danielhanchen commited on
Commit
cdb28d2
·
verified ·
1 Parent(s): 47a64ee

Add files using upload-large-folder tool

Browse files
config.json CHANGED
@@ -31,37 +31,36 @@
31
  "multi_modal_projector",
32
  "merger",
33
  "modality_projection",
34
- "model.layers.6.self_attn",
35
  "model.layers.54.mlp",
36
  "model.layers.50.mlp",
37
- "model.layers.46.mlp",
38
  "model.layers.23.mlp",
 
39
  "model.layers.49.mlp",
40
- "model.layers.51.mlp",
41
  "model.layers.48.mlp",
42
- "model.layers.24.mlp",
43
  "model.layers.47.mlp",
44
- "model.layers.53.mlp",
 
45
  "model.layers.52.mlp",
 
46
  "model.layers.43.mlp",
47
- "model.layers.44.mlp",
48
- "model.layers.40.mlp",
49
- "model.layers.42.mlp",
50
- "model.layers.27.mlp",
51
  "model.layers.45.mlp",
 
52
  "model.layers.39.mlp",
53
- "model.layers.41.mlp",
54
  "model.layers.34.mlp",
 
55
  "model.layers.37.mlp",
56
- "model.layers.26.mlp",
57
- "model.layers.32.mlp",
58
  "model.layers.30.mlp",
59
- "model.layers.35.mlp",
60
- "model.layers.38.mlp",
61
  "model.layers.33.mlp",
62
- "model.layers.31.mlp",
 
 
63
  "model.layers.36.mlp",
64
- "model.layers.25.mlp",
 
65
  "model.layers.28.mlp",
66
  "model.layers.29.mlp",
67
  "model.layers.62.mlp",
@@ -80,7 +79,7 @@
80
  "sliding_window": null,
81
  "tie_word_embeddings": false,
82
  "torch_dtype": "bfloat16",
83
- "transformers_version": "4.52.0.dev0",
84
  "unsloth_fixed": true,
85
  "use_cache": true,
86
  "use_sliding_window": false,
 
31
  "multi_modal_projector",
32
  "merger",
33
  "modality_projection",
 
34
  "model.layers.54.mlp",
35
  "model.layers.50.mlp",
 
36
  "model.layers.23.mlp",
37
+ "model.layers.46.mlp",
38
  "model.layers.49.mlp",
 
39
  "model.layers.48.mlp",
 
40
  "model.layers.47.mlp",
41
+ "model.layers.51.mlp",
42
+ "model.layers.44.mlp",
43
  "model.layers.52.mlp",
44
+ "model.layers.53.mlp",
45
  "model.layers.43.mlp",
46
+ "model.layers.24.mlp",
47
+ "model.layers.41.mlp",
 
 
48
  "model.layers.45.mlp",
49
+ "model.layers.42.mlp",
50
  "model.layers.39.mlp",
51
+ "model.layers.40.mlp",
52
  "model.layers.34.mlp",
53
+ "model.layers.27.mlp",
54
  "model.layers.37.mlp",
55
+ "model.layers.25.mlp",
 
56
  "model.layers.30.mlp",
 
 
57
  "model.layers.33.mlp",
58
+ "model.layers.26.mlp",
59
+ "model.layers.38.mlp",
60
+ "model.layers.35.mlp",
61
  "model.layers.36.mlp",
62
+ "model.layers.31.mlp",
63
+ "model.layers.32.mlp",
64
  "model.layers.28.mlp",
65
  "model.layers.29.mlp",
66
  "model.layers.62.mlp",
 
79
  "sliding_window": null,
80
  "tie_word_embeddings": false,
81
  "torch_dtype": "bfloat16",
82
+ "transformers_version": "4.51.3",
83
  "unsloth_fixed": true,
84
  "use_cache": true,
85
  "use_sliding_window": false,
generation_config.json CHANGED
@@ -10,5 +10,5 @@
10
  "temperature": 0.6,
11
  "top_k": 20,
12
  "top_p": 0.95,
13
- "transformers_version": "4.52.0.dev0"
14
  }
 
10
  "temperature": 0.6,
11
  "top_k": 20,
12
  "top_p": 0.95,
13
+ "transformers_version": "4.51.3"
14
  }
model-00001-of-00009.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:06012b4891cfa4378e4a9d161565b4189e6b15557ff598732fcb82f11dcdf596
3
- size 4989649795
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a78e5657b38dd14e0b033c10096d558728f6778585f9da1e22ed73f36d8c315d
3
+ size 4965907285
model-00002-of-00009.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:54cfc03992d2245845e6ec804d0abdc0371192e7820df608a9a0659fb44dfa3c
3
- size 4989351257
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d25c9ef160542ca2ec31168bd3bfa3afd651e548d53b3528e60e51cfdaa08563
3
+ size 4873041084
model.safetensors.index.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "metadata": {
3
- "total_size": 39450141918
4
  },
5
  "weight_map": {
6
  "lm_head.weight": "model-00009-of-00009.safetensors",
@@ -104,12 +104,12 @@
104
  "model.layers.10.mlp.down_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
105
  "model.layers.10.mlp.down_proj.weight.quant_map": "model-00002-of-00009.safetensors",
106
  "model.layers.10.mlp.down_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
107
- "model.layers.10.mlp.gate_proj.weight": "model-00002-of-00009.safetensors",
108
- "model.layers.10.mlp.gate_proj.weight.absmax": "model-00002-of-00009.safetensors",
109
- "model.layers.10.mlp.gate_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
110
- "model.layers.10.mlp.gate_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
111
- "model.layers.10.mlp.gate_proj.weight.quant_map": "model-00002-of-00009.safetensors",
112
- "model.layers.10.mlp.gate_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
113
  "model.layers.10.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
114
  "model.layers.10.mlp.up_proj.weight.absmax": "model-00002-of-00009.safetensors",
115
  "model.layers.10.mlp.up_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
@@ -117,32 +117,32 @@
117
  "model.layers.10.mlp.up_proj.weight.quant_map": "model-00002-of-00009.safetensors",
118
  "model.layers.10.mlp.up_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
119
  "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
120
- "model.layers.10.self_attn.k_norm.weight": "model-00002-of-00009.safetensors",
121
- "model.layers.10.self_attn.k_proj.weight": "model-00002-of-00009.safetensors",
122
- "model.layers.10.self_attn.k_proj.weight.absmax": "model-00002-of-00009.safetensors",
123
- "model.layers.10.self_attn.k_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
124
- "model.layers.10.self_attn.k_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
125
- "model.layers.10.self_attn.k_proj.weight.quant_map": "model-00002-of-00009.safetensors",
126
- "model.layers.10.self_attn.k_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
127
- "model.layers.10.self_attn.o_proj.weight": "model-00002-of-00009.safetensors",
128
- "model.layers.10.self_attn.o_proj.weight.absmax": "model-00002-of-00009.safetensors",
129
- "model.layers.10.self_attn.o_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
130
- "model.layers.10.self_attn.o_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
131
- "model.layers.10.self_attn.o_proj.weight.quant_map": "model-00002-of-00009.safetensors",
132
- "model.layers.10.self_attn.o_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
133
- "model.layers.10.self_attn.q_norm.weight": "model-00002-of-00009.safetensors",
134
- "model.layers.10.self_attn.q_proj.weight": "model-00002-of-00009.safetensors",
135
- "model.layers.10.self_attn.q_proj.weight.absmax": "model-00002-of-00009.safetensors",
136
- "model.layers.10.self_attn.q_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
137
- "model.layers.10.self_attn.q_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
138
- "model.layers.10.self_attn.q_proj.weight.quant_map": "model-00002-of-00009.safetensors",
139
- "model.layers.10.self_attn.q_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
140
- "model.layers.10.self_attn.v_proj.weight": "model-00002-of-00009.safetensors",
141
- "model.layers.10.self_attn.v_proj.weight.absmax": "model-00002-of-00009.safetensors",
142
- "model.layers.10.self_attn.v_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
143
- "model.layers.10.self_attn.v_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
144
- "model.layers.10.self_attn.v_proj.weight.quant_map": "model-00002-of-00009.safetensors",
145
- "model.layers.10.self_attn.v_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
146
  "model.layers.11.input_layernorm.weight": "model-00002-of-00009.safetensors",
147
  "model.layers.11.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
148
  "model.layers.11.mlp.down_proj.weight.absmax": "model-00002-of-00009.safetensors",
@@ -2103,10 +2103,30 @@
2103
  "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
2104
  "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00009.safetensors",
2105
  "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
 
 
 
 
 
2106
  "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
 
 
 
 
 
2107
  "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00009.safetensors",
2108
  "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
 
 
 
 
 
2109
  "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
 
 
 
 
 
2110
  "model.layers.60.input_layernorm.weight": "model-00008-of-00009.safetensors",
2111
  "model.layers.60.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
2112
  "model.layers.60.mlp.down_proj.weight.absmax": "model-00008-of-00009.safetensors",
 
1
  {
2
  "metadata": {
3
+ "total_size": 39310086795
4
  },
5
  "weight_map": {
6
  "lm_head.weight": "model-00009-of-00009.safetensors",
 
104
  "model.layers.10.mlp.down_proj.weight.nested_quant_map": "model-00002-of-00009.safetensors",
105
  "model.layers.10.mlp.down_proj.weight.quant_map": "model-00002-of-00009.safetensors",
106
  "model.layers.10.mlp.down_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
107
+ "model.layers.10.mlp.gate_proj.weight": "model-00001-of-00009.safetensors",
108
+ "model.layers.10.mlp.gate_proj.weight.absmax": "model-00001-of-00009.safetensors",
109
+ "model.layers.10.mlp.gate_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
110
+ "model.layers.10.mlp.gate_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
111
+ "model.layers.10.mlp.gate_proj.weight.quant_map": "model-00001-of-00009.safetensors",
112
+ "model.layers.10.mlp.gate_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
113
  "model.layers.10.mlp.up_proj.weight": "model-00002-of-00009.safetensors",
114
  "model.layers.10.mlp.up_proj.weight.absmax": "model-00002-of-00009.safetensors",
115
  "model.layers.10.mlp.up_proj.weight.nested_absmax": "model-00002-of-00009.safetensors",
 
117
  "model.layers.10.mlp.up_proj.weight.quant_map": "model-00002-of-00009.safetensors",
118
  "model.layers.10.mlp.up_proj.weight.quant_state.bitsandbytes__nf4": "model-00002-of-00009.safetensors",
119
  "model.layers.10.post_attention_layernorm.weight": "model-00002-of-00009.safetensors",
120
+ "model.layers.10.self_attn.k_norm.weight": "model-00001-of-00009.safetensors",
121
+ "model.layers.10.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
122
+ "model.layers.10.self_attn.k_proj.weight.absmax": "model-00001-of-00009.safetensors",
123
+ "model.layers.10.self_attn.k_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
124
+ "model.layers.10.self_attn.k_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
125
+ "model.layers.10.self_attn.k_proj.weight.quant_map": "model-00001-of-00009.safetensors",
126
+ "model.layers.10.self_attn.k_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
127
+ "model.layers.10.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
128
+ "model.layers.10.self_attn.o_proj.weight.absmax": "model-00001-of-00009.safetensors",
129
+ "model.layers.10.self_attn.o_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
130
+ "model.layers.10.self_attn.o_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
131
+ "model.layers.10.self_attn.o_proj.weight.quant_map": "model-00001-of-00009.safetensors",
132
+ "model.layers.10.self_attn.o_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
133
+ "model.layers.10.self_attn.q_norm.weight": "model-00001-of-00009.safetensors",
134
+ "model.layers.10.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
135
+ "model.layers.10.self_attn.q_proj.weight.absmax": "model-00001-of-00009.safetensors",
136
+ "model.layers.10.self_attn.q_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
137
+ "model.layers.10.self_attn.q_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
138
+ "model.layers.10.self_attn.q_proj.weight.quant_map": "model-00001-of-00009.safetensors",
139
+ "model.layers.10.self_attn.q_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
140
+ "model.layers.10.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
141
+ "model.layers.10.self_attn.v_proj.weight.absmax": "model-00001-of-00009.safetensors",
142
+ "model.layers.10.self_attn.v_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
143
+ "model.layers.10.self_attn.v_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
144
+ "model.layers.10.self_attn.v_proj.weight.quant_map": "model-00001-of-00009.safetensors",
145
+ "model.layers.10.self_attn.v_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
146
  "model.layers.11.input_layernorm.weight": "model-00002-of-00009.safetensors",
147
  "model.layers.11.mlp.down_proj.weight": "model-00002-of-00009.safetensors",
148
  "model.layers.11.mlp.down_proj.weight.absmax": "model-00002-of-00009.safetensors",
 
2103
  "model.layers.6.post_attention_layernorm.weight": "model-00001-of-00009.safetensors",
2104
  "model.layers.6.self_attn.k_norm.weight": "model-00001-of-00009.safetensors",
2105
  "model.layers.6.self_attn.k_proj.weight": "model-00001-of-00009.safetensors",
2106
+ "model.layers.6.self_attn.k_proj.weight.absmax": "model-00001-of-00009.safetensors",
2107
+ "model.layers.6.self_attn.k_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
2108
+ "model.layers.6.self_attn.k_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
2109
+ "model.layers.6.self_attn.k_proj.weight.quant_map": "model-00001-of-00009.safetensors",
2110
+ "model.layers.6.self_attn.k_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
2111
  "model.layers.6.self_attn.o_proj.weight": "model-00001-of-00009.safetensors",
2112
+ "model.layers.6.self_attn.o_proj.weight.absmax": "model-00001-of-00009.safetensors",
2113
+ "model.layers.6.self_attn.o_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
2114
+ "model.layers.6.self_attn.o_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
2115
+ "model.layers.6.self_attn.o_proj.weight.quant_map": "model-00001-of-00009.safetensors",
2116
+ "model.layers.6.self_attn.o_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
2117
  "model.layers.6.self_attn.q_norm.weight": "model-00001-of-00009.safetensors",
2118
  "model.layers.6.self_attn.q_proj.weight": "model-00001-of-00009.safetensors",
2119
+ "model.layers.6.self_attn.q_proj.weight.absmax": "model-00001-of-00009.safetensors",
2120
+ "model.layers.6.self_attn.q_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
2121
+ "model.layers.6.self_attn.q_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
2122
+ "model.layers.6.self_attn.q_proj.weight.quant_map": "model-00001-of-00009.safetensors",
2123
+ "model.layers.6.self_attn.q_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
2124
  "model.layers.6.self_attn.v_proj.weight": "model-00001-of-00009.safetensors",
2125
+ "model.layers.6.self_attn.v_proj.weight.absmax": "model-00001-of-00009.safetensors",
2126
+ "model.layers.6.self_attn.v_proj.weight.nested_absmax": "model-00001-of-00009.safetensors",
2127
+ "model.layers.6.self_attn.v_proj.weight.nested_quant_map": "model-00001-of-00009.safetensors",
2128
+ "model.layers.6.self_attn.v_proj.weight.quant_map": "model-00001-of-00009.safetensors",
2129
+ "model.layers.6.self_attn.v_proj.weight.quant_state.bitsandbytes__nf4": "model-00001-of-00009.safetensors",
2130
  "model.layers.60.input_layernorm.weight": "model-00008-of-00009.safetensors",
2131
  "model.layers.60.mlp.down_proj.weight": "model-00008-of-00009.safetensors",
2132
  "model.layers.60.mlp.down_proj.weight.absmax": "model-00008-of-00009.safetensors",
tokenizer_config.json CHANGED
@@ -227,6 +227,7 @@
227
  "<|video_pad|>"
228
  ],
229
  "bos_token": null,
 
230
  "clean_up_tokenization_spaces": false,
231
  "eos_token": "<|im_end|>",
232
  "errors": "replace",
@@ -236,6 +237,5 @@
236
  "padding_side": "left",
237
  "split_special_tokens": false,
238
  "tokenizer_class": "Qwen2Tokenizer",
239
- "unk_token": null,
240
- "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 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 %}"
241
  }
 
227
  "<|video_pad|>"
228
  ],
229
  "bos_token": null,
230
+ "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 defined and 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 m_content = message.content if message.content is defined and message.content is not none else '' %}\n {%- set content = m_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 m_content %}\n {%- set content = (m_content.split('</think>')|last).lstrip('\\n') %}\n {%- set reasoning_content = (m_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 (not reasoning_content.strip() == '')) %}\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 %}",
231
  "clean_up_tokenization_spaces": false,
232
  "eos_token": "<|im_end|>",
233
  "errors": "replace",
 
237
  "padding_side": "left",
238
  "split_special_tokens": false,
239
  "tokenizer_class": "Qwen2Tokenizer",
240
+ "unk_token": null
 
241
  }