Hugging Face's logo

nm-testing
/
Qwen2-VL-72B-Instruct-quantized.w4a16

Safetensors
qwen2_vl
compressed-tensors
Model card Files Community
1
Shubhra Pandit commited on
Commit
7a76119
·
1 Parent(s): 6553822

Upload model files

Browse files
Files changed (22) hide show
  1. .gitattributes +1 -0
  2. added_tokens.json +16 -0
  3. chat_template.json +3 -0
  4. config.json +910 -0
  5. generation_config.json +13 -0
  6. merges.txt +0 -0
  7. model-00001-of-00009.safetensors +3 -0
  8. model-00002-of-00009.safetensors +3 -0
  9. model-00003-of-00009.safetensors +3 -0
  10. model-00004-of-00009.safetensors +3 -0
  11. model-00005-of-00009.safetensors +3 -0
  12. model-00006-of-00009.safetensors +3 -0
  13. model-00007-of-00009.safetensors +3 -0
  14. model-00008-of-00009.safetensors +3 -0
  15. model-00009-of-00009.safetensors +3 -0
  16. model.safetensors.index.json +0 -0
  17. preprocessor_config.json +29 -0
  18. recipe.yaml +7 -0
  19. special_tokens_map.json +31 -0
  20. tokenizer.json +3 -0
  21. tokenizer_config.json +145 -0
  22. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "<|box_end|>": 151649,
3
+ "<|box_start|>": 151648,
4
+ "<|endoftext|>": 151643,
5
+ "<|im_end|>": 151645,
6
+ "<|im_start|>": 151644,
7
+ "<|image_pad|>": 151655,
8
+ "<|object_ref_end|>": 151647,
9
+ "<|object_ref_start|>": 151646,
10
+ "<|quad_end|>": 151651,
11
+ "<|quad_start|>": 151650,
12
+ "<|video_pad|>": 151656,
13
+ "<|vision_end|>": 151653,
14
+ "<|vision_pad|>": 151654,
15
+ "<|vision_start|>": 151652
16
+ }
chat_template.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}"
3
+ }
config.json ADDED
@@ -0,0 +1,910 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Qwen/Qwen2-VL-72B-Instruct",
3
+ "architectures": [
4
+ "Qwen2VLForConditionalGeneration"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 8192,
11
+ "image_token_id": 151655,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 29568,
14
+ "max_position_embeddings": 32768,
15
+ "max_window_layers": 80,
16
+ "model_type": "qwen2_vl",
17
+ "num_attention_heads": 64,
18
+ "num_hidden_layers": 80,
19
+ "num_key_value_heads": 8,
20
+ "quantization_config": {
21
+ "config_groups": {
22
+ "group_0": {
23
+ "input_activations": null,
24
+ "output_activations": null,
25
+ "targets": [
26
+ "Linear"
27
+ ],
28
+ "weights": {
29
+ "actorder": null,
30
+ "block_structure": null,
31
+ "dynamic": false,
32
+ "group_size": 128,
33
+ "num_bits": 4,
34
+ "observer": "minmax",
35
+ "observer_kwargs": {},
36
+ "strategy": "group",
37
+ "symmetric": true,
38
+ "type": "int"
39
+ }
40
+ }
41
+ },
42
+ "format": "pack-quantized",
43
+ "global_compression_ratio": 1.6039917466222644,
44
+ "ignore": [
45
+ "visual.blocks.0.attn.qkv",
46
+ "visual.blocks.0.attn.proj",
47
+ "visual.blocks.0.mlp.fc1",
48
+ "visual.blocks.0.mlp.fc2",
49
+ "visual.blocks.1.attn.qkv",
50
+ "visual.blocks.1.attn.proj",
51
+ "visual.blocks.1.mlp.fc1",
52
+ "visual.blocks.1.mlp.fc2",
53
+ "visual.blocks.2.attn.qkv",
54
+ "visual.blocks.2.attn.proj",
55
+ "visual.blocks.2.mlp.fc1",
56
+ "visual.blocks.2.mlp.fc2",
57
+ "visual.blocks.3.attn.qkv",
58
+ "visual.blocks.3.attn.proj",
59
+ "visual.blocks.3.mlp.fc1",
60
+ "visual.blocks.3.mlp.fc2",
61
+ "visual.blocks.4.attn.qkv",
62
+ "visual.blocks.4.attn.proj",
63
+ "visual.blocks.4.mlp.fc1",
64
+ "visual.blocks.4.mlp.fc2",
65
+ "visual.blocks.5.attn.qkv",
66
+ "visual.blocks.5.attn.proj",
67
+ "visual.blocks.5.mlp.fc1",
68
+ "visual.blocks.5.mlp.fc2",
69
+ "visual.blocks.6.attn.qkv",
70
+ "visual.blocks.6.attn.proj",
71
+ "visual.blocks.6.mlp.fc1",
72
+ "visual.blocks.6.mlp.fc2",
73
+ "visual.blocks.7.attn.qkv",
74
+ "visual.blocks.7.attn.proj",
75
+ "visual.blocks.7.mlp.fc1",
76
+ "visual.blocks.7.mlp.fc2",
77
+ "visual.blocks.8.attn.qkv",
78
+ "visual.blocks.8.attn.proj",
79
+ "visual.blocks.8.mlp.fc1",
80
+ "visual.blocks.8.mlp.fc2",
81
+ "visual.blocks.9.attn.qkv",
82
+ "visual.blocks.9.attn.proj",
83
+ "visual.blocks.9.mlp.fc1",
84
+ "visual.blocks.9.mlp.fc2",
85
+ "visual.blocks.10.attn.qkv",
86
+ "visual.blocks.10.attn.proj",
87
+ "visual.blocks.10.mlp.fc1",
88
+ "visual.blocks.10.mlp.fc2",
89
+ "visual.blocks.11.attn.qkv",
90
+ "visual.blocks.11.attn.proj",
91
+ "visual.blocks.11.mlp.fc1",
92
+ "visual.blocks.11.mlp.fc2",
93
+ "visual.blocks.12.attn.qkv",
94
+ "visual.blocks.12.attn.proj",
95
+ "visual.blocks.12.mlp.fc1",
96
+ "visual.blocks.12.mlp.fc2",
97
+ "visual.blocks.13.attn.qkv",
98
+ "visual.blocks.13.attn.proj",
99
+ "visual.blocks.13.mlp.fc1",
100
+ "visual.blocks.13.mlp.fc2",
101
+ "visual.blocks.14.attn.qkv",
102
+ "visual.blocks.14.attn.proj",
103
+ "visual.blocks.14.mlp.fc1",
104
+ "visual.blocks.14.mlp.fc2",
105
+ "visual.blocks.15.attn.qkv",
106
+ "visual.blocks.15.attn.proj",
107
+ "visual.blocks.15.mlp.fc1",
108
+ "visual.blocks.15.mlp.fc2",
109
+ "visual.blocks.16.attn.qkv",
110
+ "visual.blocks.16.attn.proj",
111
+ "visual.blocks.16.mlp.fc1",
112
+ "visual.blocks.16.mlp.fc2",
113
+ "visual.blocks.17.attn.qkv",
114
+ "visual.blocks.17.attn.proj",
115
+ "visual.blocks.17.mlp.fc1",
116
+ "visual.blocks.17.mlp.fc2",
117
+ "visual.blocks.18.attn.qkv",
118
+ "visual.blocks.18.attn.proj",
119
+ "visual.blocks.18.mlp.fc1",
120
+ "visual.blocks.18.mlp.fc2",
121
+ "visual.blocks.19.attn.qkv",
122
+ "visual.blocks.19.attn.proj",
123
+ "visual.blocks.19.mlp.fc1",
124
+ "visual.blocks.19.mlp.fc2",
125
+ "visual.blocks.20.attn.qkv",
126
+ "visual.blocks.20.attn.proj",
127
+ "visual.blocks.20.mlp.fc1",
128
+ "visual.blocks.20.mlp.fc2",
129
+ "visual.blocks.21.attn.qkv",
130
+ "visual.blocks.21.attn.proj",
131
+ "visual.blocks.21.mlp.fc1",
132
+ "visual.blocks.21.mlp.fc2",
133
+ "visual.blocks.22.attn.qkv",
134
+ "visual.blocks.22.attn.proj",
135
+ "visual.blocks.22.mlp.fc1",
136
+ "visual.blocks.22.mlp.fc2",
137
+ "visual.blocks.23.attn.qkv",
138
+ "visual.blocks.23.attn.proj",
139
+ "visual.blocks.23.mlp.fc1",
140
+ "visual.blocks.23.mlp.fc2",
141
+ "visual.blocks.24.attn.qkv",
142
+ "visual.blocks.24.attn.proj",
143
+ "visual.blocks.24.mlp.fc1",
144
+ "visual.blocks.24.mlp.fc2",
145
+ "visual.blocks.25.attn.qkv",
146
+ "visual.blocks.25.attn.proj",
147
+ "visual.blocks.25.mlp.fc1",
148
+ "visual.blocks.25.mlp.fc2",
149
+ "visual.blocks.26.attn.qkv",
150
+ "visual.blocks.26.attn.proj",
151
+ "visual.blocks.26.mlp.fc1",
152
+ "visual.blocks.26.mlp.fc2",
153
+ "visual.blocks.27.attn.qkv",
154
+ "visual.blocks.27.attn.proj",
155
+ "visual.blocks.27.mlp.fc1",
156
+ "visual.blocks.27.mlp.fc2",
157
+ "visual.blocks.28.attn.qkv",
158
+ "visual.blocks.28.attn.proj",
159
+ "visual.blocks.28.mlp.fc1",
160
+ "visual.blocks.28.mlp.fc2",
161
+ "visual.blocks.29.attn.qkv",
162
+ "visual.blocks.29.attn.proj",
163
+ "visual.blocks.29.mlp.fc1",
164
+ "visual.blocks.29.mlp.fc2",
165
+ "visual.blocks.30.attn.qkv",
166
+ "visual.blocks.30.attn.proj",
167
+ "visual.blocks.30.mlp.fc1",
168
+ "visual.blocks.30.mlp.fc2",
169
+ "visual.blocks.31.attn.qkv",
170
+ "visual.blocks.31.attn.proj",
171
+ "visual.blocks.31.mlp.fc1",
172
+ "visual.blocks.31.mlp.fc2",
173
+ "visual.merger.mlp.0",
174
+ "visual.merger.mlp.2",
175
+ "lm_head"
176
+ ],
177
+ "kv_cache_scheme": null,
178
+ "quant_method": "compressed-tensors",
179
+ "quantization_status": "compressed",
180
+ "sparsity_config": {
181
+ "format": "dense",
182
+ "global_sparsity": 0.15412567029899502,
183
+ "ignore": [
184
+ "visual.blocks.0.attn.qkv",
185
+ "visual.blocks.0.attn.proj",
186
+ "visual.blocks.0.mlp.fc1",
187
+ "visual.blocks.0.mlp.fc2",
188
+ "visual.blocks.1.attn.qkv",
189
+ "visual.blocks.1.attn.proj",
190
+ "visual.blocks.1.mlp.fc1",
191
+ "visual.blocks.1.mlp.fc2",
192
+ "visual.blocks.2.attn.qkv",
193
+ "visual.blocks.2.attn.proj",
194
+ "visual.blocks.2.mlp.fc1",
195
+ "visual.blocks.2.mlp.fc2",
196
+ "visual.blocks.3.attn.qkv",
197
+ "visual.blocks.3.attn.proj",
198
+ "visual.blocks.3.mlp.fc1",
199
+ "visual.blocks.3.mlp.fc2",
200
+ "visual.blocks.4.attn.qkv",
201
+ "visual.blocks.4.attn.proj",
202
+ "visual.blocks.4.mlp.fc1",
203
+ "visual.blocks.4.mlp.fc2",
204
+ "visual.blocks.5.attn.qkv",
205
+ "visual.blocks.5.attn.proj",
206
+ "visual.blocks.5.mlp.fc1",
207
+ "visual.blocks.5.mlp.fc2",
208
+ "visual.blocks.6.attn.qkv",
209
+ "visual.blocks.6.attn.proj",
210
+ "visual.blocks.6.mlp.fc1",
211
+ "visual.blocks.6.mlp.fc2",
212
+ "visual.blocks.7.attn.qkv",
213
+ "visual.blocks.7.attn.proj",
214
+ "visual.blocks.7.mlp.fc1",
215
+ "visual.blocks.7.mlp.fc2",
216
+ "visual.blocks.8.attn.qkv",
217
+ "visual.blocks.8.attn.proj",
218
+ "visual.blocks.8.mlp.fc1",
219
+ "visual.blocks.8.mlp.fc2",
220
+ "visual.blocks.9.attn.qkv",
221
+ "visual.blocks.9.attn.proj",
222
+ "visual.blocks.9.mlp.fc1",
223
+ "visual.blocks.9.mlp.fc2",
224
+ "visual.blocks.10.attn.qkv",
225
+ "visual.blocks.10.attn.proj",
226
+ "visual.blocks.10.mlp.fc1",
227
+ "visual.blocks.10.mlp.fc2",
228
+ "visual.blocks.11.attn.qkv",
229
+ "visual.blocks.11.attn.proj",
230
+ "visual.blocks.11.mlp.fc1",
231
+ "visual.blocks.11.mlp.fc2",
232
+ "visual.blocks.12.attn.qkv",
233
+ "visual.blocks.12.attn.proj",
234
+ "visual.blocks.12.mlp.fc1",
235
+ "visual.blocks.12.mlp.fc2",
236
+ "visual.blocks.13.attn.qkv",
237
+ "visual.blocks.13.attn.proj",
238
+ "visual.blocks.13.mlp.fc1",
239
+ "visual.blocks.13.mlp.fc2",
240
+ "visual.blocks.14.attn.qkv",
241
+ "visual.blocks.14.attn.proj",
242
+ "visual.blocks.14.mlp.fc1",
243
+ "visual.blocks.14.mlp.fc2",
244
+ "visual.blocks.15.attn.qkv",
245
+ "visual.blocks.15.attn.proj",
246
+ "visual.blocks.15.mlp.fc1",
247
+ "visual.blocks.15.mlp.fc2",
248
+ "visual.blocks.16.attn.qkv",
249
+ "visual.blocks.16.attn.proj",
250
+ "visual.blocks.16.mlp.fc1",
251
+ "visual.blocks.16.mlp.fc2",
252
+ "visual.blocks.17.attn.qkv",
253
+ "visual.blocks.17.attn.proj",
254
+ "visual.blocks.17.mlp.fc1",
255
+ "visual.blocks.17.mlp.fc2",
256
+ "visual.blocks.18.attn.qkv",
257
+ "visual.blocks.18.attn.proj",
258
+ "visual.blocks.18.mlp.fc1",
259
+ "visual.blocks.18.mlp.fc2",
260
+ "visual.blocks.19.attn.qkv",
261
+ "visual.blocks.19.attn.proj",
262
+ "visual.blocks.19.mlp.fc1",
263
+ "visual.blocks.19.mlp.fc2",
264
+ "visual.blocks.20.attn.qkv",
265
+ "visual.blocks.20.attn.proj",
266
+ "visual.blocks.20.mlp.fc1",
267
+ "visual.blocks.20.mlp.fc2",
268
+ "visual.blocks.21.attn.qkv",
269
+ "visual.blocks.21.attn.proj",
270
+ "visual.blocks.21.mlp.fc1",
271
+ "visual.blocks.21.mlp.fc2",
272
+ "visual.blocks.22.attn.qkv",
273
+ "visual.blocks.22.attn.proj",
274
+ "visual.blocks.22.mlp.fc1",
275
+ "visual.blocks.22.mlp.fc2",
276
+ "visual.blocks.23.attn.qkv",
277
+ "visual.blocks.23.attn.proj",
278
+ "visual.blocks.23.mlp.fc1",
279
+ "visual.blocks.23.mlp.fc2",
280
+ "visual.blocks.24.attn.qkv",
281
+ "visual.blocks.24.attn.proj",
282
+ "visual.blocks.24.mlp.fc1",
283
+ "visual.blocks.24.mlp.fc2",
284
+ "visual.blocks.25.attn.qkv",
285
+ "visual.blocks.25.attn.proj",
286
+ "visual.blocks.25.mlp.fc1",
287
+ "visual.blocks.25.mlp.fc2",
288
+ "visual.blocks.26.attn.qkv",
289
+ "visual.blocks.26.attn.proj",
290
+ "visual.blocks.26.mlp.fc1",
291
+ "visual.blocks.26.mlp.fc2",
292
+ "visual.blocks.27.attn.qkv",
293
+ "visual.blocks.27.attn.proj",
294
+ "visual.blocks.27.mlp.fc1",
295
+ "visual.blocks.27.mlp.fc2",
296
+ "visual.blocks.28.attn.qkv",
297
+ "visual.blocks.28.attn.proj",
298
+ "visual.blocks.28.mlp.fc1",
299
+ "visual.blocks.28.mlp.fc2",
300
+ "visual.blocks.29.attn.qkv",
301
+ "visual.blocks.29.attn.proj",
302
+ "visual.blocks.29.mlp.fc1",
303
+ "visual.blocks.29.mlp.fc2",
304
+ "visual.blocks.30.attn.qkv",
305
+ "visual.blocks.30.attn.proj",
306
+ "visual.blocks.30.mlp.fc1",
307
+ "visual.blocks.30.mlp.fc2",
308
+ "visual.blocks.31.attn.qkv",
309
+ "visual.blocks.31.attn.proj",
310
+ "visual.blocks.31.mlp.fc1",
311
+ "visual.blocks.31.mlp.fc2",
312
+ "visual.merger.mlp.0",
313
+ "visual.merger.mlp.2",
314
+ "model.layers.0.self_attn.q_proj",
315
+ "model.layers.0.self_attn.k_proj",
316
+ "model.layers.0.self_attn.v_proj",
317
+ "model.layers.0.self_attn.o_proj",
318
+ "model.layers.0.mlp.gate_proj",
319
+ "model.layers.0.mlp.up_proj",
320
+ "model.layers.0.mlp.down_proj",
321
+ "model.layers.1.self_attn.v_proj",
322
+ "model.layers.1.self_attn.o_proj",
323
+ "model.layers.1.mlp.gate_proj",
324
+ "model.layers.1.mlp.up_proj",
325
+ "model.layers.1.mlp.down_proj",
326
+ "model.layers.2.self_attn.v_proj",
327
+ "model.layers.2.self_attn.o_proj",
328
+ "model.layers.2.mlp.gate_proj",
329
+ "model.layers.2.mlp.up_proj",
330
+ "model.layers.2.mlp.down_proj",
331
+ "model.layers.3.self_attn.k_proj",
332
+ "model.layers.3.self_attn.v_proj",
333
+ "model.layers.3.self_attn.o_proj",
334
+ "model.layers.3.mlp.gate_proj",
335
+ "model.layers.3.mlp.up_proj",
336
+ "model.layers.3.mlp.down_proj",
337
+ "model.layers.4.self_attn.q_proj",
338
+ "model.layers.4.self_attn.k_proj",
339
+ "model.layers.4.self_attn.v_proj",
340
+ "model.layers.4.self_attn.o_proj",
341
+ "model.layers.4.mlp.gate_proj",
342
+ "model.layers.4.mlp.up_proj",
343
+ "model.layers.4.mlp.down_proj",
344
+ "model.layers.5.self_attn.k_proj",
345
+ "model.layers.5.self_attn.v_proj",
346
+ "model.layers.5.self_attn.o_proj",
347
+ "model.layers.5.mlp.gate_proj",
348
+ "model.layers.5.mlp.up_proj",
349
+ "model.layers.5.mlp.down_proj",
350
+ "model.layers.6.self_attn.q_proj",
351
+ "model.layers.6.self_attn.k_proj",
352
+ "model.layers.6.self_attn.v_proj",
353
+ "model.layers.6.self_attn.o_proj",
354
+ "model.layers.6.mlp.gate_proj",
355
+ "model.layers.6.mlp.up_proj",
356
+ "model.layers.6.mlp.down_proj",
357
+ "model.layers.7.self_attn.q_proj",
358
+ "model.layers.7.self_attn.k_proj",
359
+ "model.layers.7.self_attn.v_proj",
360
+ "model.layers.7.self_attn.o_proj",
361
+ "model.layers.7.mlp.gate_proj",
362
+ "model.layers.7.mlp.up_proj",
363
+ "model.layers.7.mlp.down_proj",
364
+ "model.layers.8.self_attn.q_proj",
365
+ "model.layers.8.self_attn.k_proj",
366
+ "model.layers.8.self_attn.v_proj",
367
+ "model.layers.8.self_attn.o_proj",
368
+ "model.layers.8.mlp.gate_proj",
369
+ "model.layers.8.mlp.up_proj",
370
+ "model.layers.8.mlp.down_proj",
371
+ "model.layers.9.self_attn.q_proj",
372
+ "model.layers.9.self_attn.k_proj",
373
+ "model.layers.9.self_attn.v_proj",
374
+ "model.layers.9.self_attn.o_proj",
375
+ "model.layers.9.mlp.gate_proj",
376
+ "model.layers.9.mlp.up_proj",
377
+ "model.layers.9.mlp.down_proj",
378
+ "model.layers.10.self_attn.q_proj",
379
+ "model.layers.10.self_attn.k_proj",
380
+ "model.layers.10.self_attn.v_proj",
381
+ "model.layers.10.self_attn.o_proj",
382
+ "model.layers.10.mlp.gate_proj",
383
+ "model.layers.10.mlp.up_proj",
384
+ "model.layers.10.mlp.down_proj",
385
+ "model.layers.11.self_attn.q_proj",
386
+ "model.layers.11.self_attn.k_proj",
387
+ "model.layers.11.self_attn.v_proj",
388
+ "model.layers.11.self_attn.o_proj",
389
+ "model.layers.11.mlp.gate_proj",
390
+ "model.layers.11.mlp.up_proj",
391
+ "model.layers.11.mlp.down_proj",
392
+ "model.layers.12.self_attn.q_proj",
393
+ "model.layers.12.self_attn.k_proj",
394
+ "model.layers.12.self_attn.v_proj",
395
+ "model.layers.12.self_attn.o_proj",
396
+ "model.layers.12.mlp.gate_proj",
397
+ "model.layers.12.mlp.up_proj",
398
+ "model.layers.12.mlp.down_proj",
399
+ "model.layers.13.self_attn.q_proj",
400
+ "model.layers.13.self_attn.k_proj",
401
+ "model.layers.13.self_attn.v_proj",
402
+ "model.layers.13.self_attn.o_proj",
403
+ "model.layers.13.mlp.gate_proj",
404
+ "model.layers.13.mlp.up_proj",
405
+ "model.layers.13.mlp.down_proj",
406
+ "model.layers.14.self_attn.q_proj",
407
+ "model.layers.14.self_attn.k_proj",
408
+ "model.layers.14.self_attn.v_proj",
409
+ "model.layers.14.self_attn.o_proj",
410
+ "model.layers.14.mlp.gate_proj",
411
+ "model.layers.14.mlp.up_proj",
412
+ "model.layers.14.mlp.down_proj",
413
+ "model.layers.15.self_attn.q_proj",
414
+ "model.layers.15.self_attn.k_proj",
415
+ "model.layers.15.self_attn.v_proj",
416
+ "model.layers.15.self_attn.o_proj",
417
+ "model.layers.15.mlp.gate_proj",
418
+ "model.layers.15.mlp.up_proj",
419
+ "model.layers.15.mlp.down_proj",
420
+ "model.layers.16.self_attn.q_proj",
421
+ "model.layers.16.self_attn.k_proj",
422
+ "model.layers.16.self_attn.v_proj",
423
+ "model.layers.16.self_attn.o_proj",
424
+ "model.layers.16.mlp.gate_proj",
425
+ "model.layers.16.mlp.up_proj",
426
+ "model.layers.16.mlp.down_proj",
427
+ "model.layers.17.self_attn.q_proj",
428
+ "model.layers.17.self_attn.k_proj",
429
+ "model.layers.17.self_attn.v_proj",
430
+ "model.layers.17.self_attn.o_proj",
431
+ "model.layers.17.mlp.gate_proj",
432
+ "model.layers.17.mlp.up_proj",
433
+ "model.layers.17.mlp.down_proj",
434
+ "model.layers.18.self_attn.q_proj",
435
+ "model.layers.18.self_attn.k_proj",
436
+ "model.layers.18.self_attn.v_proj",
437
+ "model.layers.18.self_attn.o_proj",
438
+ "model.layers.18.mlp.gate_proj",
439
+ "model.layers.18.mlp.up_proj",
440
+ "model.layers.18.mlp.down_proj",
441
+ "model.layers.19.self_attn.q_proj",
442
+ "model.layers.19.self_attn.k_proj",
443
+ "model.layers.19.self_attn.v_proj",
444
+ "model.layers.19.self_attn.o_proj",
445
+ "model.layers.19.mlp.gate_proj",
446
+ "model.layers.19.mlp.up_proj",
447
+ "model.layers.19.mlp.down_proj",
448
+ "model.layers.20.self_attn.q_proj",
449
+ "model.layers.20.self_attn.k_proj",
450
+ "model.layers.20.self_attn.v_proj",
451
+ "model.layers.20.self_attn.o_proj",
452
+ "model.layers.20.mlp.gate_proj",
453
+ "model.layers.20.mlp.up_proj",
454
+ "model.layers.20.mlp.down_proj",
455
+ "model.layers.21.self_attn.q_proj",
456
+ "model.layers.21.self_attn.k_proj",
457
+ "model.layers.21.self_attn.v_proj",
458
+ "model.layers.21.self_attn.o_proj",
459
+ "model.layers.21.mlp.gate_proj",
460
+ "model.layers.21.mlp.up_proj",
461
+ "model.layers.21.mlp.down_proj",
462
+ "model.layers.22.self_attn.q_proj",
463
+ "model.layers.22.self_attn.k_proj",
464
+ "model.layers.22.self_attn.v_proj",
465
+ "model.layers.22.self_attn.o_proj",
466
+ "model.layers.22.mlp.gate_proj",
467
+ "model.layers.22.mlp.up_proj",
468
+ "model.layers.22.mlp.down_proj",
469
+ "model.layers.23.self_attn.q_proj",
470
+ "model.layers.23.self_attn.k_proj",
471
+ "model.layers.23.self_attn.v_proj",
472
+ "model.layers.23.self_attn.o_proj",
473
+ "model.layers.23.mlp.gate_proj",
474
+ "model.layers.23.mlp.up_proj",
475
+ "model.layers.23.mlp.down_proj",
476
+ "model.layers.24.self_attn.q_proj",
477
+ "model.layers.24.self_attn.k_proj",
478
+ "model.layers.24.self_attn.v_proj",
479
+ "model.layers.24.self_attn.o_proj",
480
+ "model.layers.24.mlp.gate_proj",
481
+ "model.layers.24.mlp.up_proj",
482
+ "model.layers.24.mlp.down_proj",
483
+ "model.layers.25.self_attn.q_proj",
484
+ "model.layers.25.self_attn.k_proj",
485
+ "model.layers.25.self_attn.v_proj",
486
+ "model.layers.25.self_attn.o_proj",
487
+ "model.layers.25.mlp.gate_proj",
488
+ "model.layers.25.mlp.up_proj",
489
+ "model.layers.25.mlp.down_proj",
490
+ "model.layers.26.self_attn.q_proj",
491
+ "model.layers.26.self_attn.k_proj",
492
+ "model.layers.26.self_attn.v_proj",
493
+ "model.layers.26.self_attn.o_proj",
494
+ "model.layers.26.mlp.gate_proj",
495
+ "model.layers.26.mlp.up_proj",
496
+ "model.layers.26.mlp.down_proj",
497
+ "model.layers.27.self_attn.q_proj",
498
+ "model.layers.27.self_attn.k_proj",
499
+ "model.layers.27.self_attn.v_proj",
500
+ "model.layers.27.self_attn.o_proj",
501
+ "model.layers.27.mlp.gate_proj",
502
+ "model.layers.27.mlp.up_proj",
503
+ "model.layers.27.mlp.down_proj",
504
+ "model.layers.28.self_attn.q_proj",
505
+ "model.layers.28.self_attn.k_proj",
506
+ "model.layers.28.self_attn.v_proj",
507
+ "model.layers.28.self_attn.o_proj",
508
+ "model.layers.28.mlp.gate_proj",
509
+ "model.layers.28.mlp.up_proj",
510
+ "model.layers.28.mlp.down_proj",
511
+ "model.layers.29.self_attn.q_proj",
512
+ "model.layers.29.self_attn.k_proj",
513
+ "model.layers.29.self_attn.v_proj",
514
+ "model.layers.29.self_attn.o_proj",
515
+ "model.layers.29.mlp.gate_proj",
516
+ "model.layers.29.mlp.up_proj",
517
+ "model.layers.29.mlp.down_proj",
518
+ "model.layers.30.self_attn.q_proj",
519
+ "model.layers.30.self_attn.k_proj",
520
+ "model.layers.30.self_attn.v_proj",
521
+ "model.layers.30.self_attn.o_proj",
522
+ "model.layers.30.mlp.gate_proj",
523
+ "model.layers.30.mlp.up_proj",
524
+ "model.layers.30.mlp.down_proj",
525
+ "model.layers.31.self_attn.q_proj",
526
+ "model.layers.31.self_attn.k_proj",
527
+ "model.layers.31.self_attn.v_proj",
528
+ "model.layers.31.self_attn.o_proj",
529
+ "model.layers.31.mlp.gate_proj",
530
+ "model.layers.31.mlp.up_proj",
531
+ "model.layers.31.mlp.down_proj",
532
+ "model.layers.32.self_attn.q_proj",
533
+ "model.layers.32.self_attn.k_proj",
534
+ "model.layers.32.self_attn.v_proj",
535
+ "model.layers.32.self_attn.o_proj",
536
+ "model.layers.32.mlp.gate_proj",
537
+ "model.layers.32.mlp.up_proj",
538
+ "model.layers.32.mlp.down_proj",
539
+ "model.layers.33.self_attn.q_proj",
540
+ "model.layers.33.self_attn.k_proj",
541
+ "model.layers.33.self_attn.v_proj",
542
+ "model.layers.33.self_attn.o_proj",
543
+ "model.layers.33.mlp.gate_proj",
544
+ "model.layers.33.mlp.up_proj",
545
+ "model.layers.33.mlp.down_proj",
546
+ "model.layers.34.self_attn.q_proj",
547
+ "model.layers.34.self_attn.k_proj",
548
+ "model.layers.34.self_attn.v_proj",
549
+ "model.layers.34.self_attn.o_proj",
550
+ "model.layers.34.mlp.gate_proj",
551
+ "model.layers.34.mlp.up_proj",
552
+ "model.layers.34.mlp.down_proj",
553
+ "model.layers.35.self_attn.q_proj",
554
+ "model.layers.35.self_attn.k_proj",
555
+ "model.layers.35.self_attn.v_proj",
556
+ "model.layers.35.self_attn.o_proj",
557
+ "model.layers.35.mlp.gate_proj",
558
+ "model.layers.35.mlp.up_proj",
559
+ "model.layers.35.mlp.down_proj",
560
+ "model.layers.36.self_attn.q_proj",
561
+ "model.layers.36.self_attn.k_proj",
562
+ "model.layers.36.self_attn.v_proj",
563
+ "model.layers.36.self_attn.o_proj",
564
+ "model.layers.36.mlp.gate_proj",
565
+ "model.layers.36.mlp.up_proj",
566
+ "model.layers.36.mlp.down_proj",
567
+ "model.layers.37.self_attn.q_proj",
568
+ "model.layers.37.self_attn.k_proj",
569
+ "model.layers.37.self_attn.v_proj",
570
+ "model.layers.37.self_attn.o_proj",
571
+ "model.layers.37.mlp.gate_proj",
572
+ "model.layers.37.mlp.up_proj",
573
+ "model.layers.37.mlp.down_proj",
574
+ "model.layers.38.self_attn.q_proj",
575
+ "model.layers.38.self_attn.k_proj",
576
+ "model.layers.38.self_attn.v_proj",
577
+ "model.layers.38.self_attn.o_proj",
578
+ "model.layers.38.mlp.gate_proj",
579
+ "model.layers.38.mlp.up_proj",
580
+ "model.layers.38.mlp.down_proj",
581
+ "model.layers.39.self_attn.q_proj",
582
+ "model.layers.39.self_attn.k_proj",
583
+ "model.layers.39.self_attn.v_proj",
584
+ "model.layers.39.self_attn.o_proj",
585
+ "model.layers.39.mlp.gate_proj",
586
+ "model.layers.39.mlp.up_proj",
587
+ "model.layers.39.mlp.down_proj",
588
+ "model.layers.40.self_attn.q_proj",
589
+ "model.layers.40.self_attn.k_proj",
590
+ "model.layers.40.self_attn.v_proj",
591
+ "model.layers.40.self_attn.o_proj",
592
+ "model.layers.40.mlp.gate_proj",
593
+ "model.layers.40.mlp.up_proj",
594
+ "model.layers.40.mlp.down_proj",
595
+ "model.layers.41.self_attn.q_proj",
596
+ "model.layers.41.self_attn.k_proj",
597
+ "model.layers.41.self_attn.v_proj",
598
+ "model.layers.41.self_attn.o_proj",
599
+ "model.layers.41.mlp.gate_proj",
600
+ "model.layers.41.mlp.up_proj",
601
+ "model.layers.41.mlp.down_proj",
602
+ "model.layers.42.self_attn.q_proj",
603
+ "model.layers.42.self_attn.k_proj",
604
+ "model.layers.42.self_attn.v_proj",
605
+ "model.layers.42.self_attn.o_proj",
606
+ "model.layers.42.mlp.gate_proj",
607
+ "model.layers.42.mlp.up_proj",
608
+ "model.layers.42.mlp.down_proj",
609
+ "model.layers.43.self_attn.q_proj",
610
+ "model.layers.43.self_attn.k_proj",
611
+ "model.layers.43.self_attn.v_proj",
612
+ "model.layers.43.self_attn.o_proj",
613
+ "model.layers.43.mlp.gate_proj",
614
+ "model.layers.43.mlp.up_proj",
615
+ "model.layers.43.mlp.down_proj",
616
+ "model.layers.44.self_attn.q_proj",
617
+ "model.layers.44.self_attn.k_proj",
618
+ "model.layers.44.self_attn.v_proj",
619
+ "model.layers.44.self_attn.o_proj",
620
+ "model.layers.44.mlp.gate_proj",
621
+ "model.layers.44.mlp.up_proj",
622
+ "model.layers.44.mlp.down_proj",
623
+ "model.layers.45.self_attn.q_proj",
624
+ "model.layers.45.self_attn.k_proj",
625
+ "model.layers.45.self_attn.v_proj",
626
+ "model.layers.45.self_attn.o_proj",
627
+ "model.layers.45.mlp.gate_proj",
628
+ "model.layers.45.mlp.up_proj",
629
+ "model.layers.45.mlp.down_proj",
630
+ "model.layers.46.self_attn.q_proj",
631
+ "model.layers.46.self_attn.k_proj",
632
+ "model.layers.46.self_attn.v_proj",
633
+ "model.layers.46.self_attn.o_proj",
634
+ "model.layers.46.mlp.gate_proj",
635
+ "model.layers.46.mlp.up_proj",
636
+ "model.layers.46.mlp.down_proj",
637
+ "model.layers.47.self_attn.q_proj",
638
+ "model.layers.47.self_attn.k_proj",
639
+ "model.layers.47.self_attn.v_proj",
640
+ "model.layers.47.self_attn.o_proj",
641
+ "model.layers.47.mlp.gate_proj",
642
+ "model.layers.47.mlp.up_proj",
643
+ "model.layers.47.mlp.down_proj",
644
+ "model.layers.48.self_attn.q_proj",
645
+ "model.layers.48.self_attn.k_proj",
646
+ "model.layers.48.self_attn.v_proj",
647
+ "model.layers.48.self_attn.o_proj",
648
+ "model.layers.48.mlp.gate_proj",
649
+ "model.layers.48.mlp.up_proj",
650
+ "model.layers.48.mlp.down_proj",
651
+ "model.layers.49.self_attn.q_proj",
652
+ "model.layers.49.self_attn.k_proj",
653
+ "model.layers.49.self_attn.v_proj",
654
+ "model.layers.49.self_attn.o_proj",
655
+ "model.layers.49.mlp.gate_proj",
656
+ "model.layers.49.mlp.up_proj",
657
+ "model.layers.49.mlp.down_proj",
658
+ "model.layers.50.self_attn.q_proj",
659
+ "model.layers.50.self_attn.k_proj",
660
+ "model.layers.50.self_attn.v_proj",
661
+ "model.layers.50.self_attn.o_proj",
662
+ "model.layers.50.mlp.gate_proj",
663
+ "model.layers.50.mlp.up_proj",
664
+ "model.layers.50.mlp.down_proj",
665
+ "model.layers.51.self_attn.q_proj",
666
+ "model.layers.51.self_attn.k_proj",
667
+ "model.layers.51.self_attn.v_proj",
668
+ "model.layers.51.self_attn.o_proj",
669
+ "model.layers.51.mlp.gate_proj",
670
+ "model.layers.51.mlp.up_proj",
671
+ "model.layers.51.mlp.down_proj",
672
+ "model.layers.52.self_attn.q_proj",
673
+ "model.layers.52.self_attn.k_proj",
674
+ "model.layers.52.self_attn.v_proj",
675
+ "model.layers.52.self_attn.o_proj",
676
+ "model.layers.52.mlp.gate_proj",
677
+ "model.layers.52.mlp.up_proj",
678
+ "model.layers.52.mlp.down_proj",
679
+ "model.layers.53.self_attn.q_proj",
680
+ "model.layers.53.self_attn.k_proj",
681
+ "model.layers.53.self_attn.v_proj",
682
+ "model.layers.53.self_attn.o_proj",
683
+ "model.layers.53.mlp.gate_proj",
684
+ "model.layers.53.mlp.up_proj",
685
+ "model.layers.53.mlp.down_proj",
686
+ "model.layers.54.self_attn.q_proj",
687
+ "model.layers.54.self_attn.k_proj",
688
+ "model.layers.54.self_attn.v_proj",
689
+ "model.layers.54.self_attn.o_proj",
690
+ "model.layers.54.mlp.gate_proj",
691
+ "model.layers.54.mlp.up_proj",
692
+ "model.layers.54.mlp.down_proj",
693
+ "model.layers.55.self_attn.q_proj",
694
+ "model.layers.55.self_attn.k_proj",
695
+ "model.layers.55.self_attn.v_proj",
696
+ "model.layers.55.self_attn.o_proj",
697
+ "model.layers.55.mlp.gate_proj",
698
+ "model.layers.55.mlp.up_proj",
699
+ "model.layers.55.mlp.down_proj",
700
+ "model.layers.56.self_attn.q_proj",
701
+ "model.layers.56.self_attn.k_proj",
702
+ "model.layers.56.self_attn.v_proj",
703
+ "model.layers.56.self_attn.o_proj",
704
+ "model.layers.56.mlp.gate_proj",
705
+ "model.layers.56.mlp.up_proj",
706
+ "model.layers.56.mlp.down_proj",
707
+ "model.layers.57.self_attn.q_proj",
708
+ "model.layers.57.self_attn.k_proj",
709
+ "model.layers.57.self_attn.v_proj",
710
+ "model.layers.57.self_attn.o_proj",
711
+ "model.layers.57.mlp.gate_proj",
712
+ "model.layers.57.mlp.up_proj",
713
+ "model.layers.57.mlp.down_proj",
714
+ "model.layers.58.self_attn.q_proj",
715
+ "model.layers.58.self_attn.k_proj",
716
+ "model.layers.58.self_attn.v_proj",
717
+ "model.layers.58.self_attn.o_proj",
718
+ "model.layers.58.mlp.gate_proj",
719
+ "model.layers.58.mlp.up_proj",
720
+ "model.layers.58.mlp.down_proj",
721
+ "model.layers.59.self_attn.q_proj",
722
+ "model.layers.59.self_attn.k_proj",
723
+ "model.layers.59.self_attn.v_proj",
724
+ "model.layers.59.self_attn.o_proj",
725
+ "model.layers.59.mlp.gate_proj",
726
+ "model.layers.59.mlp.up_proj",
727
+ "model.layers.59.mlp.down_proj",
728
+ "model.layers.60.self_attn.q_proj",
729
+ "model.layers.60.self_attn.k_proj",
730
+ "model.layers.60.self_attn.v_proj",
731
+ "model.layers.60.self_attn.o_proj",
732
+ "model.layers.60.mlp.gate_proj",
733
+ "model.layers.60.mlp.up_proj",
734
+ "model.layers.60.mlp.down_proj",
735
+ "model.layers.61.self_attn.q_proj",
736
+ "model.layers.61.self_attn.k_proj",
737
+ "model.layers.61.self_attn.v_proj",
738
+ "model.layers.61.self_attn.o_proj",
739
+ "model.layers.61.mlp.gate_proj",
740
+ "model.layers.61.mlp.up_proj",
741
+ "model.layers.61.mlp.down_proj",
742
+ "model.layers.62.self_attn.q_proj",
743
+ "model.layers.62.self_attn.k_proj",
744
+ "model.layers.62.self_attn.v_proj",
745
+ "model.layers.62.self_attn.o_proj",
746
+ "model.layers.62.mlp.gate_proj",
747
+ "model.layers.62.mlp.up_proj",
748
+ "model.layers.62.mlp.down_proj",
749
+ "model.layers.63.self_attn.q_proj",
750
+ "model.layers.63.self_attn.k_proj",
751
+ "model.layers.63.self_attn.v_proj",
752
+ "model.layers.63.self_attn.o_proj",
753
+ "model.layers.63.mlp.gate_proj",
754
+ "model.layers.63.mlp.up_proj",
755
+ "model.layers.63.mlp.down_proj",
756
+ "model.layers.64.self_attn.q_proj",
757
+ "model.layers.64.self_attn.k_proj",
758
+ "model.layers.64.self_attn.v_proj",
759
+ "model.layers.64.self_attn.o_proj",
760
+ "model.layers.64.mlp.gate_proj",
761
+ "model.layers.64.mlp.up_proj",
762
+ "model.layers.64.mlp.down_proj",
763
+ "model.layers.65.self_attn.q_proj",
764
+ "model.layers.65.self_attn.k_proj",
765
+ "model.layers.65.self_attn.v_proj",
766
+ "model.layers.65.self_attn.o_proj",
767
+ "model.layers.65.mlp.gate_proj",
768
+ "model.layers.65.mlp.up_proj",
769
+ "model.layers.65.mlp.down_proj",
770
+ "model.layers.66.self_attn.q_proj",
771
+ "model.layers.66.self_attn.k_proj",
772
+ "model.layers.66.self_attn.v_proj",
773
+ "model.layers.66.self_attn.o_proj",
774
+ "model.layers.66.mlp.gate_proj",
775
+ "model.layers.66.mlp.up_proj",
776
+ "model.layers.66.mlp.down_proj",
777
+ "model.layers.67.self_attn.q_proj",
778
+ "model.layers.67.self_attn.k_proj",
779
+ "model.layers.67.self_attn.v_proj",
780
+ "model.layers.67.self_attn.o_proj",
781
+ "model.layers.67.mlp.gate_proj",
782
+ "model.layers.67.mlp.up_proj",
783
+ "model.layers.67.mlp.down_proj",
784
+ "model.layers.68.self_attn.q_proj",
785
+ "model.layers.68.self_attn.k_proj",
786
+ "model.layers.68.self_attn.v_proj",
787
+ "model.layers.68.self_attn.o_proj",
788
+ "model.layers.68.mlp.gate_proj",
789
+ "model.layers.68.mlp.up_proj",
790
+ "model.layers.68.mlp.down_proj",
791
+ "model.layers.69.self_attn.q_proj",
792
+ "model.layers.69.self_attn.k_proj",
793
+ "model.layers.69.self_attn.v_proj",
794
+ "model.layers.69.self_attn.o_proj",
795
+ "model.layers.69.mlp.gate_proj",
796
+ "model.layers.69.mlp.up_proj",
797
+ "model.layers.69.mlp.down_proj",
798
+ "model.layers.70.self_attn.q_proj",
799
+ "model.layers.70.self_attn.k_proj",
800
+ "model.layers.70.self_attn.v_proj",
801
+ "model.layers.70.self_attn.o_proj",
802
+ "model.layers.70.mlp.gate_proj",
803
+ "model.layers.70.mlp.up_proj",
804
+ "model.layers.70.mlp.down_proj",
805
+ "model.layers.71.self_attn.q_proj",
806
+ "model.layers.71.self_attn.k_proj",
807
+ "model.layers.71.self_attn.v_proj",
808
+ "model.layers.71.self_attn.o_proj",
809
+ "model.layers.71.mlp.gate_proj",
810
+ "model.layers.71.mlp.up_proj",
811
+ "model.layers.71.mlp.down_proj",
812
+ "model.layers.72.self_attn.q_proj",
813
+ "model.layers.72.self_attn.k_proj",
814
+ "model.layers.72.self_attn.v_proj",
815
+ "model.layers.72.self_attn.o_proj",
816
+ "model.layers.72.mlp.gate_proj",
817
+ "model.layers.72.mlp.up_proj",
818
+ "model.layers.72.mlp.down_proj",
819
+ "model.layers.73.self_attn.q_proj",
820
+ "model.layers.73.self_attn.k_proj",
821
+ "model.layers.73.self_attn.v_proj",
822
+ "model.layers.73.self_attn.o_proj",
823
+ "model.layers.73.mlp.gate_proj",
824
+ "model.layers.73.mlp.up_proj",
825
+ "model.layers.73.mlp.down_proj",
826
+ "model.layers.74.self_attn.q_proj",
827
+ "model.layers.74.self_attn.k_proj",
828
+ "model.layers.74.self_attn.v_proj",
829
+ "model.layers.74.self_attn.o_proj",
830
+ "model.layers.74.mlp.gate_proj",
831
+ "model.layers.74.mlp.up_proj",
832
+ "model.layers.74.mlp.down_proj",
833
+ "model.layers.75.self_attn.q_proj",
834
+ "model.layers.75.self_attn.k_proj",
835
+ "model.layers.75.self_attn.v_proj",
836
+ "model.layers.75.self_attn.o_proj",
837
+ "model.layers.75.mlp.gate_proj",
838
+ "model.layers.75.mlp.up_proj",
839
+ "model.layers.75.mlp.down_proj",
840
+ "model.layers.76.self_attn.q_proj",
841
+ "model.layers.76.self_attn.k_proj",
842
+ "model.layers.76.self_attn.v_proj",
843
+ "model.layers.76.self_attn.o_proj",
844
+ "model.layers.76.mlp.gate_proj",
845
+ "model.layers.76.mlp.up_proj",
846
+ "model.layers.76.mlp.down_proj",
847
+ "model.layers.77.self_attn.q_proj",
848
+ "model.layers.77.self_attn.k_proj",
849
+ "model.layers.77.self_attn.v_proj",
850
+ "model.layers.77.self_attn.o_proj",
851
+ "model.layers.77.mlp.gate_proj",
852
+ "model.layers.77.mlp.up_proj",
853
+ "model.layers.77.mlp.down_proj",
854
+ "model.layers.78.self_attn.q_proj",
855
+ "model.layers.78.self_attn.k_proj",
856
+ "model.layers.78.self_attn.v_proj",
857
+ "model.layers.78.self_attn.o_proj",
858
+ "model.layers.78.mlp.gate_proj",
859
+ "model.layers.78.mlp.up_proj",
860
+ "model.layers.78.mlp.down_proj",
861
+ "model.layers.79.self_attn.q_proj",
862
+ "model.layers.79.self_attn.k_proj",
863
+ "model.layers.79.self_attn.v_proj",
864
+ "model.layers.79.self_attn.o_proj",
865
+ "model.layers.79.mlp.gate_proj",
866
+ "model.layers.79.mlp.up_proj",
867
+ "model.layers.79.mlp.down_proj",
868
+ "lm_head"
869
+ ],
870
+ "registry_requires_subclass": false,
871
+ "sparsity_structure": "unstructured",
872
+ "targets": [
873
+ "model.layers.1.self_attn.q_proj",
874
+ "model.layers.1.self_attn.k_proj",
875
+ "model.layers.2.self_attn.q_proj",
876
+ "model.layers.2.self_attn.k_proj",
877
+ "model.layers.3.self_attn.q_proj",
878
+ "model.layers.5.self_attn.q_proj"
879
+ ]
880
+ }
881
+ },
882
+ "rms_norm_eps": 1e-06,
883
+ "rope_scaling": {
884
+ "mrope_section": [
885
+ 16,
886
+ 24,
887
+ 24
888
+ ],
889
+ "rope_type": "default",
890
+ "type": "default"
891
+ },
892
+ "rope_theta": 1000000.0,
893
+ "sliding_window": 32768,
894
+ "tie_word_embeddings": false,
895
+ "torch_dtype": "bfloat16",
896
+ "transformers_version": "4.48.1",
897
+ "use_cache": true,
898
+ "use_sliding_window": false,
899
+ "video_token_id": 151656,
900
+ "vision_config": {
901
+ "hidden_size": 8192,
902
+ "in_chans": 3,
903
+ "model_type": "qwen2_vl",
904
+ "spatial_patch_size": 14
905
+ },
906
+ "vision_end_token_id": 151653,
907
+ "vision_start_token_id": 151652,
908
+ "vision_token_id": 151654,
909
+ "vocab_size": 152064
910
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 151643,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.05,
10
+ "top_k": 1,
11
+ "top_p": 0.001,
12
+ "transformers_version": "4.48.1"
13
+ }
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0aceb9c56fe71793960188acc88db8048a7dba241df591701081d84b63023f8
3
+ size 4998143368
model-00002-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53771ee9263e5fe1b8a5bd97d0b46ef4dcd0ebea8875fd72eebb35d0d6492616
3
+ size 4978586536
model-00003-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5103af8f27f1745f65f11ac3af4b62ae709eaf5e5d30abfcda30c5eb18469bb9
3
+ size 4978586728
model-00004-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:268750612712ca03b35171793bf3222cd8832eba974a624a5136467a46d9e46a
3
+ size 4978586728
model-00005-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01ffdf53de7def09b55034b7f67b175ffb3997f1cbc8c315ea3258b0bcdbf7d9
3
+ size 4978586728
model-00006-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32bd376650c22ee21b584151baf332d174f87983c5fb778da3fd400a5d04e418
3
+ size 4978586728
model-00007-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2f4ef21454d3fc12f18456a5394e2eb87d900851ca4650e1eb5dbaf8120b3a7
3
+ size 4978586728
model-00008-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35cd263f038a68e5bc795139f4a2705c8d42156ae18fa5ddda8b9356db6fb970
3
+ size 4978586728
model-00009-of-00009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4862d37dd0e87eb89786a3565a8175265258d78a25cba60036f234c0339b5ac2
3
+ size 2741257232
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
preprocessor_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": true,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.48145466,
8
+ 0.4578275,
9
+ 0.40821073
10
+ ],
11
+ "image_processor_type": "Qwen2VLImageProcessor",
12
+ "image_std": [
13
+ 0.26862954,
14
+ 0.26130258,
15
+ 0.27577711
16
+ ],
17
+ "max_pixels": 12845056,
18
+ "merge_size": 2,
19
+ "min_pixels": 3136,
20
+ "patch_size": 14,
21
+ "processor_class": "Qwen2VLProcessor",
22
+ "resample": 3,
23
+ "rescale_factor": 0.00392156862745098,
24
+ "size": {
25
+ "max_pixels": 12845056,
26
+ "min_pixels": 3136
27
+ },
28
+ "temporal_patch_size": 2
29
+ }
recipe.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ DEFAULT_stage:
2
+ DEFAULT_modifiers:
3
+ GPTQModifier:
4
+ sequential_targets: [Qwen2VLDecoderLayer]
5
+ scheme: W4A16
6
+ targets: Linear
7
+ ignore: [lm_head, 're:visual.*']
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:268c96d8611372617e20f95ad1eedab8c90f816f1f99ce13355e41b7700f7111
3
+ size 11420470
tokenizer_config.json ADDED
@@ -0,0 +1,145 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "151643": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "151644": {
13
+ "content": "<|im_start|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "151645": {
21
+ "content": "<|im_end|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "151646": {
29
+ "content": "<|object_ref_start|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "151647": {
37
+ "content": "<|object_ref_end|>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "151648": {
45
+ "content": "<|box_start|>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": true
51
+ },
52
+ "151649": {
53
+ "content": "<|box_end|>",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": true
59
+ },
60
+ "151650": {
61
+ "content": "<|quad_start|>",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": true
67
+ },
68
+ "151651": {
69
+ "content": "<|quad_end|>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "151652": {
77
+ "content": "<|vision_start|>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "151653": {
85
+ "content": "<|vision_end|>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": true
91
+ },
92
+ "151654": {
93
+ "content": "<|vision_pad|>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": true
99
+ },
100
+ "151655": {
101
+ "content": "<|image_pad|>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": true
107
+ },
108
+ "151656": {
109
+ "content": "<|video_pad|>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": true
115
+ }
116
+ },
117
+ "additional_special_tokens": [
118
+ "<|im_start|>",
119
+ "<|im_end|>",
120
+ "<|object_ref_start|>",
121
+ "<|object_ref_end|>",
122
+ "<|box_start|>",
123
+ "<|box_end|>",
124
+ "<|quad_start|>",
125
+ "<|quad_end|>",
126
+ "<|vision_start|>",
127
+ "<|vision_end|>",
128
+ "<|vision_pad|>",
129
+ "<|image_pad|>",
130
+ "<|video_pad|>"
131
+ ],
132
+ "bos_token": null,
133
+ "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}",
134
+ "clean_up_tokenization_spaces": false,
135
+ "eos_token": "<|im_end|>",
136
+ "errors": "replace",
137
+ "extra_special_tokens": {},
138
+ "model_max_length": 32768,
139
+ "pad_token": "<|endoftext|>",
140
+ "padding_side": "left",
141
+ "processor_class": "Qwen2VLProcessor",
142
+ "split_special_tokens": false,
143
+ "tokenizer_class": "Qwen2Tokenizer",
144
+ "unk_token": null
145
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff