Hugging Face's logo

Intel
/
BAGEL-7B-MoT-int4-AutoRound

Safetensors
bagel
Text-to-Image
4-bit precision
auto-round
Model card Files
xet
 Community
lianglv commited on
Commit
634107f
·
verified ·
1 Parent(s): b18a8ec

Upload folder using huggingface_hub

Browse files
Files changed (33) hide show
  1. .gitattributes +1 -0
  2. chat_template.jinja +54 -0
  3. config.json +1372 -0
  4. llm_config.json +27 -0
  5. model-00001-of-00022.safetensors +3 -0
  6. model-00002-of-00022.safetensors +3 -0
  7. model-00003-of-00022.safetensors +3 -0
  8. model-00004-of-00022.safetensors +3 -0
  9. model-00005-of-00022.safetensors +3 -0
  10. model-00006-of-00022.safetensors +3 -0
  11. model-00007-of-00022.safetensors +3 -0
  12. model-00008-of-00022.safetensors +3 -0
  13. model-00009-of-00022.safetensors +3 -0
  14. model-00010-of-00022.safetensors +3 -0
  15. model-00011-of-00022.safetensors +3 -0
  16. model-00012-of-00022.safetensors +3 -0
  17. model-00013-of-00022.safetensors +3 -0
  18. model-00014-of-00022.safetensors +3 -0
  19. model-00015-of-00022.safetensors +3 -0
  20. model-00016-of-00022.safetensors +3 -0
  21. model-00017-of-00022.safetensors +3 -0
  22. model-00018-of-00022.safetensors +3 -0
  23. model-00019-of-00022.safetensors +3 -0
  24. model-00020-of-00022.safetensors +3 -0
  25. model-00021-of-00022.safetensors +3 -0
  26. model-00022-of-00022.safetensors +3 -0
  27. model.safetensors.index.json +0 -0
  28. model_extra_tensors.safetensors +3 -0
  29. preprocessor_config.json +24 -0
  30. quantization_config.json +1265 -0
  31. tokenizer.json +3 -0
  32. tokenizer_config.json +29 -0
  33. vit_config.json +9 -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
chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,1372 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BagelForConditionalGeneration"
4
+ ],
5
+ "connector_act": "gelu_pytorch_tanh",
6
+ "dtype": "bfloat16",
7
+ "interpolate_pos": false,
8
+ "latent_patch_size": 2,
9
+ "llm_config": {
10
+ "_name_or_path": "",
11
+ "architectures": [
12
+ "Qwen2ForCausalLM"
13
+ ],
14
+ "attention_dropout": 0.0,
15
+ "bos_token_id": 151643,
16
+ "chunk_size_feed_forward": 0,
17
+ "dtype": "bfloat16",
18
+ "eos_token_id": 151645,
19
+ "hidden_act": "silu",
20
+ "hidden_size": 3584,
21
+ "id2label": {
22
+ "0": "LABEL_0",
23
+ "1": "LABEL_1"
24
+ },
25
+ "initializer_range": 0.02,
26
+ "intermediate_size": 18944,
27
+ "is_encoder_decoder": false,
28
+ "label2id": {
29
+ "LABEL_0": 0,
30
+ "LABEL_1": 1
31
+ },
32
+ "layer_types": [
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention"
61
+ ],
62
+ "max_position_embeddings": 32768,
63
+ "max_window_layers": 28,
64
+ "model_type": "qwen2",
65
+ "num_attention_heads": 28,
66
+ "num_hidden_layers": 28,
67
+ "num_key_value_heads": 4,
68
+ "output_attentions": false,
69
+ "output_hidden_states": false,
70
+ "pad_token_id": null,
71
+ "problem_type": null,
72
+ "qk_norm": true,
73
+ "return_dict": true,
74
+ "rms_norm_eps": 1e-06,
75
+ "rope_parameters": {
76
+ "rope_theta": 1000000.0,
77
+ "rope_type": "default"
78
+ },
79
+ "sliding_window": null,
80
+ "tie_word_embeddings": false,
81
+ "transformers_version": "5.3.0",
82
+ "use_cache": true,
83
+ "use_sliding_window": false,
84
+ "vocab_size": 152064
85
+ },
86
+ "max_latent_size": 32,
87
+ "model_type": "bagel",
88
+ "quantization_config": {
89
+ "autoround_version": "0.12.0",
90
+ "bits": 4,
91
+ "block_name_to_quantize": "language_model.model.layers",
92
+ "data_type": "int",
93
+ "extra_config": {
94
+ ".*moe_gen.*": {
95
+ "bits": 16,
96
+ "data_type": "float"
97
+ },
98
+ ".*self_attn\\.k_proj.*": {
99
+ "bits": 16,
100
+ "data_type": "float"
101
+ },
102
+ ".*self_attn\\.o_proj.*": {
103
+ "bits": 16,
104
+ "data_type": "float"
105
+ },
106
+ ".*self_attn\\.q_proj.*": {
107
+ "bits": 16,
108
+ "data_type": "float"
109
+ },
110
+ ".*self_attn\\.v_proj.*": {
111
+ "bits": 16,
112
+ "data_type": "float"
113
+ },
114
+ "language_model.model.layers.0.mlp_moe_gen.down_proj": {
115
+ "bits": 16,
116
+ "data_type": "float"
117
+ },
118
+ "language_model.model.layers.0.mlp_moe_gen.gate_proj": {
119
+ "bits": 16,
120
+ "data_type": "float"
121
+ },
122
+ "language_model.model.layers.0.mlp_moe_gen.up_proj": {
123
+ "bits": 16,
124
+ "data_type": "float"
125
+ },
126
+ "language_model.model.layers.0.self_attn.k_proj": {
127
+ "bits": 16,
128
+ "data_type": "float"
129
+ },
130
+ "language_model.model.layers.0.self_attn.k_proj_moe_gen": {
131
+ "bits": 16,
132
+ "data_type": "float"
133
+ },
134
+ "language_model.model.layers.0.self_attn.o_proj": {
135
+ "bits": 16,
136
+ "data_type": "float"
137
+ },
138
+ "language_model.model.layers.0.self_attn.o_proj_moe_gen": {
139
+ "bits": 16,
140
+ "data_type": "float"
141
+ },
142
+ "language_model.model.layers.0.self_attn.q_proj": {
143
+ "bits": 16,
144
+ "data_type": "float"
145
+ },
146
+ "language_model.model.layers.0.self_attn.q_proj_moe_gen": {
147
+ "bits": 16,
148
+ "data_type": "float"
149
+ },
150
+ "language_model.model.layers.0.self_attn.v_proj": {
151
+ "bits": 16,
152
+ "data_type": "float"
153
+ },
154
+ "language_model.model.layers.0.self_attn.v_proj_moe_gen": {
155
+ "bits": 16,
156
+ "data_type": "float"
157
+ },
158
+ "language_model.model.layers.1.mlp_moe_gen.down_proj": {
159
+ "bits": 16,
160
+ "data_type": "float"
161
+ },
162
+ "language_model.model.layers.1.mlp_moe_gen.gate_proj": {
163
+ "bits": 16,
164
+ "data_type": "float"
165
+ },
166
+ "language_model.model.layers.1.mlp_moe_gen.up_proj": {
167
+ "bits": 16,
168
+ "data_type": "float"
169
+ },
170
+ "language_model.model.layers.1.self_attn.k_proj": {
171
+ "bits": 16,
172
+ "data_type": "float"
173
+ },
174
+ "language_model.model.layers.1.self_attn.k_proj_moe_gen": {
175
+ "bits": 16,
176
+ "data_type": "float"
177
+ },
178
+ "language_model.model.layers.1.self_attn.o_proj": {
179
+ "bits": 16,
180
+ "data_type": "float"
181
+ },
182
+ "language_model.model.layers.1.self_attn.o_proj_moe_gen": {
183
+ "bits": 16,
184
+ "data_type": "float"
185
+ },
186
+ "language_model.model.layers.1.self_attn.q_proj": {
187
+ "bits": 16,
188
+ "data_type": "float"
189
+ },
190
+ "language_model.model.layers.1.self_attn.q_proj_moe_gen": {
191
+ "bits": 16,
192
+ "data_type": "float"
193
+ },
194
+ "language_model.model.layers.1.self_attn.v_proj": {
195
+ "bits": 16,
196
+ "data_type": "float"
197
+ },
198
+ "language_model.model.layers.1.self_attn.v_proj_moe_gen": {
199
+ "bits": 16,
200
+ "data_type": "float"
201
+ },
202
+ "language_model.model.layers.10.mlp_moe_gen.down_proj": {
203
+ "bits": 16,
204
+ "data_type": "float"
205
+ },
206
+ "language_model.model.layers.10.mlp_moe_gen.gate_proj": {
207
+ "bits": 16,
208
+ "data_type": "float"
209
+ },
210
+ "language_model.model.layers.10.mlp_moe_gen.up_proj": {
211
+ "bits": 16,
212
+ "data_type": "float"
213
+ },
214
+ "language_model.model.layers.10.self_attn.k_proj": {
215
+ "bits": 16,
216
+ "data_type": "float"
217
+ },
218
+ "language_model.model.layers.10.self_attn.k_proj_moe_gen": {
219
+ "bits": 16,
220
+ "data_type": "float"
221
+ },
222
+ "language_model.model.layers.10.self_attn.o_proj": {
223
+ "bits": 16,
224
+ "data_type": "float"
225
+ },
226
+ "language_model.model.layers.10.self_attn.o_proj_moe_gen": {
227
+ "bits": 16,
228
+ "data_type": "float"
229
+ },
230
+ "language_model.model.layers.10.self_attn.q_proj": {
231
+ "bits": 16,
232
+ "data_type": "float"
233
+ },
234
+ "language_model.model.layers.10.self_attn.q_proj_moe_gen": {
235
+ "bits": 16,
236
+ "data_type": "float"
237
+ },
238
+ "language_model.model.layers.10.self_attn.v_proj": {
239
+ "bits": 16,
240
+ "data_type": "float"
241
+ },
242
+ "language_model.model.layers.10.self_attn.v_proj_moe_gen": {
243
+ "bits": 16,
244
+ "data_type": "float"
245
+ },
246
+ "language_model.model.layers.11.mlp_moe_gen.down_proj": {
247
+ "bits": 16,
248
+ "data_type": "float"
249
+ },
250
+ "language_model.model.layers.11.mlp_moe_gen.gate_proj": {
251
+ "bits": 16,
252
+ "data_type": "float"
253
+ },
254
+ "language_model.model.layers.11.mlp_moe_gen.up_proj": {
255
+ "bits": 16,
256
+ "data_type": "float"
257
+ },
258
+ "language_model.model.layers.11.self_attn.k_proj": {
259
+ "bits": 16,
260
+ "data_type": "float"
261
+ },
262
+ "language_model.model.layers.11.self_attn.k_proj_moe_gen": {
263
+ "bits": 16,
264
+ "data_type": "float"
265
+ },
266
+ "language_model.model.layers.11.self_attn.o_proj": {
267
+ "bits": 16,
268
+ "data_type": "float"
269
+ },
270
+ "language_model.model.layers.11.self_attn.o_proj_moe_gen": {
271
+ "bits": 16,
272
+ "data_type": "float"
273
+ },
274
+ "language_model.model.layers.11.self_attn.q_proj": {
275
+ "bits": 16,
276
+ "data_type": "float"
277
+ },
278
+ "language_model.model.layers.11.self_attn.q_proj_moe_gen": {
279
+ "bits": 16,
280
+ "data_type": "float"
281
+ },
282
+ "language_model.model.layers.11.self_attn.v_proj": {
283
+ "bits": 16,
284
+ "data_type": "float"
285
+ },
286
+ "language_model.model.layers.11.self_attn.v_proj_moe_gen": {
287
+ "bits": 16,
288
+ "data_type": "float"
289
+ },
290
+ "language_model.model.layers.12.mlp_moe_gen.down_proj": {
291
+ "bits": 16,
292
+ "data_type": "float"
293
+ },
294
+ "language_model.model.layers.12.mlp_moe_gen.gate_proj": {
295
+ "bits": 16,
296
+ "data_type": "float"
297
+ },
298
+ "language_model.model.layers.12.mlp_moe_gen.up_proj": {
299
+ "bits": 16,
300
+ "data_type": "float"
301
+ },
302
+ "language_model.model.layers.12.self_attn.k_proj": {
303
+ "bits": 16,
304
+ "data_type": "float"
305
+ },
306
+ "language_model.model.layers.12.self_attn.k_proj_moe_gen": {
307
+ "bits": 16,
308
+ "data_type": "float"
309
+ },
310
+ "language_model.model.layers.12.self_attn.o_proj": {
311
+ "bits": 16,
312
+ "data_type": "float"
313
+ },
314
+ "language_model.model.layers.12.self_attn.o_proj_moe_gen": {
315
+ "bits": 16,
316
+ "data_type": "float"
317
+ },
318
+ "language_model.model.layers.12.self_attn.q_proj": {
319
+ "bits": 16,
320
+ "data_type": "float"
321
+ },
322
+ "language_model.model.layers.12.self_attn.q_proj_moe_gen": {
323
+ "bits": 16,
324
+ "data_type": "float"
325
+ },
326
+ "language_model.model.layers.12.self_attn.v_proj": {
327
+ "bits": 16,
328
+ "data_type": "float"
329
+ },
330
+ "language_model.model.layers.12.self_attn.v_proj_moe_gen": {
331
+ "bits": 16,
332
+ "data_type": "float"
333
+ },
334
+ "language_model.model.layers.13.mlp_moe_gen.down_proj": {
335
+ "bits": 16,
336
+ "data_type": "float"
337
+ },
338
+ "language_model.model.layers.13.mlp_moe_gen.gate_proj": {
339
+ "bits": 16,
340
+ "data_type": "float"
341
+ },
342
+ "language_model.model.layers.13.mlp_moe_gen.up_proj": {
343
+ "bits": 16,
344
+ "data_type": "float"
345
+ },
346
+ "language_model.model.layers.13.self_attn.k_proj": {
347
+ "bits": 16,
348
+ "data_type": "float"
349
+ },
350
+ "language_model.model.layers.13.self_attn.k_proj_moe_gen": {
351
+ "bits": 16,
352
+ "data_type": "float"
353
+ },
354
+ "language_model.model.layers.13.self_attn.o_proj": {
355
+ "bits": 16,
356
+ "data_type": "float"
357
+ },
358
+ "language_model.model.layers.13.self_attn.o_proj_moe_gen": {
359
+ "bits": 16,
360
+ "data_type": "float"
361
+ },
362
+ "language_model.model.layers.13.self_attn.q_proj": {
363
+ "bits": 16,
364
+ "data_type": "float"
365
+ },
366
+ "language_model.model.layers.13.self_attn.q_proj_moe_gen": {
367
+ "bits": 16,
368
+ "data_type": "float"
369
+ },
370
+ "language_model.model.layers.13.self_attn.v_proj": {
371
+ "bits": 16,
372
+ "data_type": "float"
373
+ },
374
+ "language_model.model.layers.13.self_attn.v_proj_moe_gen": {
375
+ "bits": 16,
376
+ "data_type": "float"
377
+ },
378
+ "language_model.model.layers.14.mlp_moe_gen.down_proj": {
379
+ "bits": 16,
380
+ "data_type": "float"
381
+ },
382
+ "language_model.model.layers.14.mlp_moe_gen.gate_proj": {
383
+ "bits": 16,
384
+ "data_type": "float"
385
+ },
386
+ "language_model.model.layers.14.mlp_moe_gen.up_proj": {
387
+ "bits": 16,
388
+ "data_type": "float"
389
+ },
390
+ "language_model.model.layers.14.self_attn.k_proj": {
391
+ "bits": 16,
392
+ "data_type": "float"
393
+ },
394
+ "language_model.model.layers.14.self_attn.k_proj_moe_gen": {
395
+ "bits": 16,
396
+ "data_type": "float"
397
+ },
398
+ "language_model.model.layers.14.self_attn.o_proj": {
399
+ "bits": 16,
400
+ "data_type": "float"
401
+ },
402
+ "language_model.model.layers.14.self_attn.o_proj_moe_gen": {
403
+ "bits": 16,
404
+ "data_type": "float"
405
+ },
406
+ "language_model.model.layers.14.self_attn.q_proj": {
407
+ "bits": 16,
408
+ "data_type": "float"
409
+ },
410
+ "language_model.model.layers.14.self_attn.q_proj_moe_gen": {
411
+ "bits": 16,
412
+ "data_type": "float"
413
+ },
414
+ "language_model.model.layers.14.self_attn.v_proj": {
415
+ "bits": 16,
416
+ "data_type": "float"
417
+ },
418
+ "language_model.model.layers.14.self_attn.v_proj_moe_gen": {
419
+ "bits": 16,
420
+ "data_type": "float"
421
+ },
422
+ "language_model.model.layers.15.mlp_moe_gen.down_proj": {
423
+ "bits": 16,
424
+ "data_type": "float"
425
+ },
426
+ "language_model.model.layers.15.mlp_moe_gen.gate_proj": {
427
+ "bits": 16,
428
+ "data_type": "float"
429
+ },
430
+ "language_model.model.layers.15.mlp_moe_gen.up_proj": {
431
+ "bits": 16,
432
+ "data_type": "float"
433
+ },
434
+ "language_model.model.layers.15.self_attn.k_proj": {
435
+ "bits": 16,
436
+ "data_type": "float"
437
+ },
438
+ "language_model.model.layers.15.self_attn.k_proj_moe_gen": {
439
+ "bits": 16,
440
+ "data_type": "float"
441
+ },
442
+ "language_model.model.layers.15.self_attn.o_proj": {
443
+ "bits": 16,
444
+ "data_type": "float"
445
+ },
446
+ "language_model.model.layers.15.self_attn.o_proj_moe_gen": {
447
+ "bits": 16,
448
+ "data_type": "float"
449
+ },
450
+ "language_model.model.layers.15.self_attn.q_proj": {
451
+ "bits": 16,
452
+ "data_type": "float"
453
+ },
454
+ "language_model.model.layers.15.self_attn.q_proj_moe_gen": {
455
+ "bits": 16,
456
+ "data_type": "float"
457
+ },
458
+ "language_model.model.layers.15.self_attn.v_proj": {
459
+ "bits": 16,
460
+ "data_type": "float"
461
+ },
462
+ "language_model.model.layers.15.self_attn.v_proj_moe_gen": {
463
+ "bits": 16,
464
+ "data_type": "float"
465
+ },
466
+ "language_model.model.layers.16.mlp_moe_gen.down_proj": {
467
+ "bits": 16,
468
+ "data_type": "float"
469
+ },
470
+ "language_model.model.layers.16.mlp_moe_gen.gate_proj": {
471
+ "bits": 16,
472
+ "data_type": "float"
473
+ },
474
+ "language_model.model.layers.16.mlp_moe_gen.up_proj": {
475
+ "bits": 16,
476
+ "data_type": "float"
477
+ },
478
+ "language_model.model.layers.16.self_attn.k_proj": {
479
+ "bits": 16,
480
+ "data_type": "float"
481
+ },
482
+ "language_model.model.layers.16.self_attn.k_proj_moe_gen": {
483
+ "bits": 16,
484
+ "data_type": "float"
485
+ },
486
+ "language_model.model.layers.16.self_attn.o_proj": {
487
+ "bits": 16,
488
+ "data_type": "float"
489
+ },
490
+ "language_model.model.layers.16.self_attn.o_proj_moe_gen": {
491
+ "bits": 16,
492
+ "data_type": "float"
493
+ },
494
+ "language_model.model.layers.16.self_attn.q_proj": {
495
+ "bits": 16,
496
+ "data_type": "float"
497
+ },
498
+ "language_model.model.layers.16.self_attn.q_proj_moe_gen": {
499
+ "bits": 16,
500
+ "data_type": "float"
501
+ },
502
+ "language_model.model.layers.16.self_attn.v_proj": {
503
+ "bits": 16,
504
+ "data_type": "float"
505
+ },
506
+ "language_model.model.layers.16.self_attn.v_proj_moe_gen": {
507
+ "bits": 16,
508
+ "data_type": "float"
509
+ },
510
+ "language_model.model.layers.17.mlp_moe_gen.down_proj": {
511
+ "bits": 16,
512
+ "data_type": "float"
513
+ },
514
+ "language_model.model.layers.17.mlp_moe_gen.gate_proj": {
515
+ "bits": 16,
516
+ "data_type": "float"
517
+ },
518
+ "language_model.model.layers.17.mlp_moe_gen.up_proj": {
519
+ "bits": 16,
520
+ "data_type": "float"
521
+ },
522
+ "language_model.model.layers.17.self_attn.k_proj": {
523
+ "bits": 16,
524
+ "data_type": "float"
525
+ },
526
+ "language_model.model.layers.17.self_attn.k_proj_moe_gen": {
527
+ "bits": 16,
528
+ "data_type": "float"
529
+ },
530
+ "language_model.model.layers.17.self_attn.o_proj": {
531
+ "bits": 16,
532
+ "data_type": "float"
533
+ },
534
+ "language_model.model.layers.17.self_attn.o_proj_moe_gen": {
535
+ "bits": 16,
536
+ "data_type": "float"
537
+ },
538
+ "language_model.model.layers.17.self_attn.q_proj": {
539
+ "bits": 16,
540
+ "data_type": "float"
541
+ },
542
+ "language_model.model.layers.17.self_attn.q_proj_moe_gen": {
543
+ "bits": 16,
544
+ "data_type": "float"
545
+ },
546
+ "language_model.model.layers.17.self_attn.v_proj": {
547
+ "bits": 16,
548
+ "data_type": "float"
549
+ },
550
+ "language_model.model.layers.17.self_attn.v_proj_moe_gen": {
551
+ "bits": 16,
552
+ "data_type": "float"
553
+ },
554
+ "language_model.model.layers.18.mlp_moe_gen.down_proj": {
555
+ "bits": 16,
556
+ "data_type": "float"
557
+ },
558
+ "language_model.model.layers.18.mlp_moe_gen.gate_proj": {
559
+ "bits": 16,
560
+ "data_type": "float"
561
+ },
562
+ "language_model.model.layers.18.mlp_moe_gen.up_proj": {
563
+ "bits": 16,
564
+ "data_type": "float"
565
+ },
566
+ "language_model.model.layers.18.self_attn.k_proj": {
567
+ "bits": 16,
568
+ "data_type": "float"
569
+ },
570
+ "language_model.model.layers.18.self_attn.k_proj_moe_gen": {
571
+ "bits": 16,
572
+ "data_type": "float"
573
+ },
574
+ "language_model.model.layers.18.self_attn.o_proj": {
575
+ "bits": 16,
576
+ "data_type": "float"
577
+ },
578
+ "language_model.model.layers.18.self_attn.o_proj_moe_gen": {
579
+ "bits": 16,
580
+ "data_type": "float"
581
+ },
582
+ "language_model.model.layers.18.self_attn.q_proj": {
583
+ "bits": 16,
584
+ "data_type": "float"
585
+ },
586
+ "language_model.model.layers.18.self_attn.q_proj_moe_gen": {
587
+ "bits": 16,
588
+ "data_type": "float"
589
+ },
590
+ "language_model.model.layers.18.self_attn.v_proj": {
591
+ "bits": 16,
592
+ "data_type": "float"
593
+ },
594
+ "language_model.model.layers.18.self_attn.v_proj_moe_gen": {
595
+ "bits": 16,
596
+ "data_type": "float"
597
+ },
598
+ "language_model.model.layers.19.mlp_moe_gen.down_proj": {
599
+ "bits": 16,
600
+ "data_type": "float"
601
+ },
602
+ "language_model.model.layers.19.mlp_moe_gen.gate_proj": {
603
+ "bits": 16,
604
+ "data_type": "float"
605
+ },
606
+ "language_model.model.layers.19.mlp_moe_gen.up_proj": {
607
+ "bits": 16,
608
+ "data_type": "float"
609
+ },
610
+ "language_model.model.layers.19.self_attn.k_proj": {
611
+ "bits": 16,
612
+ "data_type": "float"
613
+ },
614
+ "language_model.model.layers.19.self_attn.k_proj_moe_gen": {
615
+ "bits": 16,
616
+ "data_type": "float"
617
+ },
618
+ "language_model.model.layers.19.self_attn.o_proj": {
619
+ "bits": 16,
620
+ "data_type": "float"
621
+ },
622
+ "language_model.model.layers.19.self_attn.o_proj_moe_gen": {
623
+ "bits": 16,
624
+ "data_type": "float"
625
+ },
626
+ "language_model.model.layers.19.self_attn.q_proj": {
627
+ "bits": 16,
628
+ "data_type": "float"
629
+ },
630
+ "language_model.model.layers.19.self_attn.q_proj_moe_gen": {
631
+ "bits": 16,
632
+ "data_type": "float"
633
+ },
634
+ "language_model.model.layers.19.self_attn.v_proj": {
635
+ "bits": 16,
636
+ "data_type": "float"
637
+ },
638
+ "language_model.model.layers.19.self_attn.v_proj_moe_gen": {
639
+ "bits": 16,
640
+ "data_type": "float"
641
+ },
642
+ "language_model.model.layers.2.mlp_moe_gen.down_proj": {
643
+ "bits": 16,
644
+ "data_type": "float"
645
+ },
646
+ "language_model.model.layers.2.mlp_moe_gen.gate_proj": {
647
+ "bits": 16,
648
+ "data_type": "float"
649
+ },
650
+ "language_model.model.layers.2.mlp_moe_gen.up_proj": {
651
+ "bits": 16,
652
+ "data_type": "float"
653
+ },
654
+ "language_model.model.layers.2.self_attn.k_proj": {
655
+ "bits": 16,
656
+ "data_type": "float"
657
+ },
658
+ "language_model.model.layers.2.self_attn.k_proj_moe_gen": {
659
+ "bits": 16,
660
+ "data_type": "float"
661
+ },
662
+ "language_model.model.layers.2.self_attn.o_proj": {
663
+ "bits": 16,
664
+ "data_type": "float"
665
+ },
666
+ "language_model.model.layers.2.self_attn.o_proj_moe_gen": {
667
+ "bits": 16,
668
+ "data_type": "float"
669
+ },
670
+ "language_model.model.layers.2.self_attn.q_proj": {
671
+ "bits": 16,
672
+ "data_type": "float"
673
+ },
674
+ "language_model.model.layers.2.self_attn.q_proj_moe_gen": {
675
+ "bits": 16,
676
+ "data_type": "float"
677
+ },
678
+ "language_model.model.layers.2.self_attn.v_proj": {
679
+ "bits": 16,
680
+ "data_type": "float"
681
+ },
682
+ "language_model.model.layers.2.self_attn.v_proj_moe_gen": {
683
+ "bits": 16,
684
+ "data_type": "float"
685
+ },
686
+ "language_model.model.layers.20.mlp_moe_gen.down_proj": {
687
+ "bits": 16,
688
+ "data_type": "float"
689
+ },
690
+ "language_model.model.layers.20.mlp_moe_gen.gate_proj": {
691
+ "bits": 16,
692
+ "data_type": "float"
693
+ },
694
+ "language_model.model.layers.20.mlp_moe_gen.up_proj": {
695
+ "bits": 16,
696
+ "data_type": "float"
697
+ },
698
+ "language_model.model.layers.20.self_attn.k_proj": {
699
+ "bits": 16,
700
+ "data_type": "float"
701
+ },
702
+ "language_model.model.layers.20.self_attn.k_proj_moe_gen": {
703
+ "bits": 16,
704
+ "data_type": "float"
705
+ },
706
+ "language_model.model.layers.20.self_attn.o_proj": {
707
+ "bits": 16,
708
+ "data_type": "float"
709
+ },
710
+ "language_model.model.layers.20.self_attn.o_proj_moe_gen": {
711
+ "bits": 16,
712
+ "data_type": "float"
713
+ },
714
+ "language_model.model.layers.20.self_attn.q_proj": {
715
+ "bits": 16,
716
+ "data_type": "float"
717
+ },
718
+ "language_model.model.layers.20.self_attn.q_proj_moe_gen": {
719
+ "bits": 16,
720
+ "data_type": "float"
721
+ },
722
+ "language_model.model.layers.20.self_attn.v_proj": {
723
+ "bits": 16,
724
+ "data_type": "float"
725
+ },
726
+ "language_model.model.layers.20.self_attn.v_proj_moe_gen": {
727
+ "bits": 16,
728
+ "data_type": "float"
729
+ },
730
+ "language_model.model.layers.21.mlp_moe_gen.down_proj": {
731
+ "bits": 16,
732
+ "data_type": "float"
733
+ },
734
+ "language_model.model.layers.21.mlp_moe_gen.gate_proj": {
735
+ "bits": 16,
736
+ "data_type": "float"
737
+ },
738
+ "language_model.model.layers.21.mlp_moe_gen.up_proj": {
739
+ "bits": 16,
740
+ "data_type": "float"
741
+ },
742
+ "language_model.model.layers.21.self_attn.k_proj": {
743
+ "bits": 16,
744
+ "data_type": "float"
745
+ },
746
+ "language_model.model.layers.21.self_attn.k_proj_moe_gen": {
747
+ "bits": 16,
748
+ "data_type": "float"
749
+ },
750
+ "language_model.model.layers.21.self_attn.o_proj": {
751
+ "bits": 16,
752
+ "data_type": "float"
753
+ },
754
+ "language_model.model.layers.21.self_attn.o_proj_moe_gen": {
755
+ "bits": 16,
756
+ "data_type": "float"
757
+ },
758
+ "language_model.model.layers.21.self_attn.q_proj": {
759
+ "bits": 16,
760
+ "data_type": "float"
761
+ },
762
+ "language_model.model.layers.21.self_attn.q_proj_moe_gen": {
763
+ "bits": 16,
764
+ "data_type": "float"
765
+ },
766
+ "language_model.model.layers.21.self_attn.v_proj": {
767
+ "bits": 16,
768
+ "data_type": "float"
769
+ },
770
+ "language_model.model.layers.21.self_attn.v_proj_moe_gen": {
771
+ "bits": 16,
772
+ "data_type": "float"
773
+ },
774
+ "language_model.model.layers.22.mlp_moe_gen.down_proj": {
775
+ "bits": 16,
776
+ "data_type": "float"
777
+ },
778
+ "language_model.model.layers.22.mlp_moe_gen.gate_proj": {
779
+ "bits": 16,
780
+ "data_type": "float"
781
+ },
782
+ "language_model.model.layers.22.mlp_moe_gen.up_proj": {
783
+ "bits": 16,
784
+ "data_type": "float"
785
+ },
786
+ "language_model.model.layers.22.self_attn.k_proj": {
787
+ "bits": 16,
788
+ "data_type": "float"
789
+ },
790
+ "language_model.model.layers.22.self_attn.k_proj_moe_gen": {
791
+ "bits": 16,
792
+ "data_type": "float"
793
+ },
794
+ "language_model.model.layers.22.self_attn.o_proj": {
795
+ "bits": 16,
796
+ "data_type": "float"
797
+ },
798
+ "language_model.model.layers.22.self_attn.o_proj_moe_gen": {
799
+ "bits": 16,
800
+ "data_type": "float"
801
+ },
802
+ "language_model.model.layers.22.self_attn.q_proj": {
803
+ "bits": 16,
804
+ "data_type": "float"
805
+ },
806
+ "language_model.model.layers.22.self_attn.q_proj_moe_gen": {
807
+ "bits": 16,
808
+ "data_type": "float"
809
+ },
810
+ "language_model.model.layers.22.self_attn.v_proj": {
811
+ "bits": 16,
812
+ "data_type": "float"
813
+ },
814
+ "language_model.model.layers.22.self_attn.v_proj_moe_gen": {
815
+ "bits": 16,
816
+ "data_type": "float"
817
+ },
818
+ "language_model.model.layers.23.mlp_moe_gen.down_proj": {
819
+ "bits": 16,
820
+ "data_type": "float"
821
+ },
822
+ "language_model.model.layers.23.mlp_moe_gen.gate_proj": {
823
+ "bits": 16,
824
+ "data_type": "float"
825
+ },
826
+ "language_model.model.layers.23.mlp_moe_gen.up_proj": {
827
+ "bits": 16,
828
+ "data_type": "float"
829
+ },
830
+ "language_model.model.layers.23.self_attn.k_proj": {
831
+ "bits": 16,
832
+ "data_type": "float"
833
+ },
834
+ "language_model.model.layers.23.self_attn.k_proj_moe_gen": {
835
+ "bits": 16,
836
+ "data_type": "float"
837
+ },
838
+ "language_model.model.layers.23.self_attn.o_proj": {
839
+ "bits": 16,
840
+ "data_type": "float"
841
+ },
842
+ "language_model.model.layers.23.self_attn.o_proj_moe_gen": {
843
+ "bits": 16,
844
+ "data_type": "float"
845
+ },
846
+ "language_model.model.layers.23.self_attn.q_proj": {
847
+ "bits": 16,
848
+ "data_type": "float"
849
+ },
850
+ "language_model.model.layers.23.self_attn.q_proj_moe_gen": {
851
+ "bits": 16,
852
+ "data_type": "float"
853
+ },
854
+ "language_model.model.layers.23.self_attn.v_proj": {
855
+ "bits": 16,
856
+ "data_type": "float"
857
+ },
858
+ "language_model.model.layers.23.self_attn.v_proj_moe_gen": {
859
+ "bits": 16,
860
+ "data_type": "float"
861
+ },
862
+ "language_model.model.layers.24.mlp_moe_gen.down_proj": {
863
+ "bits": 16,
864
+ "data_type": "float"
865
+ },
866
+ "language_model.model.layers.24.mlp_moe_gen.gate_proj": {
867
+ "bits": 16,
868
+ "data_type": "float"
869
+ },
870
+ "language_model.model.layers.24.mlp_moe_gen.up_proj": {
871
+ "bits": 16,
872
+ "data_type": "float"
873
+ },
874
+ "language_model.model.layers.24.self_attn.k_proj": {
875
+ "bits": 16,
876
+ "data_type": "float"
877
+ },
878
+ "language_model.model.layers.24.self_attn.k_proj_moe_gen": {
879
+ "bits": 16,
880
+ "data_type": "float"
881
+ },
882
+ "language_model.model.layers.24.self_attn.o_proj": {
883
+ "bits": 16,
884
+ "data_type": "float"
885
+ },
886
+ "language_model.model.layers.24.self_attn.o_proj_moe_gen": {
887
+ "bits": 16,
888
+ "data_type": "float"
889
+ },
890
+ "language_model.model.layers.24.self_attn.q_proj": {
891
+ "bits": 16,
892
+ "data_type": "float"
893
+ },
894
+ "language_model.model.layers.24.self_attn.q_proj_moe_gen": {
895
+ "bits": 16,
896
+ "data_type": "float"
897
+ },
898
+ "language_model.model.layers.24.self_attn.v_proj": {
899
+ "bits": 16,
900
+ "data_type": "float"
901
+ },
902
+ "language_model.model.layers.24.self_attn.v_proj_moe_gen": {
903
+ "bits": 16,
904
+ "data_type": "float"
905
+ },
906
+ "language_model.model.layers.25.mlp_moe_gen.down_proj": {
907
+ "bits": 16,
908
+ "data_type": "float"
909
+ },
910
+ "language_model.model.layers.25.mlp_moe_gen.gate_proj": {
911
+ "bits": 16,
912
+ "data_type": "float"
913
+ },
914
+ "language_model.model.layers.25.mlp_moe_gen.up_proj": {
915
+ "bits": 16,
916
+ "data_type": "float"
917
+ },
918
+ "language_model.model.layers.25.self_attn.k_proj": {
919
+ "bits": 16,
920
+ "data_type": "float"
921
+ },
922
+ "language_model.model.layers.25.self_attn.k_proj_moe_gen": {
923
+ "bits": 16,
924
+ "data_type": "float"
925
+ },
926
+ "language_model.model.layers.25.self_attn.o_proj": {
927
+ "bits": 16,
928
+ "data_type": "float"
929
+ },
930
+ "language_model.model.layers.25.self_attn.o_proj_moe_gen": {
931
+ "bits": 16,
932
+ "data_type": "float"
933
+ },
934
+ "language_model.model.layers.25.self_attn.q_proj": {
935
+ "bits": 16,
936
+ "data_type": "float"
937
+ },
938
+ "language_model.model.layers.25.self_attn.q_proj_moe_gen": {
939
+ "bits": 16,
940
+ "data_type": "float"
941
+ },
942
+ "language_model.model.layers.25.self_attn.v_proj": {
943
+ "bits": 16,
944
+ "data_type": "float"
945
+ },
946
+ "language_model.model.layers.25.self_attn.v_proj_moe_gen": {
947
+ "bits": 16,
948
+ "data_type": "float"
949
+ },
950
+ "language_model.model.layers.26.mlp_moe_gen.down_proj": {
951
+ "bits": 16,
952
+ "data_type": "float"
953
+ },
954
+ "language_model.model.layers.26.mlp_moe_gen.gate_proj": {
955
+ "bits": 16,
956
+ "data_type": "float"
957
+ },
958
+ "language_model.model.layers.26.mlp_moe_gen.up_proj": {
959
+ "bits": 16,
960
+ "data_type": "float"
961
+ },
962
+ "language_model.model.layers.26.self_attn.k_proj": {
963
+ "bits": 16,
964
+ "data_type": "float"
965
+ },
966
+ "language_model.model.layers.26.self_attn.k_proj_moe_gen": {
967
+ "bits": 16,
968
+ "data_type": "float"
969
+ },
970
+ "language_model.model.layers.26.self_attn.o_proj": {
971
+ "bits": 16,
972
+ "data_type": "float"
973
+ },
974
+ "language_model.model.layers.26.self_attn.o_proj_moe_gen": {
975
+ "bits": 16,
976
+ "data_type": "float"
977
+ },
978
+ "language_model.model.layers.26.self_attn.q_proj": {
979
+ "bits": 16,
980
+ "data_type": "float"
981
+ },
982
+ "language_model.model.layers.26.self_attn.q_proj_moe_gen": {
983
+ "bits": 16,
984
+ "data_type": "float"
985
+ },
986
+ "language_model.model.layers.26.self_attn.v_proj": {
987
+ "bits": 16,
988
+ "data_type": "float"
989
+ },
990
+ "language_model.model.layers.26.self_attn.v_proj_moe_gen": {
991
+ "bits": 16,
992
+ "data_type": "float"
993
+ },
994
+ "language_model.model.layers.27.mlp_moe_gen.down_proj": {
995
+ "bits": 16,
996
+ "data_type": "float"
997
+ },
998
+ "language_model.model.layers.27.mlp_moe_gen.gate_proj": {
999
+ "bits": 16,
1000
+ "data_type": "float"
1001
+ },
1002
+ "language_model.model.layers.27.mlp_moe_gen.up_proj": {
1003
+ "bits": 16,
1004
+ "data_type": "float"
1005
+ },
1006
+ "language_model.model.layers.27.self_attn.k_proj": {
1007
+ "bits": 16,
1008
+ "data_type": "float"
1009
+ },
1010
+ "language_model.model.layers.27.self_attn.k_proj_moe_gen": {
1011
+ "bits": 16,
1012
+ "data_type": "float"
1013
+ },
1014
+ "language_model.model.layers.27.self_attn.o_proj": {
1015
+ "bits": 16,
1016
+ "data_type": "float"
1017
+ },
1018
+ "language_model.model.layers.27.self_attn.o_proj_moe_gen": {
1019
+ "bits": 16,
1020
+ "data_type": "float"
1021
+ },
1022
+ "language_model.model.layers.27.self_attn.q_proj": {
1023
+ "bits": 16,
1024
+ "data_type": "float"
1025
+ },
1026
+ "language_model.model.layers.27.self_attn.q_proj_moe_gen": {
1027
+ "bits": 16,
1028
+ "data_type": "float"
1029
+ },
1030
+ "language_model.model.layers.27.self_attn.v_proj": {
1031
+ "bits": 16,
1032
+ "data_type": "float"
1033
+ },
1034
+ "language_model.model.layers.27.self_attn.v_proj_moe_gen": {
1035
+ "bits": 16,
1036
+ "data_type": "float"
1037
+ },
1038
+ "language_model.model.layers.3.mlp_moe_gen.down_proj": {
1039
+ "bits": 16,
1040
+ "data_type": "float"
1041
+ },
1042
+ "language_model.model.layers.3.mlp_moe_gen.gate_proj": {
1043
+ "bits": 16,
1044
+ "data_type": "float"
1045
+ },
1046
+ "language_model.model.layers.3.mlp_moe_gen.up_proj": {
1047
+ "bits": 16,
1048
+ "data_type": "float"
1049
+ },
1050
+ "language_model.model.layers.3.self_attn.k_proj": {
1051
+ "bits": 16,
1052
+ "data_type": "float"
1053
+ },
1054
+ "language_model.model.layers.3.self_attn.k_proj_moe_gen": {
1055
+ "bits": 16,
1056
+ "data_type": "float"
1057
+ },
1058
+ "language_model.model.layers.3.self_attn.o_proj": {
1059
+ "bits": 16,
1060
+ "data_type": "float"
1061
+ },
1062
+ "language_model.model.layers.3.self_attn.o_proj_moe_gen": {
1063
+ "bits": 16,
1064
+ "data_type": "float"
1065
+ },
1066
+ "language_model.model.layers.3.self_attn.q_proj": {
1067
+ "bits": 16,
1068
+ "data_type": "float"
1069
+ },
1070
+ "language_model.model.layers.3.self_attn.q_proj_moe_gen": {
1071
+ "bits": 16,
1072
+ "data_type": "float"
1073
+ },
1074
+ "language_model.model.layers.3.self_attn.v_proj": {
1075
+ "bits": 16,
1076
+ "data_type": "float"
1077
+ },
1078
+ "language_model.model.layers.3.self_attn.v_proj_moe_gen": {
1079
+ "bits": 16,
1080
+ "data_type": "float"
1081
+ },
1082
+ "language_model.model.layers.4.mlp_moe_gen.down_proj": {
1083
+ "bits": 16,
1084
+ "data_type": "float"
1085
+ },
1086
+ "language_model.model.layers.4.mlp_moe_gen.gate_proj": {
1087
+ "bits": 16,
1088
+ "data_type": "float"
1089
+ },
1090
+ "language_model.model.layers.4.mlp_moe_gen.up_proj": {
1091
+ "bits": 16,
1092
+ "data_type": "float"
1093
+ },
1094
+ "language_model.model.layers.4.self_attn.k_proj": {
1095
+ "bits": 16,
1096
+ "data_type": "float"
1097
+ },
1098
+ "language_model.model.layers.4.self_attn.k_proj_moe_gen": {
1099
+ "bits": 16,
1100
+ "data_type": "float"
1101
+ },
1102
+ "language_model.model.layers.4.self_attn.o_proj": {
1103
+ "bits": 16,
1104
+ "data_type": "float"
1105
+ },
1106
+ "language_model.model.layers.4.self_attn.o_proj_moe_gen": {
1107
+ "bits": 16,
1108
+ "data_type": "float"
1109
+ },
1110
+ "language_model.model.layers.4.self_attn.q_proj": {
1111
+ "bits": 16,
1112
+ "data_type": "float"
1113
+ },
1114
+ "language_model.model.layers.4.self_attn.q_proj_moe_gen": {
1115
+ "bits": 16,
1116
+ "data_type": "float"
1117
+ },
1118
+ "language_model.model.layers.4.self_attn.v_proj": {
1119
+ "bits": 16,
1120
+ "data_type": "float"
1121
+ },
1122
+ "language_model.model.layers.4.self_attn.v_proj_moe_gen": {
1123
+ "bits": 16,
1124
+ "data_type": "float"
1125
+ },
1126
+ "language_model.model.layers.5.mlp_moe_gen.down_proj": {
1127
+ "bits": 16,
1128
+ "data_type": "float"
1129
+ },
1130
+ "language_model.model.layers.5.mlp_moe_gen.gate_proj": {
1131
+ "bits": 16,
1132
+ "data_type": "float"
1133
+ },
1134
+ "language_model.model.layers.5.mlp_moe_gen.up_proj": {
1135
+ "bits": 16,
1136
+ "data_type": "float"
1137
+ },
1138
+ "language_model.model.layers.5.self_attn.k_proj": {
1139
+ "bits": 16,
1140
+ "data_type": "float"
1141
+ },
1142
+ "language_model.model.layers.5.self_attn.k_proj_moe_gen": {
1143
+ "bits": 16,
1144
+ "data_type": "float"
1145
+ },
1146
+ "language_model.model.layers.5.self_attn.o_proj": {
1147
+ "bits": 16,
1148
+ "data_type": "float"
1149
+ },
1150
+ "language_model.model.layers.5.self_attn.o_proj_moe_gen": {
1151
+ "bits": 16,
1152
+ "data_type": "float"
1153
+ },
1154
+ "language_model.model.layers.5.self_attn.q_proj": {
1155
+ "bits": 16,
1156
+ "data_type": "float"
1157
+ },
1158
+ "language_model.model.layers.5.self_attn.q_proj_moe_gen": {
1159
+ "bits": 16,
1160
+ "data_type": "float"
1161
+ },
1162
+ "language_model.model.layers.5.self_attn.v_proj": {
1163
+ "bits": 16,
1164
+ "data_type": "float"
1165
+ },
1166
+ "language_model.model.layers.5.self_attn.v_proj_moe_gen": {
1167
+ "bits": 16,
1168
+ "data_type": "float"
1169
+ },
1170
+ "language_model.model.layers.6.mlp_moe_gen.down_proj": {
1171
+ "bits": 16,
1172
+ "data_type": "float"
1173
+ },
1174
+ "language_model.model.layers.6.mlp_moe_gen.gate_proj": {
1175
+ "bits": 16,
1176
+ "data_type": "float"
1177
+ },
1178
+ "language_model.model.layers.6.mlp_moe_gen.up_proj": {
1179
+ "bits": 16,
1180
+ "data_type": "float"
1181
+ },
1182
+ "language_model.model.layers.6.self_attn.k_proj": {
1183
+ "bits": 16,
1184
+ "data_type": "float"
1185
+ },
1186
+ "language_model.model.layers.6.self_attn.k_proj_moe_gen": {
1187
+ "bits": 16,
1188
+ "data_type": "float"
1189
+ },
1190
+ "language_model.model.layers.6.self_attn.o_proj": {
1191
+ "bits": 16,
1192
+ "data_type": "float"
1193
+ },
1194
+ "language_model.model.layers.6.self_attn.o_proj_moe_gen": {
1195
+ "bits": 16,
1196
+ "data_type": "float"
1197
+ },
1198
+ "language_model.model.layers.6.self_attn.q_proj": {
1199
+ "bits": 16,
1200
+ "data_type": "float"
1201
+ },
1202
+ "language_model.model.layers.6.self_attn.q_proj_moe_gen": {
1203
+ "bits": 16,
1204
+ "data_type": "float"
1205
+ },
1206
+ "language_model.model.layers.6.self_attn.v_proj": {
1207
+ "bits": 16,
1208
+ "data_type": "float"
1209
+ },
1210
+ "language_model.model.layers.6.self_attn.v_proj_moe_gen": {
1211
+ "bits": 16,
1212
+ "data_type": "float"
1213
+ },
1214
+ "language_model.model.layers.7.mlp_moe_gen.down_proj": {
1215
+ "bits": 16,
1216
+ "data_type": "float"
1217
+ },
1218
+ "language_model.model.layers.7.mlp_moe_gen.gate_proj": {
1219
+ "bits": 16,
1220
+ "data_type": "float"
1221
+ },
1222
+ "language_model.model.layers.7.mlp_moe_gen.up_proj": {
1223
+ "bits": 16,
1224
+ "data_type": "float"
1225
+ },
1226
+ "language_model.model.layers.7.self_attn.k_proj": {
1227
+ "bits": 16,
1228
+ "data_type": "float"
1229
+ },
1230
+ "language_model.model.layers.7.self_attn.k_proj_moe_gen": {
1231
+ "bits": 16,
1232
+ "data_type": "float"
1233
+ },
1234
+ "language_model.model.layers.7.self_attn.o_proj": {
1235
+ "bits": 16,
1236
+ "data_type": "float"
1237
+ },
1238
+ "language_model.model.layers.7.self_attn.o_proj_moe_gen": {
1239
+ "bits": 16,
1240
+ "data_type": "float"
1241
+ },
1242
+ "language_model.model.layers.7.self_attn.q_proj": {
1243
+ "bits": 16,
1244
+ "data_type": "float"
1245
+ },
1246
+ "language_model.model.layers.7.self_attn.q_proj_moe_gen": {
1247
+ "bits": 16,
1248
+ "data_type": "float"
1249
+ },
1250
+ "language_model.model.layers.7.self_attn.v_proj": {
1251
+ "bits": 16,
1252
+ "data_type": "float"
1253
+ },
1254
+ "language_model.model.layers.7.self_attn.v_proj_moe_gen": {
1255
+ "bits": 16,
1256
+ "data_type": "float"
1257
+ },
1258
+ "language_model.model.layers.8.mlp_moe_gen.down_proj": {
1259
+ "bits": 16,
1260
+ "data_type": "float"
1261
+ },
1262
+ "language_model.model.layers.8.mlp_moe_gen.gate_proj": {
1263
+ "bits": 16,
1264
+ "data_type": "float"
1265
+ },
1266
+ "language_model.model.layers.8.mlp_moe_gen.up_proj": {
1267
+ "bits": 16,
1268
+ "data_type": "float"
1269
+ },
1270
+ "language_model.model.layers.8.self_attn.k_proj": {
1271
+ "bits": 16,
1272
+ "data_type": "float"
1273
+ },
1274
+ "language_model.model.layers.8.self_attn.k_proj_moe_gen": {
1275
+ "bits": 16,
1276
+ "data_type": "float"
1277
+ },
1278
+ "language_model.model.layers.8.self_attn.o_proj": {
1279
+ "bits": 16,
1280
+ "data_type": "float"
1281
+ },
1282
+ "language_model.model.layers.8.self_attn.o_proj_moe_gen": {
1283
+ "bits": 16,
1284
+ "data_type": "float"
1285
+ },
1286
+ "language_model.model.layers.8.self_attn.q_proj": {
1287
+ "bits": 16,
1288
+ "data_type": "float"
1289
+ },
1290
+ "language_model.model.layers.8.self_attn.q_proj_moe_gen": {
1291
+ "bits": 16,
1292
+ "data_type": "float"
1293
+ },
1294
+ "language_model.model.layers.8.self_attn.v_proj": {
1295
+ "bits": 16,
1296
+ "data_type": "float"
1297
+ },
1298
+ "language_model.model.layers.8.self_attn.v_proj_moe_gen": {
1299
+ "bits": 16,
1300
+ "data_type": "float"
1301
+ },
1302
+ "language_model.model.layers.9.mlp_moe_gen.down_proj": {
1303
+ "bits": 16,
1304
+ "data_type": "float"
1305
+ },
1306
+ "language_model.model.layers.9.mlp_moe_gen.gate_proj": {
1307
+ "bits": 16,
1308
+ "data_type": "float"
1309
+ },
1310
+ "language_model.model.layers.9.mlp_moe_gen.up_proj": {
1311
+ "bits": 16,
1312
+ "data_type": "float"
1313
+ },
1314
+ "language_model.model.layers.9.self_attn.k_proj": {
1315
+ "bits": 16,
1316
+ "data_type": "float"
1317
+ },
1318
+ "language_model.model.layers.9.self_attn.k_proj_moe_gen": {
1319
+ "bits": 16,
1320
+ "data_type": "float"
1321
+ },
1322
+ "language_model.model.layers.9.self_attn.o_proj": {
1323
+ "bits": 16,
1324
+ "data_type": "float"
1325
+ },
1326
+ "language_model.model.layers.9.self_attn.o_proj_moe_gen": {
1327
+ "bits": 16,
1328
+ "data_type": "float"
1329
+ },
1330
+ "language_model.model.layers.9.self_attn.q_proj": {
1331
+ "bits": 16,
1332
+ "data_type": "float"
1333
+ },
1334
+ "language_model.model.layers.9.self_attn.q_proj_moe_gen": {
1335
+ "bits": 16,
1336
+ "data_type": "float"
1337
+ },
1338
+ "language_model.model.layers.9.self_attn.v_proj": {
1339
+ "bits": 16,
1340
+ "data_type": "float"
1341
+ },
1342
+ "language_model.model.layers.9.self_attn.v_proj_moe_gen": {
1343
+ "bits": 16,
1344
+ "data_type": "float"
1345
+ }
1346
+ },
1347
+ "group_size": 128,
1348
+ "packing_format": "auto_round:auto_gptq",
1349
+ "quant_method": "auto-round",
1350
+ "sym": true,
1351
+ "transform_config": {}
1352
+ },
1353
+ "timestep_shift": 1.0,
1354
+ "transformers_version": "5.3.0",
1355
+ "vae_config": {
1356
+ "downsample": 8,
1357
+ "z_channels": 16
1358
+ },
1359
+ "visual_gen": true,
1360
+ "visual_und": true,
1361
+ "vit_config": {
1362
+ "hidden_size": 1152,
1363
+ "image_size": 980,
1364
+ "intermediate_size": 4304,
1365
+ "model_type": "siglip_vision_model",
1366
+ "num_attention_heads": 16,
1367
+ "num_channels": 3,
1368
+ "num_hidden_layers": 27,
1369
+ "patch_size": 14
1370
+ },
1371
+ "vit_max_num_patch_per_side": 70
1372
+ }
llm_config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2ForCausalLM"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": 151643,
7
+ "eos_token_id": 151645,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 3584,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 18944,
12
+ "max_position_embeddings": 32768,
13
+ "max_window_layers": 28,
14
+ "model_type": "qwen2",
15
+ "num_attention_heads": 28,
16
+ "num_hidden_layers": 28,
17
+ "num_key_value_heads": 4,
18
+ "rms_norm_eps": 1e-06,
19
+ "rope_theta": 1000000.0,
20
+ "sliding_window": 131072,
21
+ "tie_word_embeddings": false,
22
+ "torch_dtype": "bfloat16",
23
+ "transformers_version": "4.43.1",
24
+ "use_cache": true,
25
+ "use_sliding_window": false,
26
+ "vocab_size": 152064
27
+ }
model-00001-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e42647305c83d2d293c16cab3f15cecffcc035fdd136c176a9d73b7bd5582646
3
+ size 989775928
model-00002-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1467393987e0c0f41f3fd168bcd96e179b1c3dc9cbf0d730c298502fd287984f
3
+ size 1055018488
model-00003-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d164f06156d431746f6c3da9ce1f94ab815096135e6402f81cd83dd1dc165134
3
+ size 972830240
model-00004-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:88d4a5543b4ba0a7617b6d66b87b37fb1d732caf8e2d0fc3f283f8fba116adc7
3
+ size 989790536
model-00005-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd299bf29896060826913c0d06b432dcde100c939ba50ecdabf0458d9598c5bb
3
+ size 1071112928
model-00006-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fd2e2f08a9d35d4ef3ffa9d6d0c48930f892b4ddcd44da714f81873f6e0de58
3
+ size 956735808
model-00007-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e26dc52a915335dc7a3eb43a05d30863402ef00c763bc196b286e75050f9b89f
3
+ size 1055018528
model-00008-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cdba5a93337d7adc91aea2c025f7ad3b9a165980460581c4e83c38868198a34e
3
+ size 972830288
model-00009-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59f9a372c49424f1d001406703a03020244adc8cfb5383ed4347d2fb5462c494
3
+ size 989790600
model-00010-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a2623fcaaead92ca5413d7be214debce744dfd2fff06d274f769de48a5eff144
3
+ size 1071112976
model-00011-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a1ec57b0f2219e100e0a3f10f56948a5d5f08262e77b3772ee5398b7ad653c4
3
+ size 956735872
model-00012-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29663f1d2c21cd6e6851e2679f4cea3636094f7cd622fcdf5a4d2c9c113ffa49
3
+ size 1055018544
model-00013-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:06ef33f42a11f91439a1fd3df6f2cf6ce982cc77259cd6da31038b68dc65d856
3
+ size 972830288
model-00014-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f6c99af5adbf2b2b51acde06be86721d406bb5b437e5b5288079e2f697b3e71a
3
+ size 989790600
model-00015-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2aab798c888104790ee1bbf0956b1048081199d2c3d8710237eb3a46c87d89d2
3
+ size 1071112976
model-00016-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:481c4874c429434fe299bbedb652115cc166af2023a999e05e143b96de848367
3
+ size 956735872
model-00017-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:08ecedb8d54b361e4c1e8ada314d6bcbc53376211b9cfbc00de93db3f42408f9
3
+ size 1055018544
model-00018-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a7b50431ec4796d7eb1dbc6b9b9fc6d63f6eb0eff8ac95e5d0e87fed163acbf
3
+ size 477949592
model-00019-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b9e833650aa9a7768f52e6c4872f89f930bb9919b096f80f91049798f9f82a4
3
+ size 1089994880
model-00020-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aeb96bb95d2e7942ebb166fa60b3971b131143e2445c6dbafee25c0f16a6f438
3
+ size 7272
model-00021-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:096c0e03e04eba60fa04162ecb29fa8635c81215d8be9267ea2612558ff4bd94
3
+ size 1089994864
model-00022-of-00022.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9d8a3b71111916dba940f6418eab9c0d0b50d905b297f08a5c821891fa3ad320
3
+ size 932072376
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
model_extra_tensors.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b7db23f4437b320d798637e324a6d19e2f799b79e192b037407be8d4a20af87a
3
+ size 335311196
preprocessor_config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_convert_rgb": true,
3
+ "do_normalize": true,
4
+ "do_rescale": true,
5
+ "do_resize": true,
6
+ "image_mean": [
7
+ 0.5,
8
+ 0.5,
9
+ 0.5
10
+ ],
11
+ "image_processor_type": "SiglipImageProcessor",
12
+ "image_std": [
13
+ 0.5,
14
+ 0.5,
15
+ 0.5
16
+ ],
17
+ "processor_class": "BagelProcessor",
18
+ "rescale_factor": 0.00392156862745098,
19
+ "resample": 3,
20
+ "size": {
21
+ "height": 980,
22
+ "width": 980
23
+ }
24
+ }
quantization_config.json ADDED
@@ -0,0 +1,1265 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bits": 4,
3
+ "data_type": "int",
4
+ "group_size": 128,
5
+ "sym": true,
6
+ "transform_config": {},
7
+ "autoround_version": "0.12.0",
8
+ "block_name_to_quantize": "language_model.model.layers",
9
+ "quant_method": "auto-round",
10
+ "packing_format": "auto_round:auto_gptq",
11
+ "extra_config": {
12
+ "language_model.model.layers.0.self_attn.q_proj_moe_gen": {
13
+ "bits": 16,
14
+ "data_type": "float"
15
+ },
16
+ "language_model.model.layers.0.self_attn.k_proj_moe_gen": {
17
+ "bits": 16,
18
+ "data_type": "float"
19
+ },
20
+ "language_model.model.layers.0.self_attn.v_proj_moe_gen": {
21
+ "bits": 16,
22
+ "data_type": "float"
23
+ },
24
+ "language_model.model.layers.0.self_attn.o_proj_moe_gen": {
25
+ "bits": 16,
26
+ "data_type": "float"
27
+ },
28
+ "language_model.model.layers.0.mlp_moe_gen.gate_proj": {
29
+ "bits": 16,
30
+ "data_type": "float"
31
+ },
32
+ "language_model.model.layers.0.mlp_moe_gen.up_proj": {
33
+ "bits": 16,
34
+ "data_type": "float"
35
+ },
36
+ "language_model.model.layers.0.mlp_moe_gen.down_proj": {
37
+ "bits": 16,
38
+ "data_type": "float"
39
+ },
40
+ "language_model.model.layers.1.self_attn.q_proj_moe_gen": {
41
+ "bits": 16,
42
+ "data_type": "float"
43
+ },
44
+ "language_model.model.layers.1.self_attn.k_proj_moe_gen": {
45
+ "bits": 16,
46
+ "data_type": "float"
47
+ },
48
+ "language_model.model.layers.1.self_attn.v_proj_moe_gen": {
49
+ "bits": 16,
50
+ "data_type": "float"
51
+ },
52
+ "language_model.model.layers.1.self_attn.o_proj_moe_gen": {
53
+ "bits": 16,
54
+ "data_type": "float"
55
+ },
56
+ "language_model.model.layers.1.mlp_moe_gen.gate_proj": {
57
+ "bits": 16,
58
+ "data_type": "float"
59
+ },
60
+ "language_model.model.layers.1.mlp_moe_gen.up_proj": {
61
+ "bits": 16,
62
+ "data_type": "float"
63
+ },
64
+ "language_model.model.layers.1.mlp_moe_gen.down_proj": {
65
+ "bits": 16,
66
+ "data_type": "float"
67
+ },
68
+ "language_model.model.layers.2.self_attn.q_proj_moe_gen": {
69
+ "bits": 16,
70
+ "data_type": "float"
71
+ },
72
+ "language_model.model.layers.2.self_attn.k_proj_moe_gen": {
73
+ "bits": 16,
74
+ "data_type": "float"
75
+ },
76
+ "language_model.model.layers.2.self_attn.v_proj_moe_gen": {
77
+ "bits": 16,
78
+ "data_type": "float"
79
+ },
80
+ "language_model.model.layers.2.self_attn.o_proj_moe_gen": {
81
+ "bits": 16,
82
+ "data_type": "float"
83
+ },
84
+ "language_model.model.layers.2.mlp_moe_gen.gate_proj": {
85
+ "bits": 16,
86
+ "data_type": "float"
87
+ },
88
+ "language_model.model.layers.2.mlp_moe_gen.up_proj": {
89
+ "bits": 16,
90
+ "data_type": "float"
91
+ },
92
+ "language_model.model.layers.2.mlp_moe_gen.down_proj": {
93
+ "bits": 16,
94
+ "data_type": "float"
95
+ },
96
+ "language_model.model.layers.3.self_attn.q_proj_moe_gen": {
97
+ "bits": 16,
98
+ "data_type": "float"
99
+ },
100
+ "language_model.model.layers.3.self_attn.k_proj_moe_gen": {
101
+ "bits": 16,
102
+ "data_type": "float"
103
+ },
104
+ "language_model.model.layers.3.self_attn.v_proj_moe_gen": {
105
+ "bits": 16,
106
+ "data_type": "float"
107
+ },
108
+ "language_model.model.layers.3.self_attn.o_proj_moe_gen": {
109
+ "bits": 16,
110
+ "data_type": "float"
111
+ },
112
+ "language_model.model.layers.3.mlp_moe_gen.gate_proj": {
113
+ "bits": 16,
114
+ "data_type": "float"
115
+ },
116
+ "language_model.model.layers.3.mlp_moe_gen.up_proj": {
117
+ "bits": 16,
118
+ "data_type": "float"
119
+ },
120
+ "language_model.model.layers.3.mlp_moe_gen.down_proj": {
121
+ "bits": 16,
122
+ "data_type": "float"
123
+ },
124
+ "language_model.model.layers.4.self_attn.q_proj_moe_gen": {
125
+ "bits": 16,
126
+ "data_type": "float"
127
+ },
128
+ "language_model.model.layers.4.self_attn.k_proj_moe_gen": {
129
+ "bits": 16,
130
+ "data_type": "float"
131
+ },
132
+ "language_model.model.layers.4.self_attn.v_proj_moe_gen": {
133
+ "bits": 16,
134
+ "data_type": "float"
135
+ },
136
+ "language_model.model.layers.4.self_attn.o_proj_moe_gen": {
137
+ "bits": 16,
138
+ "data_type": "float"
139
+ },
140
+ "language_model.model.layers.4.mlp_moe_gen.gate_proj": {
141
+ "bits": 16,
142
+ "data_type": "float"
143
+ },
144
+ "language_model.model.layers.4.mlp_moe_gen.up_proj": {
145
+ "bits": 16,
146
+ "data_type": "float"
147
+ },
148
+ "language_model.model.layers.4.mlp_moe_gen.down_proj": {
149
+ "bits": 16,
150
+ "data_type": "float"
151
+ },
152
+ "language_model.model.layers.5.self_attn.q_proj_moe_gen": {
153
+ "bits": 16,
154
+ "data_type": "float"
155
+ },
156
+ "language_model.model.layers.5.self_attn.k_proj_moe_gen": {
157
+ "bits": 16,
158
+ "data_type": "float"
159
+ },
160
+ "language_model.model.layers.5.self_attn.v_proj_moe_gen": {
161
+ "bits": 16,
162
+ "data_type": "float"
163
+ },
164
+ "language_model.model.layers.5.self_attn.o_proj_moe_gen": {
165
+ "bits": 16,
166
+ "data_type": "float"
167
+ },
168
+ "language_model.model.layers.5.mlp_moe_gen.gate_proj": {
169
+ "bits": 16,
170
+ "data_type": "float"
171
+ },
172
+ "language_model.model.layers.5.mlp_moe_gen.up_proj": {
173
+ "bits": 16,
174
+ "data_type": "float"
175
+ },
176
+ "language_model.model.layers.5.mlp_moe_gen.down_proj": {
177
+ "bits": 16,
178
+ "data_type": "float"
179
+ },
180
+ "language_model.model.layers.6.self_attn.q_proj_moe_gen": {
181
+ "bits": 16,
182
+ "data_type": "float"
183
+ },
184
+ "language_model.model.layers.6.self_attn.k_proj_moe_gen": {
185
+ "bits": 16,
186
+ "data_type": "float"
187
+ },
188
+ "language_model.model.layers.6.self_attn.v_proj_moe_gen": {
189
+ "bits": 16,
190
+ "data_type": "float"
191
+ },
192
+ "language_model.model.layers.6.self_attn.o_proj_moe_gen": {
193
+ "bits": 16,
194
+ "data_type": "float"
195
+ },
196
+ "language_model.model.layers.6.mlp_moe_gen.gate_proj": {
197
+ "bits": 16,
198
+ "data_type": "float"
199
+ },
200
+ "language_model.model.layers.6.mlp_moe_gen.up_proj": {
201
+ "bits": 16,
202
+ "data_type": "float"
203
+ },
204
+ "language_model.model.layers.6.mlp_moe_gen.down_proj": {
205
+ "bits": 16,
206
+ "data_type": "float"
207
+ },
208
+ "language_model.model.layers.7.self_attn.q_proj_moe_gen": {
209
+ "bits": 16,
210
+ "data_type": "float"
211
+ },
212
+ "language_model.model.layers.7.self_attn.k_proj_moe_gen": {
213
+ "bits": 16,
214
+ "data_type": "float"
215
+ },
216
+ "language_model.model.layers.7.self_attn.v_proj_moe_gen": {
217
+ "bits": 16,
218
+ "data_type": "float"
219
+ },
220
+ "language_model.model.layers.7.self_attn.o_proj_moe_gen": {
221
+ "bits": 16,
222
+ "data_type": "float"
223
+ },
224
+ "language_model.model.layers.7.mlp_moe_gen.gate_proj": {
225
+ "bits": 16,
226
+ "data_type": "float"
227
+ },
228
+ "language_model.model.layers.7.mlp_moe_gen.up_proj": {
229
+ "bits": 16,
230
+ "data_type": "float"
231
+ },
232
+ "language_model.model.layers.7.mlp_moe_gen.down_proj": {
233
+ "bits": 16,
234
+ "data_type": "float"
235
+ },
236
+ "language_model.model.layers.8.self_attn.q_proj_moe_gen": {
237
+ "bits": 16,
238
+ "data_type": "float"
239
+ },
240
+ "language_model.model.layers.8.self_attn.k_proj_moe_gen": {
241
+ "bits": 16,
242
+ "data_type": "float"
243
+ },
244
+ "language_model.model.layers.8.self_attn.v_proj_moe_gen": {
245
+ "bits": 16,
246
+ "data_type": "float"
247
+ },
248
+ "language_model.model.layers.8.self_attn.o_proj_moe_gen": {
249
+ "bits": 16,
250
+ "data_type": "float"
251
+ },
252
+ "language_model.model.layers.8.mlp_moe_gen.gate_proj": {
253
+ "bits": 16,
254
+ "data_type": "float"
255
+ },
256
+ "language_model.model.layers.8.mlp_moe_gen.up_proj": {
257
+ "bits": 16,
258
+ "data_type": "float"
259
+ },
260
+ "language_model.model.layers.8.mlp_moe_gen.down_proj": {
261
+ "bits": 16,
262
+ "data_type": "float"
263
+ },
264
+ "language_model.model.layers.9.self_attn.q_proj_moe_gen": {
265
+ "bits": 16,
266
+ "data_type": "float"
267
+ },
268
+ "language_model.model.layers.9.self_attn.k_proj_moe_gen": {
269
+ "bits": 16,
270
+ "data_type": "float"
271
+ },
272
+ "language_model.model.layers.9.self_attn.v_proj_moe_gen": {
273
+ "bits": 16,
274
+ "data_type": "float"
275
+ },
276
+ "language_model.model.layers.9.self_attn.o_proj_moe_gen": {
277
+ "bits": 16,
278
+ "data_type": "float"
279
+ },
280
+ "language_model.model.layers.9.mlp_moe_gen.gate_proj": {
281
+ "bits": 16,
282
+ "data_type": "float"
283
+ },
284
+ "language_model.model.layers.9.mlp_moe_gen.up_proj": {
285
+ "bits": 16,
286
+ "data_type": "float"
287
+ },
288
+ "language_model.model.layers.9.mlp_moe_gen.down_proj": {
289
+ "bits": 16,
290
+ "data_type": "float"
291
+ },
292
+ "language_model.model.layers.10.self_attn.q_proj_moe_gen": {
293
+ "bits": 16,
294
+ "data_type": "float"
295
+ },
296
+ "language_model.model.layers.10.self_attn.k_proj_moe_gen": {
297
+ "bits": 16,
298
+ "data_type": "float"
299
+ },
300
+ "language_model.model.layers.10.self_attn.v_proj_moe_gen": {
301
+ "bits": 16,
302
+ "data_type": "float"
303
+ },
304
+ "language_model.model.layers.10.self_attn.o_proj_moe_gen": {
305
+ "bits": 16,
306
+ "data_type": "float"
307
+ },
308
+ "language_model.model.layers.10.mlp_moe_gen.gate_proj": {
309
+ "bits": 16,
310
+ "data_type": "float"
311
+ },
312
+ "language_model.model.layers.10.mlp_moe_gen.up_proj": {
313
+ "bits": 16,
314
+ "data_type": "float"
315
+ },
316
+ "language_model.model.layers.10.mlp_moe_gen.down_proj": {
317
+ "bits": 16,
318
+ "data_type": "float"
319
+ },
320
+ "language_model.model.layers.11.self_attn.q_proj_moe_gen": {
321
+ "bits": 16,
322
+ "data_type": "float"
323
+ },
324
+ "language_model.model.layers.11.self_attn.k_proj_moe_gen": {
325
+ "bits": 16,
326
+ "data_type": "float"
327
+ },
328
+ "language_model.model.layers.11.self_attn.v_proj_moe_gen": {
329
+ "bits": 16,
330
+ "data_type": "float"
331
+ },
332
+ "language_model.model.layers.11.self_attn.o_proj_moe_gen": {
333
+ "bits": 16,
334
+ "data_type": "float"
335
+ },
336
+ "language_model.model.layers.11.mlp_moe_gen.gate_proj": {
337
+ "bits": 16,
338
+ "data_type": "float"
339
+ },
340
+ "language_model.model.layers.11.mlp_moe_gen.up_proj": {
341
+ "bits": 16,
342
+ "data_type": "float"
343
+ },
344
+ "language_model.model.layers.11.mlp_moe_gen.down_proj": {
345
+ "bits": 16,
346
+ "data_type": "float"
347
+ },
348
+ "language_model.model.layers.12.self_attn.q_proj_moe_gen": {
349
+ "bits": 16,
350
+ "data_type": "float"
351
+ },
352
+ "language_model.model.layers.12.self_attn.k_proj_moe_gen": {
353
+ "bits": 16,
354
+ "data_type": "float"
355
+ },
356
+ "language_model.model.layers.12.self_attn.v_proj_moe_gen": {
357
+ "bits": 16,
358
+ "data_type": "float"
359
+ },
360
+ "language_model.model.layers.12.self_attn.o_proj_moe_gen": {
361
+ "bits": 16,
362
+ "data_type": "float"
363
+ },
364
+ "language_model.model.layers.12.mlp_moe_gen.gate_proj": {
365
+ "bits": 16,
366
+ "data_type": "float"
367
+ },
368
+ "language_model.model.layers.12.mlp_moe_gen.up_proj": {
369
+ "bits": 16,
370
+ "data_type": "float"
371
+ },
372
+ "language_model.model.layers.12.mlp_moe_gen.down_proj": {
373
+ "bits": 16,
374
+ "data_type": "float"
375
+ },
376
+ "language_model.model.layers.13.self_attn.q_proj_moe_gen": {
377
+ "bits": 16,
378
+ "data_type": "float"
379
+ },
380
+ "language_model.model.layers.13.self_attn.k_proj_moe_gen": {
381
+ "bits": 16,
382
+ "data_type": "float"
383
+ },
384
+ "language_model.model.layers.13.self_attn.v_proj_moe_gen": {
385
+ "bits": 16,
386
+ "data_type": "float"
387
+ },
388
+ "language_model.model.layers.13.self_attn.o_proj_moe_gen": {
389
+ "bits": 16,
390
+ "data_type": "float"
391
+ },
392
+ "language_model.model.layers.13.mlp_moe_gen.gate_proj": {
393
+ "bits": 16,
394
+ "data_type": "float"
395
+ },
396
+ "language_model.model.layers.13.mlp_moe_gen.up_proj": {
397
+ "bits": 16,
398
+ "data_type": "float"
399
+ },
400
+ "language_model.model.layers.13.mlp_moe_gen.down_proj": {
401
+ "bits": 16,
402
+ "data_type": "float"
403
+ },
404
+ "language_model.model.layers.14.self_attn.q_proj_moe_gen": {
405
+ "bits": 16,
406
+ "data_type": "float"
407
+ },
408
+ "language_model.model.layers.14.self_attn.k_proj_moe_gen": {
409
+ "bits": 16,
410
+ "data_type": "float"
411
+ },
412
+ "language_model.model.layers.14.self_attn.v_proj_moe_gen": {
413
+ "bits": 16,
414
+ "data_type": "float"
415
+ },
416
+ "language_model.model.layers.14.self_attn.o_proj_moe_gen": {
417
+ "bits": 16,
418
+ "data_type": "float"
419
+ },
420
+ "language_model.model.layers.14.mlp_moe_gen.gate_proj": {
421
+ "bits": 16,
422
+ "data_type": "float"
423
+ },
424
+ "language_model.model.layers.14.mlp_moe_gen.up_proj": {
425
+ "bits": 16,
426
+ "data_type": "float"
427
+ },
428
+ "language_model.model.layers.14.mlp_moe_gen.down_proj": {
429
+ "bits": 16,
430
+ "data_type": "float"
431
+ },
432
+ "language_model.model.layers.15.self_attn.q_proj_moe_gen": {
433
+ "bits": 16,
434
+ "data_type": "float"
435
+ },
436
+ "language_model.model.layers.15.self_attn.k_proj_moe_gen": {
437
+ "bits": 16,
438
+ "data_type": "float"
439
+ },
440
+ "language_model.model.layers.15.self_attn.v_proj_moe_gen": {
441
+ "bits": 16,
442
+ "data_type": "float"
443
+ },
444
+ "language_model.model.layers.15.self_attn.o_proj_moe_gen": {
445
+ "bits": 16,
446
+ "data_type": "float"
447
+ },
448
+ "language_model.model.layers.15.mlp_moe_gen.gate_proj": {
449
+ "bits": 16,
450
+ "data_type": "float"
451
+ },
452
+ "language_model.model.layers.15.mlp_moe_gen.up_proj": {
453
+ "bits": 16,
454
+ "data_type": "float"
455
+ },
456
+ "language_model.model.layers.15.mlp_moe_gen.down_proj": {
457
+ "bits": 16,
458
+ "data_type": "float"
459
+ },
460
+ "language_model.model.layers.16.self_attn.q_proj_moe_gen": {
461
+ "bits": 16,
462
+ "data_type": "float"
463
+ },
464
+ "language_model.model.layers.16.self_attn.k_proj_moe_gen": {
465
+ "bits": 16,
466
+ "data_type": "float"
467
+ },
468
+ "language_model.model.layers.16.self_attn.v_proj_moe_gen": {
469
+ "bits": 16,
470
+ "data_type": "float"
471
+ },
472
+ "language_model.model.layers.16.self_attn.o_proj_moe_gen": {
473
+ "bits": 16,
474
+ "data_type": "float"
475
+ },
476
+ "language_model.model.layers.16.mlp_moe_gen.gate_proj": {
477
+ "bits": 16,
478
+ "data_type": "float"
479
+ },
480
+ "language_model.model.layers.16.mlp_moe_gen.up_proj": {
481
+ "bits": 16,
482
+ "data_type": "float"
483
+ },
484
+ "language_model.model.layers.16.mlp_moe_gen.down_proj": {
485
+ "bits": 16,
486
+ "data_type": "float"
487
+ },
488
+ "language_model.model.layers.17.self_attn.q_proj_moe_gen": {
489
+ "bits": 16,
490
+ "data_type": "float"
491
+ },
492
+ "language_model.model.layers.17.self_attn.k_proj_moe_gen": {
493
+ "bits": 16,
494
+ "data_type": "float"
495
+ },
496
+ "language_model.model.layers.17.self_attn.v_proj_moe_gen": {
497
+ "bits": 16,
498
+ "data_type": "float"
499
+ },
500
+ "language_model.model.layers.17.self_attn.o_proj_moe_gen": {
501
+ "bits": 16,
502
+ "data_type": "float"
503
+ },
504
+ "language_model.model.layers.17.mlp_moe_gen.gate_proj": {
505
+ "bits": 16,
506
+ "data_type": "float"
507
+ },
508
+ "language_model.model.layers.17.mlp_moe_gen.up_proj": {
509
+ "bits": 16,
510
+ "data_type": "float"
511
+ },
512
+ "language_model.model.layers.17.mlp_moe_gen.down_proj": {
513
+ "bits": 16,
514
+ "data_type": "float"
515
+ },
516
+ "language_model.model.layers.18.self_attn.q_proj_moe_gen": {
517
+ "bits": 16,
518
+ "data_type": "float"
519
+ },
520
+ "language_model.model.layers.18.self_attn.k_proj_moe_gen": {
521
+ "bits": 16,
522
+ "data_type": "float"
523
+ },
524
+ "language_model.model.layers.18.self_attn.v_proj_moe_gen": {
525
+ "bits": 16,
526
+ "data_type": "float"
527
+ },
528
+ "language_model.model.layers.18.self_attn.o_proj_moe_gen": {
529
+ "bits": 16,
530
+ "data_type": "float"
531
+ },
532
+ "language_model.model.layers.18.mlp_moe_gen.gate_proj": {
533
+ "bits": 16,
534
+ "data_type": "float"
535
+ },
536
+ "language_model.model.layers.18.mlp_moe_gen.up_proj": {
537
+ "bits": 16,
538
+ "data_type": "float"
539
+ },
540
+ "language_model.model.layers.18.mlp_moe_gen.down_proj": {
541
+ "bits": 16,
542
+ "data_type": "float"
543
+ },
544
+ "language_model.model.layers.19.self_attn.q_proj_moe_gen": {
545
+ "bits": 16,
546
+ "data_type": "float"
547
+ },
548
+ "language_model.model.layers.19.self_attn.k_proj_moe_gen": {
549
+ "bits": 16,
550
+ "data_type": "float"
551
+ },
552
+ "language_model.model.layers.19.self_attn.v_proj_moe_gen": {
553
+ "bits": 16,
554
+ "data_type": "float"
555
+ },
556
+ "language_model.model.layers.19.self_attn.o_proj_moe_gen": {
557
+ "bits": 16,
558
+ "data_type": "float"
559
+ },
560
+ "language_model.model.layers.19.mlp_moe_gen.gate_proj": {
561
+ "bits": 16,
562
+ "data_type": "float"
563
+ },
564
+ "language_model.model.layers.19.mlp_moe_gen.up_proj": {
565
+ "bits": 16,
566
+ "data_type": "float"
567
+ },
568
+ "language_model.model.layers.19.mlp_moe_gen.down_proj": {
569
+ "bits": 16,
570
+ "data_type": "float"
571
+ },
572
+ "language_model.model.layers.20.self_attn.q_proj_moe_gen": {
573
+ "bits": 16,
574
+ "data_type": "float"
575
+ },
576
+ "language_model.model.layers.20.self_attn.k_proj_moe_gen": {
577
+ "bits": 16,
578
+ "data_type": "float"
579
+ },
580
+ "language_model.model.layers.20.self_attn.v_proj_moe_gen": {
581
+ "bits": 16,
582
+ "data_type": "float"
583
+ },
584
+ "language_model.model.layers.20.self_attn.o_proj_moe_gen": {
585
+ "bits": 16,
586
+ "data_type": "float"
587
+ },
588
+ "language_model.model.layers.20.mlp_moe_gen.gate_proj": {
589
+ "bits": 16,
590
+ "data_type": "float"
591
+ },
592
+ "language_model.model.layers.20.mlp_moe_gen.up_proj": {
593
+ "bits": 16,
594
+ "data_type": "float"
595
+ },
596
+ "language_model.model.layers.20.mlp_moe_gen.down_proj": {
597
+ "bits": 16,
598
+ "data_type": "float"
599
+ },
600
+ "language_model.model.layers.21.self_attn.q_proj_moe_gen": {
601
+ "bits": 16,
602
+ "data_type": "float"
603
+ },
604
+ "language_model.model.layers.21.self_attn.k_proj_moe_gen": {
605
+ "bits": 16,
606
+ "data_type": "float"
607
+ },
608
+ "language_model.model.layers.21.self_attn.v_proj_moe_gen": {
609
+ "bits": 16,
610
+ "data_type": "float"
611
+ },
612
+ "language_model.model.layers.21.self_attn.o_proj_moe_gen": {
613
+ "bits": 16,
614
+ "data_type": "float"
615
+ },
616
+ "language_model.model.layers.21.mlp_moe_gen.gate_proj": {
617
+ "bits": 16,
618
+ "data_type": "float"
619
+ },
620
+ "language_model.model.layers.21.mlp_moe_gen.up_proj": {
621
+ "bits": 16,
622
+ "data_type": "float"
623
+ },
624
+ "language_model.model.layers.21.mlp_moe_gen.down_proj": {
625
+ "bits": 16,
626
+ "data_type": "float"
627
+ },
628
+ "language_model.model.layers.22.self_attn.q_proj_moe_gen": {
629
+ "bits": 16,
630
+ "data_type": "float"
631
+ },
632
+ "language_model.model.layers.22.self_attn.k_proj_moe_gen": {
633
+ "bits": 16,
634
+ "data_type": "float"
635
+ },
636
+ "language_model.model.layers.22.self_attn.v_proj_moe_gen": {
637
+ "bits": 16,
638
+ "data_type": "float"
639
+ },
640
+ "language_model.model.layers.22.self_attn.o_proj_moe_gen": {
641
+ "bits": 16,
642
+ "data_type": "float"
643
+ },
644
+ "language_model.model.layers.22.mlp_moe_gen.gate_proj": {
645
+ "bits": 16,
646
+ "data_type": "float"
647
+ },
648
+ "language_model.model.layers.22.mlp_moe_gen.up_proj": {
649
+ "bits": 16,
650
+ "data_type": "float"
651
+ },
652
+ "language_model.model.layers.22.mlp_moe_gen.down_proj": {
653
+ "bits": 16,
654
+ "data_type": "float"
655
+ },
656
+ "language_model.model.layers.23.self_attn.q_proj_moe_gen": {
657
+ "bits": 16,
658
+ "data_type": "float"
659
+ },
660
+ "language_model.model.layers.23.self_attn.k_proj_moe_gen": {
661
+ "bits": 16,
662
+ "data_type": "float"
663
+ },
664
+ "language_model.model.layers.23.self_attn.v_proj_moe_gen": {
665
+ "bits": 16,
666
+ "data_type": "float"
667
+ },
668
+ "language_model.model.layers.23.self_attn.o_proj_moe_gen": {
669
+ "bits": 16,
670
+ "data_type": "float"
671
+ },
672
+ "language_model.model.layers.23.mlp_moe_gen.gate_proj": {
673
+ "bits": 16,
674
+ "data_type": "float"
675
+ },
676
+ "language_model.model.layers.23.mlp_moe_gen.up_proj": {
677
+ "bits": 16,
678
+ "data_type": "float"
679
+ },
680
+ "language_model.model.layers.23.mlp_moe_gen.down_proj": {
681
+ "bits": 16,
682
+ "data_type": "float"
683
+ },
684
+ "language_model.model.layers.24.self_attn.q_proj_moe_gen": {
685
+ "bits": 16,
686
+ "data_type": "float"
687
+ },
688
+ "language_model.model.layers.24.self_attn.k_proj_moe_gen": {
689
+ "bits": 16,
690
+ "data_type": "float"
691
+ },
692
+ "language_model.model.layers.24.self_attn.v_proj_moe_gen": {
693
+ "bits": 16,
694
+ "data_type": "float"
695
+ },
696
+ "language_model.model.layers.24.self_attn.o_proj_moe_gen": {
697
+ "bits": 16,
698
+ "data_type": "float"
699
+ },
700
+ "language_model.model.layers.24.mlp_moe_gen.gate_proj": {
701
+ "bits": 16,
702
+ "data_type": "float"
703
+ },
704
+ "language_model.model.layers.24.mlp_moe_gen.up_proj": {
705
+ "bits": 16,
706
+ "data_type": "float"
707
+ },
708
+ "language_model.model.layers.24.mlp_moe_gen.down_proj": {
709
+ "bits": 16,
710
+ "data_type": "float"
711
+ },
712
+ "language_model.model.layers.25.self_attn.q_proj_moe_gen": {
713
+ "bits": 16,
714
+ "data_type": "float"
715
+ },
716
+ "language_model.model.layers.25.self_attn.k_proj_moe_gen": {
717
+ "bits": 16,
718
+ "data_type": "float"
719
+ },
720
+ "language_model.model.layers.25.self_attn.v_proj_moe_gen": {
721
+ "bits": 16,
722
+ "data_type": "float"
723
+ },
724
+ "language_model.model.layers.25.self_attn.o_proj_moe_gen": {
725
+ "bits": 16,
726
+ "data_type": "float"
727
+ },
728
+ "language_model.model.layers.25.mlp_moe_gen.gate_proj": {
729
+ "bits": 16,
730
+ "data_type": "float"
731
+ },
732
+ "language_model.model.layers.25.mlp_moe_gen.up_proj": {
733
+ "bits": 16,
734
+ "data_type": "float"
735
+ },
736
+ "language_model.model.layers.25.mlp_moe_gen.down_proj": {
737
+ "bits": 16,
738
+ "data_type": "float"
739
+ },
740
+ "language_model.model.layers.26.self_attn.q_proj_moe_gen": {
741
+ "bits": 16,
742
+ "data_type": "float"
743
+ },
744
+ "language_model.model.layers.26.self_attn.k_proj_moe_gen": {
745
+ "bits": 16,
746
+ "data_type": "float"
747
+ },
748
+ "language_model.model.layers.26.self_attn.v_proj_moe_gen": {
749
+ "bits": 16,
750
+ "data_type": "float"
751
+ },
752
+ "language_model.model.layers.26.self_attn.o_proj_moe_gen": {
753
+ "bits": 16,
754
+ "data_type": "float"
755
+ },
756
+ "language_model.model.layers.26.mlp_moe_gen.gate_proj": {
757
+ "bits": 16,
758
+ "data_type": "float"
759
+ },
760
+ "language_model.model.layers.26.mlp_moe_gen.up_proj": {
761
+ "bits": 16,
762
+ "data_type": "float"
763
+ },
764
+ "language_model.model.layers.26.mlp_moe_gen.down_proj": {
765
+ "bits": 16,
766
+ "data_type": "float"
767
+ },
768
+ "language_model.model.layers.27.self_attn.q_proj_moe_gen": {
769
+ "bits": 16,
770
+ "data_type": "float"
771
+ },
772
+ "language_model.model.layers.27.self_attn.k_proj_moe_gen": {
773
+ "bits": 16,
774
+ "data_type": "float"
775
+ },
776
+ "language_model.model.layers.27.self_attn.v_proj_moe_gen": {
777
+ "bits": 16,
778
+ "data_type": "float"
779
+ },
780
+ "language_model.model.layers.27.self_attn.o_proj_moe_gen": {
781
+ "bits": 16,
782
+ "data_type": "float"
783
+ },
784
+ "language_model.model.layers.27.mlp_moe_gen.gate_proj": {
785
+ "bits": 16,
786
+ "data_type": "float"
787
+ },
788
+ "language_model.model.layers.27.mlp_moe_gen.up_proj": {
789
+ "bits": 16,
790
+ "data_type": "float"
791
+ },
792
+ "language_model.model.layers.27.mlp_moe_gen.down_proj": {
793
+ "bits": 16,
794
+ "data_type": "float"
795
+ },
796
+ "language_model.model.layers.0.self_attn.q_proj": {
797
+ "bits": 16,
798
+ "data_type": "float"
799
+ },
800
+ "language_model.model.layers.1.self_attn.q_proj": {
801
+ "bits": 16,
802
+ "data_type": "float"
803
+ },
804
+ "language_model.model.layers.2.self_attn.q_proj": {
805
+ "bits": 16,
806
+ "data_type": "float"
807
+ },
808
+ "language_model.model.layers.3.self_attn.q_proj": {
809
+ "bits": 16,
810
+ "data_type": "float"
811
+ },
812
+ "language_model.model.layers.4.self_attn.q_proj": {
813
+ "bits": 16,
814
+ "data_type": "float"
815
+ },
816
+ "language_model.model.layers.5.self_attn.q_proj": {
817
+ "bits": 16,
818
+ "data_type": "float"
819
+ },
820
+ "language_model.model.layers.6.self_attn.q_proj": {
821
+ "bits": 16,
822
+ "data_type": "float"
823
+ },
824
+ "language_model.model.layers.7.self_attn.q_proj": {
825
+ "bits": 16,
826
+ "data_type": "float"
827
+ },
828
+ "language_model.model.layers.8.self_attn.q_proj": {
829
+ "bits": 16,
830
+ "data_type": "float"
831
+ },
832
+ "language_model.model.layers.9.self_attn.q_proj": {
833
+ "bits": 16,
834
+ "data_type": "float"
835
+ },
836
+ "language_model.model.layers.10.self_attn.q_proj": {
837
+ "bits": 16,
838
+ "data_type": "float"
839
+ },
840
+ "language_model.model.layers.11.self_attn.q_proj": {
841
+ "bits": 16,
842
+ "data_type": "float"
843
+ },
844
+ "language_model.model.layers.12.self_attn.q_proj": {
845
+ "bits": 16,
846
+ "data_type": "float"
847
+ },
848
+ "language_model.model.layers.13.self_attn.q_proj": {
849
+ "bits": 16,
850
+ "data_type": "float"
851
+ },
852
+ "language_model.model.layers.14.self_attn.q_proj": {
853
+ "bits": 16,
854
+ "data_type": "float"
855
+ },
856
+ "language_model.model.layers.15.self_attn.q_proj": {
857
+ "bits": 16,
858
+ "data_type": "float"
859
+ },
860
+ "language_model.model.layers.16.self_attn.q_proj": {
861
+ "bits": 16,
862
+ "data_type": "float"
863
+ },
864
+ "language_model.model.layers.17.self_attn.q_proj": {
865
+ "bits": 16,
866
+ "data_type": "float"
867
+ },
868
+ "language_model.model.layers.18.self_attn.q_proj": {
869
+ "bits": 16,
870
+ "data_type": "float"
871
+ },
872
+ "language_model.model.layers.19.self_attn.q_proj": {
873
+ "bits": 16,
874
+ "data_type": "float"
875
+ },
876
+ "language_model.model.layers.20.self_attn.q_proj": {
877
+ "bits": 16,
878
+ "data_type": "float"
879
+ },
880
+ "language_model.model.layers.21.self_attn.q_proj": {
881
+ "bits": 16,
882
+ "data_type": "float"
883
+ },
884
+ "language_model.model.layers.22.self_attn.q_proj": {
885
+ "bits": 16,
886
+ "data_type": "float"
887
+ },
888
+ "language_model.model.layers.23.self_attn.q_proj": {
889
+ "bits": 16,
890
+ "data_type": "float"
891
+ },
892
+ "language_model.model.layers.24.self_attn.q_proj": {
893
+ "bits": 16,
894
+ "data_type": "float"
895
+ },
896
+ "language_model.model.layers.25.self_attn.q_proj": {
897
+ "bits": 16,
898
+ "data_type": "float"
899
+ },
900
+ "language_model.model.layers.26.self_attn.q_proj": {
901
+ "bits": 16,
902
+ "data_type": "float"
903
+ },
904
+ "language_model.model.layers.27.self_attn.q_proj": {
905
+ "bits": 16,
906
+ "data_type": "float"
907
+ },
908
+ "language_model.model.layers.0.self_attn.k_proj": {
909
+ "bits": 16,
910
+ "data_type": "float"
911
+ },
912
+ "language_model.model.layers.1.self_attn.k_proj": {
913
+ "bits": 16,
914
+ "data_type": "float"
915
+ },
916
+ "language_model.model.layers.2.self_attn.k_proj": {
917
+ "bits": 16,
918
+ "data_type": "float"
919
+ },
920
+ "language_model.model.layers.3.self_attn.k_proj": {
921
+ "bits": 16,
922
+ "data_type": "float"
923
+ },
924
+ "language_model.model.layers.4.self_attn.k_proj": {
925
+ "bits": 16,
926
+ "data_type": "float"
927
+ },
928
+ "language_model.model.layers.5.self_attn.k_proj": {
929
+ "bits": 16,
930
+ "data_type": "float"
931
+ },
932
+ "language_model.model.layers.6.self_attn.k_proj": {
933
+ "bits": 16,
934
+ "data_type": "float"
935
+ },
936
+ "language_model.model.layers.7.self_attn.k_proj": {
937
+ "bits": 16,
938
+ "data_type": "float"
939
+ },
940
+ "language_model.model.layers.8.self_attn.k_proj": {
941
+ "bits": 16,
942
+ "data_type": "float"
943
+ },
944
+ "language_model.model.layers.9.self_attn.k_proj": {
945
+ "bits": 16,
946
+ "data_type": "float"
947
+ },
948
+ "language_model.model.layers.10.self_attn.k_proj": {
949
+ "bits": 16,
950
+ "data_type": "float"
951
+ },
952
+ "language_model.model.layers.11.self_attn.k_proj": {
953
+ "bits": 16,
954
+ "data_type": "float"
955
+ },
956
+ "language_model.model.layers.12.self_attn.k_proj": {
957
+ "bits": 16,
958
+ "data_type": "float"
959
+ },
960
+ "language_model.model.layers.13.self_attn.k_proj": {
961
+ "bits": 16,
962
+ "data_type": "float"
963
+ },
964
+ "language_model.model.layers.14.self_attn.k_proj": {
965
+ "bits": 16,
966
+ "data_type": "float"
967
+ },
968
+ "language_model.model.layers.15.self_attn.k_proj": {
969
+ "bits": 16,
970
+ "data_type": "float"
971
+ },
972
+ "language_model.model.layers.16.self_attn.k_proj": {
973
+ "bits": 16,
974
+ "data_type": "float"
975
+ },
976
+ "language_model.model.layers.17.self_attn.k_proj": {
977
+ "bits": 16,
978
+ "data_type": "float"
979
+ },
980
+ "language_model.model.layers.18.self_attn.k_proj": {
981
+ "bits": 16,
982
+ "data_type": "float"
983
+ },
984
+ "language_model.model.layers.19.self_attn.k_proj": {
985
+ "bits": 16,
986
+ "data_type": "float"
987
+ },
988
+ "language_model.model.layers.20.self_attn.k_proj": {
989
+ "bits": 16,
990
+ "data_type": "float"
991
+ },
992
+ "language_model.model.layers.21.self_attn.k_proj": {
993
+ "bits": 16,
994
+ "data_type": "float"
995
+ },
996
+ "language_model.model.layers.22.self_attn.k_proj": {
997
+ "bits": 16,
998
+ "data_type": "float"
999
+ },
1000
+ "language_model.model.layers.23.self_attn.k_proj": {
1001
+ "bits": 16,
1002
+ "data_type": "float"
1003
+ },
1004
+ "language_model.model.layers.24.self_attn.k_proj": {
1005
+ "bits": 16,
1006
+ "data_type": "float"
1007
+ },
1008
+ "language_model.model.layers.25.self_attn.k_proj": {
1009
+ "bits": 16,
1010
+ "data_type": "float"
1011
+ },
1012
+ "language_model.model.layers.26.self_attn.k_proj": {
1013
+ "bits": 16,
1014
+ "data_type": "float"
1015
+ },
1016
+ "language_model.model.layers.27.self_attn.k_proj": {
1017
+ "bits": 16,
1018
+ "data_type": "float"
1019
+ },
1020
+ "language_model.model.layers.0.self_attn.v_proj": {
1021
+ "bits": 16,
1022
+ "data_type": "float"
1023
+ },
1024
+ "language_model.model.layers.1.self_attn.v_proj": {
1025
+ "bits": 16,
1026
+ "data_type": "float"
1027
+ },
1028
+ "language_model.model.layers.2.self_attn.v_proj": {
1029
+ "bits": 16,
1030
+ "data_type": "float"
1031
+ },
1032
+ "language_model.model.layers.3.self_attn.v_proj": {
1033
+ "bits": 16,
1034
+ "data_type": "float"
1035
+ },
1036
+ "language_model.model.layers.4.self_attn.v_proj": {
1037
+ "bits": 16,
1038
+ "data_type": "float"
1039
+ },
1040
+ "language_model.model.layers.5.self_attn.v_proj": {
1041
+ "bits": 16,
1042
+ "data_type": "float"
1043
+ },
1044
+ "language_model.model.layers.6.self_attn.v_proj": {
1045
+ "bits": 16,
1046
+ "data_type": "float"
1047
+ },
1048
+ "language_model.model.layers.7.self_attn.v_proj": {
1049
+ "bits": 16,
1050
+ "data_type": "float"
1051
+ },
1052
+ "language_model.model.layers.8.self_attn.v_proj": {
1053
+ "bits": 16,
1054
+ "data_type": "float"
1055
+ },
1056
+ "language_model.model.layers.9.self_attn.v_proj": {
1057
+ "bits": 16,
1058
+ "data_type": "float"
1059
+ },
1060
+ "language_model.model.layers.10.self_attn.v_proj": {
1061
+ "bits": 16,
1062
+ "data_type": "float"
1063
+ },
1064
+ "language_model.model.layers.11.self_attn.v_proj": {
1065
+ "bits": 16,
1066
+ "data_type": "float"
1067
+ },
1068
+ "language_model.model.layers.12.self_attn.v_proj": {
1069
+ "bits": 16,
1070
+ "data_type": "float"
1071
+ },
1072
+ "language_model.model.layers.13.self_attn.v_proj": {
1073
+ "bits": 16,
1074
+ "data_type": "float"
1075
+ },
1076
+ "language_model.model.layers.14.self_attn.v_proj": {
1077
+ "bits": 16,
1078
+ "data_type": "float"
1079
+ },
1080
+ "language_model.model.layers.15.self_attn.v_proj": {
1081
+ "bits": 16,
1082
+ "data_type": "float"
1083
+ },
1084
+ "language_model.model.layers.16.self_attn.v_proj": {
1085
+ "bits": 16,
1086
+ "data_type": "float"
1087
+ },
1088
+ "language_model.model.layers.17.self_attn.v_proj": {
1089
+ "bits": 16,
1090
+ "data_type": "float"
1091
+ },
1092
+ "language_model.model.layers.18.self_attn.v_proj": {
1093
+ "bits": 16,
1094
+ "data_type": "float"
1095
+ },
1096
+ "language_model.model.layers.19.self_attn.v_proj": {
1097
+ "bits": 16,
1098
+ "data_type": "float"
1099
+ },
1100
+ "language_model.model.layers.20.self_attn.v_proj": {
1101
+ "bits": 16,
1102
+ "data_type": "float"
1103
+ },
1104
+ "language_model.model.layers.21.self_attn.v_proj": {
1105
+ "bits": 16,
1106
+ "data_type": "float"
1107
+ },
1108
+ "language_model.model.layers.22.self_attn.v_proj": {
1109
+ "bits": 16,
1110
+ "data_type": "float"
1111
+ },
1112
+ "language_model.model.layers.23.self_attn.v_proj": {
1113
+ "bits": 16,
1114
+ "data_type": "float"
1115
+ },
1116
+ "language_model.model.layers.24.self_attn.v_proj": {
1117
+ "bits": 16,
1118
+ "data_type": "float"
1119
+ },
1120
+ "language_model.model.layers.25.self_attn.v_proj": {
1121
+ "bits": 16,
1122
+ "data_type": "float"
1123
+ },
1124
+ "language_model.model.layers.26.self_attn.v_proj": {
1125
+ "bits": 16,
1126
+ "data_type": "float"
1127
+ },
1128
+ "language_model.model.layers.27.self_attn.v_proj": {
1129
+ "bits": 16,
1130
+ "data_type": "float"
1131
+ },
1132
+ "language_model.model.layers.0.self_attn.o_proj": {
1133
+ "bits": 16,
1134
+ "data_type": "float"
1135
+ },
1136
+ "language_model.model.layers.1.self_attn.o_proj": {
1137
+ "bits": 16,
1138
+ "data_type": "float"
1139
+ },
1140
+ "language_model.model.layers.2.self_attn.o_proj": {
1141
+ "bits": 16,
1142
+ "data_type": "float"
1143
+ },
1144
+ "language_model.model.layers.3.self_attn.o_proj": {
1145
+ "bits": 16,
1146
+ "data_type": "float"
1147
+ },
1148
+ "language_model.model.layers.4.self_attn.o_proj": {
1149
+ "bits": 16,
1150
+ "data_type": "float"
1151
+ },
1152
+ "language_model.model.layers.5.self_attn.o_proj": {
1153
+ "bits": 16,
1154
+ "data_type": "float"
1155
+ },
1156
+ "language_model.model.layers.6.self_attn.o_proj": {
1157
+ "bits": 16,
1158
+ "data_type": "float"
1159
+ },
1160
+ "language_model.model.layers.7.self_attn.o_proj": {
1161
+ "bits": 16,
1162
+ "data_type": "float"
1163
+ },
1164
+ "language_model.model.layers.8.self_attn.o_proj": {
1165
+ "bits": 16,
1166
+ "data_type": "float"
1167
+ },
1168
+ "language_model.model.layers.9.self_attn.o_proj": {
1169
+ "bits": 16,
1170
+ "data_type": "float"
1171
+ },
1172
+ "language_model.model.layers.10.self_attn.o_proj": {
1173
+ "bits": 16,
1174
+ "data_type": "float"
1175
+ },
1176
+ "language_model.model.layers.11.self_attn.o_proj": {
1177
+ "bits": 16,
1178
+ "data_type": "float"
1179
+ },
1180
+ "language_model.model.layers.12.self_attn.o_proj": {
1181
+ "bits": 16,
1182
+ "data_type": "float"
1183
+ },
1184
+ "language_model.model.layers.13.self_attn.o_proj": {
1185
+ "bits": 16,
1186
+ "data_type": "float"
1187
+ },
1188
+ "language_model.model.layers.14.self_attn.o_proj": {
1189
+ "bits": 16,
1190
+ "data_type": "float"
1191
+ },
1192
+ "language_model.model.layers.15.self_attn.o_proj": {
1193
+ "bits": 16,
1194
+ "data_type": "float"
1195
+ },
1196
+ "language_model.model.layers.16.self_attn.o_proj": {
1197
+ "bits": 16,
1198
+ "data_type": "float"
1199
+ },
1200
+ "language_model.model.layers.17.self_attn.o_proj": {
1201
+ "bits": 16,
1202
+ "data_type": "float"
1203
+ },
1204
+ "language_model.model.layers.18.self_attn.o_proj": {
1205
+ "bits": 16,
1206
+ "data_type": "float"
1207
+ },
1208
+ "language_model.model.layers.19.self_attn.o_proj": {
1209
+ "bits": 16,
1210
+ "data_type": "float"
1211
+ },
1212
+ "language_model.model.layers.20.self_attn.o_proj": {
1213
+ "bits": 16,
1214
+ "data_type": "float"
1215
+ },
1216
+ "language_model.model.layers.21.self_attn.o_proj": {
1217
+ "bits": 16,
1218
+ "data_type": "float"
1219
+ },
1220
+ "language_model.model.layers.22.self_attn.o_proj": {
1221
+ "bits": 16,
1222
+ "data_type": "float"
1223
+ },
1224
+ "language_model.model.layers.23.self_attn.o_proj": {
1225
+ "bits": 16,
1226
+ "data_type": "float"
1227
+ },
1228
+ "language_model.model.layers.24.self_attn.o_proj": {
1229
+ "bits": 16,
1230
+ "data_type": "float"
1231
+ },
1232
+ "language_model.model.layers.25.self_attn.o_proj": {
1233
+ "bits": 16,
1234
+ "data_type": "float"
1235
+ },
1236
+ "language_model.model.layers.26.self_attn.o_proj": {
1237
+ "bits": 16,
1238
+ "data_type": "float"
1239
+ },
1240
+ "language_model.model.layers.27.self_attn.o_proj": {
1241
+ "bits": 16,
1242
+ "data_type": "float"
1243
+ },
1244
+ ".*moe_gen.*": {
1245
+ "bits": 16,
1246
+ "data_type": "float"
1247
+ },
1248
+ ".*self_attn\\.q_proj.*": {
1249
+ "bits": 16,
1250
+ "data_type": "float"
1251
+ },
1252
+ ".*self_attn\\.k_proj.*": {
1253
+ "bits": 16,
1254
+ "data_type": "float"
1255
+ },
1256
+ ".*self_attn\\.v_proj.*": {
1257
+ "bits": 16,
1258
+ "data_type": "float"
1259
+ },
1260
+ ".*self_attn\\.o_proj.*": {
1261
+ "bits": 16,
1262
+ "data_type": "float"
1263
+ }
1264
+ }
1265
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fd169731d2cbde95e10bf356d66d5997fd885dd8dbb6fb4684da3f23b2585d8
3
+ size 11421892
tokenizer_config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": true,
24
+ "model_max_length": 131072,
25
+ "pad_token": "<|endoftext|>",
26
+ "split_special_tokens": false,
27
+ "tokenizer_class": "Qwen2Tokenizer",
28
+ "unk_token": null
29
+ }
vit_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "hidden_size": 1152,
3
+ "image_size": 980,
4
+ "intermediate_size": 4304,
5
+ "model_type": "siglip_vision_model",
6
+ "num_attention_heads": 16,
7
+ "num_hidden_layers": 27,
8
+ "patch_size": 14
9
+ }