n    {%- set tool_start_length = tool_start|length %}\n    {%- set start_of_message = current_content[:tool_start_length] %}\n    {%- set tool_end = '</tool_response>' %}\n    {%- set tool_end_length = tool_end|length %}\n    {%- set start_pos = (current_content|length) - tool_end_length %}\n    {%- if start_pos < 0 %}\n        {%- set start_pos = 0 %}\n    {%- endif %}\n    {%- set end_of_message = current_content[start_pos:] %}\n    {%- if ns.multi_step_tool and message.role == \"user\" and not(start_of_message == tool_start and end_of_message == tool_end) %}\n        {%- set ns.multi_step_tool = false %}\n        {%- set ns.last_query_index = index %}\n    {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {%- set content = message.content %}\n        {%- set reasoning_content = '' %}\n        {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n            {%- set reasoning_content = message.reasoning_content %}\n        {%- else %}\n            {%- if '</think>' in message.content %}\n                {%- set content = (message.content.split('</think>')|last).lstrip('\\n') %}\n                {%- set reasoning_content = (message.content.split('</think>')|first).rstrip('\\n') %}\n                {%- set reasoning_content = (reasoning_content.split('<think>')|last).lstrip('\\n') %}\n            {%- endif %}\n        {%- endif %}\n        {%- if loop.index0 > ns.last_query_index %}\n            {%- if loop.last or (not loop.last and reasoning_content) %}\n                {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n            {%- else %}\n                {{- '<|im_start|>' + message.role + '\\n' + content }}\n            {%- endif %}\n        {%- else %}\n            {{- '<|im_start|>' + message.role + '\\n' + content }}\n        {%- endif %}\n        {%- if message.tool_calls %}\n            {%- for tool_call in message.tool_calls %}\n                {%- if (loop.first and content) or (not loop.first) %}\n                    {{- '\\n' }}\n                {%- endif %}\n                {%- if tool_call.function %}\n                    {%- set tool_call = tool_call.function %}\n                {%- endif %}\n                {{- '<tool_call>\\n{\"name\": \"' }}\n                {{- tool_call.name }}\n                {{- '\", \"arguments\": ' }}\n                {%- if tool_call.arguments is string %}\n                    {{- tool_call.arguments }}\n                {%- else %}\n                    {{- tool_call.arguments | tojson }}\n                {%- endif %}\n                {{- '}\\n</tool_call>' }}\n            {%- endfor %}\n        {%- endif %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n    {%- if enable_thinking is defined and enable_thinking is false %}\n        {{- '<think>\\n\\n</think>\\n\\n' }}\n    {%- endif %}\n{%- endif %}",
+
+  "clean_up_tokenization_spaces": false,
   "eos_token": "<|im_end|>",
   "errors": "replace",
   "extra_special_tokens": {},