Hugging Face's logo

INCModel2
/
Inkling-Small-MXFP4-AutoRound

Image-Text-to-Text
Transformers
Safetensors
inkling_mm_model
conversational
audio-text-to-text
Mixture of Experts
Eval Results
4-bit precision
auto-round
Model card Files
xet
 Community
xinhe commited on
Commit
b1cf417
·
verified ·
1 Parent(s): 7a3fc03

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .eval_results/aime_2026.yaml +7 -0
  2. .eval_results/gpqa.yaml +7 -0
  3. .eval_results/hle.yaml +8 -0
  4. .eval_results/hmmt_feb_2026.yaml +7 -0
  5. .eval_results/mmmu_pro.yaml +7 -0
  6. .eval_results/swe-bench_pro.yaml +7 -0
  7. .eval_results/swe-bench_verified.yaml +7 -0
  8. .gitattributes +1 -0
  9. README.md +787 -0
  10. chat_template.jinja +129 -0
  11. config.json +2077 -0
  12. model-00001-of-00033.safetensors +3 -0
  13. model-00002-of-00033.safetensors +3 -0
  14. model-00003-of-00033.safetensors +3 -0
  15. model-00004-of-00033.safetensors +3 -0
  16. model-00005-of-00033.safetensors +3 -0
  17. model-00006-of-00033.safetensors +3 -0
  18. model-00007-of-00033.safetensors +3 -0
  19. model-00008-of-00033.safetensors +3 -0
  20. model-00009-of-00033.safetensors +3 -0
  21. model-00010-of-00033.safetensors +3 -0
  22. model-00011-of-00033.safetensors +3 -0
  23. model-00012-of-00033.safetensors +3 -0
  24. model-00013-of-00033.safetensors +3 -0
  25. model-00014-of-00033.safetensors +3 -0
  26. model-00015-of-00033.safetensors +3 -0
  27. model-00016-of-00033.safetensors +3 -0
  28. model-00017-of-00033.safetensors +3 -0
  29. model-00018-of-00033.safetensors +3 -0
  30. model-00019-of-00033.safetensors +3 -0
  31. model-00020-of-00033.safetensors +3 -0
  32. model-00021-of-00033.safetensors +3 -0
  33. model-00022-of-00033.safetensors +3 -0
  34. model-00023-of-00033.safetensors +3 -0
  35. model-00024-of-00033.safetensors +3 -0
  36. model-00025-of-00033.safetensors +3 -0
  37. model-00026-of-00033.safetensors +3 -0
  38. model-00027-of-00033.safetensors +3 -0
  39. model-00028-of-00033.safetensors +3 -0
  40. model-00029-of-00033.safetensors +3 -0
  41. model-00030-of-00033.safetensors +3 -0
  42. model-00031-of-00033.safetensors +3 -0
  43. model-00032-of-00033.safetensors +3 -0
  44. model-00033-of-00033.safetensors +3 -0
  45. model.safetensors.index.json +1135 -0
  46. processor_config.json +46 -0
  47. quantization_config.json +1963 -0
  48. special_tokens_map.json +22 -0
  49. tiktoken/tokenizer.model +3 -0
  50. tokenizer.json +3 -0
.eval_results/aime_2026.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: MathArena/aime_2026
3
+ task_id: MathArena/aime_2026
4
+ value: 95.5
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.eval_results/gpqa.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: Idavidrein/gpqa
3
+ task_id: diamond
4
+ value: 89.5
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.eval_results/hle.yaml ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: cais/hle
3
+ task_id: hle
4
+ value: 31.6
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
8
+ notes: text only
.eval_results/hmmt_feb_2026.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: MathArena/hmmt_feb_2026
3
+ task_id: MathArena/hmmt_feb_2026
4
+ value: 90.2
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.eval_results/mmmu_pro.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: MMMU/MMMU_Pro
3
+ task_id: mmmu_pro_standard_10_options
4
+ value: 74
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.eval_results/swe-bench_pro.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: ScaleAI/SWE-bench_Pro
3
+ task_id: SWE_Bench_Pro
4
+ value: 55.9
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.eval_results/swe-bench_verified.yaml ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ - dataset:
2
+ id: SWE-bench/SWE-bench_Verified
3
+ task_id: swe_bench_%_resolved
4
+ value: 80.2
5
+ source:
6
+ url: https://huggingface.co/thinkingmachines/Inkling-Small
7
+ name: Model Card
.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
README.md ADDED
@@ -0,0 +1,787 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ license_link: https://www.apache.org/licenses/LICENSE-2.0
4
+ pipeline_tag: image-text-to-text
5
+ tags:
6
+ - conversational
7
+ - image-text-to-text
8
+ - audio-text-to-text
9
+ - moe
10
+ library_name: transformers
11
+ ---
12
+
13
+ # Inkling
14
+
15
+ <img src="https://cdn-uploads.huggingface.co/production/uploads/630e8f0bf6f6d700f50ebd2e/AvmDwmrWRMnKjOWvmLieg.png" style="display: block;margin-left: auto;margin-right: auto;width: 30%;">
16
+
17
+ <p align="center">
18
+ <a href="https://huggingface.co/thinkingmachines/Inkling-Small">BF16</a> |
19
+ <a href="https://huggingface.co/thinkingmachines/Inkling-Small-NVFP4">NVFP4</a> |
20
+ <a href="https://tinker.thinkingmachines.ai/playground">Playground</a> |
21
+ <a href="https://github.com/thinking-machines-lab/tinker-cookbook">Tinker Cookbook</a> |
22
+ <a href="https://thinkingmachines.ai/model-acceptable-use-policy">Acceptable Use</a>
23
+ </p>
24
+
25
+ ## 1. General Information
26
+
27
+ Inkling-Small is a general-purpose multimodal model that accepts text, image and audio inputs and generates text outputs. It is intended for use in English and other languages, and across multiple coding languages. The model is designed to be used by developers building AI-powered applications, including agentic and tool-use systems, coding assistants, chatbots, and retrieval-augmented generation systems, and is suitable for general-purpose conversational use, instruction-following, and other natural language and multimodal tasks. It is released with open weights to support research, fine-tuning and integration into third-party products by downstream developers.
28
+
29
+ **Languages:** English, with general multilingual capabilities across other languages.
30
+
31
+ ## 2. Getting Started
32
+
33
+ Try Inkling-Small on the [Tinker Playground](https://tinker.thinkingmachines.ai/playground) or access via API using the [Tinker Cookbook](https://github.com/thinking-machines-lab/tinker-cookbook).
34
+
35
+ Inkling-Small supports local deployment using the following open-source libraries:
36
+
37
+ * SGLang ([recipe](https://docs.sglang.io/cookbook/autoregressive/ThinkingMachines/Inkling-Small))
38
+ * vLLM ([recipe](https://recipes.vllm.ai/thinkingmachines/Inkling-Small))
39
+ * TokenSpeed ([recipe](https://lightseek.org/tokenspeed/recipes/models#Inkling))
40
+ * Unsloth ([recipe](https://unsloth.ai/docs/models/inkling))
41
+ * Huggingface ([recipe](https://hf.co/blog/thinkingmachines-inkling))
42
+
43
+ API access is also available through third party inference providers.
44
+
45
+ ## 3. Model Properties
46
+
47
+ ### Model type
48
+
49
+ Multimodal autoregressive transformer
50
+
51
+ ### Architecture type
52
+
53
+ A 42-layer decoder-only transformer with a sparse Mixture-of-Experts (MoE) feed-forward backbone: each token is routed to 6 of 256 experts, plus 2 shared experts active on every token. Attention is a hybrid of local and global layers. The model is natively multimodal — images are encoded via a hierarchical patch encoder, and audio via discrete token encoding — with all modalities projected into a shared hidden space and processed jointly by the decoder.
54
+
55
+ ### Parameters
56
+
57
+ 276B total, 12B active
58
+
59
+ ### Numerics support
60
+
61
+ BF16 and NVFP4
62
+
63
+ ### Input modalities
64
+
65
+ Inkling-Small accepts text, image, and audio inputs:
66
+
67
+ - Text: UTF-8 encoded text
68
+ - Image: Any pixel-based image input. For optimal performance, each image dimension should be between 40px to 4096px.
69
+ - Audio: WAV format, sampled at 16kHz. For optimal performance, audio length should ideally be under 2 mins.
70
+
71
+ ### Output modalities
72
+
73
+ Inkling-Small generates output as UTF-8 encoded text.
74
+
75
+ ## 4. Training
76
+
77
+ Training data includes a broad variety of content types, including text, images, audio, video.
78
+
79
+ Training data for the model was drawn from publicly available sources, acquired from third-parties, or synthetically generated or augmented. Publicly available data includes content from the public internet and publicly accessible repositories.
80
+
81
+ The training data curation process includes cleaning, processing, and modifying datasets. These processing steps, which vary by data type, may include deduplication and filtering to remove junk or other low-quality data, or to advance safety or other objectives.
82
+
83
+ ## 5. Evaluations
84
+
85
+ <table class="benchmark-results-table benchmark-results-categorized">
86
+ <colgroup>
87
+ <col class="benchmark-name-column">
88
+ <col class="benchmark-model-column" span="10">
89
+ </colgroup>
90
+ <thead>
91
+ <tr class="benchmark-model-group-row">
92
+ <td class="benchmark-heading-spacer" aria-hidden="true"></td>
93
+ <th class="benchmark-model-group" colspan="7" scope="colgroup">
94
+ <span class="benchmark-model-group-anchor"><span class="benchmark-model-group-label">Open weights</span></span>
95
+ </th>
96
+ <th class="benchmark-model-group" colspan="3" scope="colgroup">
97
+ <span class="benchmark-model-group-anchor"><span class="benchmark-model-group-label">Closed weights</span></span>
98
+ </th>
99
+ </tr>
100
+ <tr class="benchmark-model-header-row">
101
+ <td class="benchmark-heading-spacer" aria-hidden="true"></td>
102
+ <th class="benchmark-model benchmark-instant-start"><span class="benchmark-lock-content"><span class="benchmark-model-name">
103
+ <span class="benchmark-model-name-line">Inkling-Small</span></span></span></th>
104
+ <th class="benchmark-model"><span class="benchmark-model-name">
105
+ <span class="benchmark-model-name-line">Qwen3.5</span> <span class="benchmark-model-name-line">397B-A17B</span></span></th>
106
+ <th class="benchmark-model"><span class="benchmark-model-name">
107
+ <span class="benchmark-model-name-line">MiMo V2.5</span></span></th>
108
+ <th class="benchmark-model"><span class="benchmark-model-name">
109
+ <span class="benchmark-model-name-line">Minimax M2.7</span></span></th>
110
+ <th class="benchmark-model"><span class="benchmark-model-name">
111
+ <span class="benchmark-model-name-line">DeepSeek V4</span> <span class="benchmark-model-name-line">Flash</span></span></th>
112
+ <th class="benchmark-model"><span class="benchmark-model-name">
113
+ <span class="benchmark-model-name-line">Nemotron 3</span> <span class="benchmark-model-name-line">Ultra</span></span></th>
114
+ <th class="benchmark-model benchmark-instant-start"><span class="benchmark-lock-content"><span class="benchmark-model-name">
115
+ <span class="benchmark-model-name-line">Inkling</span></span></span></th>
116
+ <th class="benchmark-model"><span class="benchmark-model-name">
117
+ <span class="benchmark-model-name-line">Claude 4.5</span> <span class="benchmark-model-name-line">Haiku</span></span></th>
118
+ <th class="benchmark-model"><span class="benchmark-model-name">
119
+ <span class="benchmark-model-name-line">Gemini 3.5</span> <span class="benchmark-model-name-line">Flash-Lite</span></span></th>
120
+ <th class="benchmark-model"><span class="benchmark-model-name">
121
+ <span class="benchmark-model-name-line">GPT 5.6</span> <span class="benchmark-model-name-line">Luna</span></span></th>
122
+ </tr>
123
+ </thead>
124
+ <tbody class="benchmark-category-group">
125
+ <tr class="benchmark-category-row">
126
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Model Info</span></th>
127
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
128
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
129
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
130
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
131
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
132
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
133
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
134
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
135
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
136
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
137
+ </tr>
138
+ <tr>
139
+ <td class="benchmark-name">
140
+ <span class="benchmark-title">AA Index</span> <span class="benchmark-subtitle">(v4.1)</span>
141
+ </td>
142
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">40.0%</span></td>
143
+ <td class="benchmark-value">34.0%</td>
144
+ <td class="benchmark-value">37.0%</td>
145
+ <td class="benchmark-value">38.0%</td>
146
+ <td class="benchmark-value">40.0%</td>
147
+ <td class="benchmark-value">38.0%</td>
148
+ <td class="benchmark-value">41.0%</td>
149
+ <td class="benchmark-value">30.0%</td>
150
+ <td class="benchmark-value">36.0%</td>
151
+ <td class="benchmark-value">49.0%</td>
152
+ </tr>
153
+ <tr>
154
+ <td class="benchmark-name">
155
+ <span class="benchmark-title">Params (B)</span> <span class="benchmark-subtitle">(activated / total)</span>
156
+ </td>
157
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">12 / 276</span></td>
158
+ <td class="benchmark-value">17 / 397</td>
159
+ <td class="benchmark-value">15 / 310</td>
160
+ <td class="benchmark-value">10 / 230</td>
161
+ <td class="benchmark-value">13 / 284</td>
162
+ <td class="benchmark-value">55 / 550</td>
163
+ <td class="benchmark-value">41 / 975</td>
164
+ <td class="benchmark-value">–</td>
165
+ <td class="benchmark-value">–</td>
166
+ <td class="benchmark-value">–</td>
167
+ </tr>
168
+ </tbody>
169
+ <tbody class="benchmark-category-group">
170
+ <tr class="benchmark-category-row">
171
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Agentic (coding)</span></th>
172
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
173
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
174
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
175
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
176
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
177
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
178
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
179
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
180
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
181
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
182
+ </tr>
183
+ <tr>
184
+ <td class="benchmark-name">
185
+ <span class="benchmark-title">SWEBench Verified</span>
186
+ </td>
187
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">80.2%</span></td>
188
+ <td class="benchmark-value">76.4%</td>
189
+ <td class="benchmark-value">71.0%</td>
190
+ <td class="benchmark-value">79.9%</td>
191
+ <td class="benchmark-value">79.0%</td>
192
+ <td class="benchmark-value">70.7%</td>
193
+ <td class="benchmark-value">77.6%</td>
194
+ <td class="benchmark-value">73.3%</td>
195
+ <td class="benchmark-value">75.0%</td>
196
+ <td class="benchmark-value">93.0%</td>
197
+ </tr>
198
+ <tr>
199
+ <td class="benchmark-name">
200
+ <span class="benchmark-title">SWEBench Pro</span> <span class="benchmark-subtitle">(public)</span>
201
+ </td>
202
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">55.9%</span></td>
203
+ <td class="benchmark-value">50.9%</td>
204
+ <td class="benchmark-value">56.1%</td>
205
+ <td class="benchmark-value">56.2%</td>
206
+ <td class="benchmark-value">52.6%</td>
207
+ <td class="benchmark-value">46.4%</td>
208
+ <td class="benchmark-value">54.3%</td>
209
+ <td class="benchmark-value">39.5%</td>
210
+ <td class="benchmark-value">54.2%</td>
211
+ <td class="benchmark-value">62.7%</td>
212
+ </tr>
213
+ <tr>
214
+ <td class="benchmark-name">
215
+ <span class="benchmark-title">Terminal Bench 2.1</span> <span class="benchmark-subtitle">(best harness)</span>
216
+ </td>
217
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">64.7%</span></td>
218
+ <td class="benchmark-value">51.3%</td>
219
+ <td class="benchmark-value">63.7%</td>
220
+ <td class="benchmark-value">55.4%</td>
221
+ <td class="benchmark-value">61.8%</td>
222
+ <td class="benchmark-value">56.4%</td>
223
+ <td class="benchmark-value">63.8%</td>
224
+ <td class="benchmark-value">44.2%</td>
225
+ <td class="benchmark-value">54.0%</td>
226
+ <td class="benchmark-value">82.5%</td>
227
+ </tr>
228
+ <tr>
229
+ <td class="benchmark-name">
230
+ <span class="benchmark-title">SciCode</span>
231
+ </td>
232
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">48.7%</span></td>
233
+ <td class="benchmark-value">42.0%</td>
234
+ <td class="benchmark-value">43.1%</td>
235
+ <td class="benchmark-value">47.0%</td>
236
+ <td class="benchmark-value">44.9%</td>
237
+ <td class="benchmark-value">39.9%</td>
238
+ <td class="benchmark-value">46.1%</td>
239
+ <td class="benchmark-value">43.3%</td>
240
+ <td class="benchmark-value">40.9%</td>
241
+ <td class="benchmark-value">50.0%</td>
242
+ </tr>
243
+ </tbody>
244
+ <tbody class="benchmark-category-group">
245
+ <tr class="benchmark-category-row">
246
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Agentic (general)</span></th>
247
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
248
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
249
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
250
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
251
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
252
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
253
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
254
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
255
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
256
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
257
+ </tr>
258
+ <tr>
259
+ <td class="benchmark-name">
260
+ <span class="benchmark-title">GDPval-AA v2</span>
261
+ </td>
262
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">1269</span></td>
263
+ <td class="benchmark-value">962</td>
264
+ <td class="benchmark-value">1145</td>
265
+ <td class="benchmark-value">1159</td>
266
+ <td class="benchmark-value">1189</td>
267
+ <td class="benchmark-value">1164</td>
268
+ <td class="benchmark-value">1238</td>
269
+ <td class="benchmark-value">911</td>
270
+ <td class="benchmark-value">1139</td>
271
+ <td class="benchmark-value">1530</td>
272
+ </tr>
273
+ <tr>
274
+ <td class="benchmark-name">
275
+ <span class="benchmark-title">MCP Atlas</span> <span class="benchmark-subtitle">(public / all)</span>
276
+ </td>
277
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">79.6/79.2%</span></td>
278
+ <td class="benchmark-value">74.2%/–</td>
279
+ <td class="benchmark-value">–</td>
280
+ <td class="benchmark-value">49.4%/–</td>
281
+ <td class="benchmark-value">69.0%/–</td>
282
+ <td class="benchmark-value">47.4/44.7%</td>
283
+ <td class="benchmark-value">78.8/76.0%</td>
284
+ <td class="benchmark-value">41.2/40.2%</td>
285
+ <td class="benchmark-value">79.8/76.8%</td>
286
+ <td class="benchmark-value">77.0/75.0%</td>
287
+ </tr>
288
+ <tr>
289
+ <td class="benchmark-name">
290
+ <span class="benchmark-title">Tau 3 Banking</span>
291
+ </td>
292
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">15.5%</span></td>
293
+ <td class="benchmark-value">13.4%</td>
294
+ <td class="benchmark-value">6.6%</td>
295
+ <td class="benchmark-value">8.9%</td>
296
+ <td class="benchmark-value">22.9%</td>
297
+ <td class="benchmark-value">13.8%</td>
298
+ <td class="benchmark-value">23.7%</td>
299
+ <td class="benchmark-value">9.1%</td>
300
+ <td class="benchmark-value">16.5%</td>
301
+ <td class="benchmark-value">24.3%</td>
302
+ </tr>
303
+ <tr>
304
+ <td class="benchmark-name">
305
+ <span class="benchmark-title">BrowseComp</span> <span class="benchmark-subtitle">(with context management)</span>
306
+ </td>
307
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">77.4%</span></td>
308
+ <td class="benchmark-value">78.6%</td>
309
+ <td class="benchmark-value">–</td>
310
+ <td class="benchmark-value">76.3%</td>
311
+ <td class="benchmark-value">73.2%</td>
312
+ <td class="benchmark-value">63.0%</td>
313
+ <td class="benchmark-value">77.1%</td>
314
+ <td class="benchmark-value">–</td>
315
+ <td class="benchmark-value">–</td>
316
+ <td class="benchmark-value">84.0%</td>
317
+ </tr>
318
+ <tr>
319
+ <td class="benchmark-name">
320
+ <span class="benchmark-title">Toolathlon Verified</span>
321
+ </td>
322
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">54.4%</span></td>
323
+ <td class="benchmark-value">40.7%</td>
324
+ <td class="benchmark-value">49.1%</td>
325
+ <td class="benchmark-value">47.5%</td>
326
+ <td class="benchmark-value">50.9%</td>
327
+ <td class="benchmark-value">34.3%</td>
328
+ <td class="benchmark-value">45.5%</td>
329
+ <td class="benchmark-value">26.9%</td>
330
+ <td class="benchmark-value">57.1%</td>
331
+ <td class="benchmark-value">67.9%</td>
332
+ </tr>
333
+ <tr>
334
+ <td class="benchmark-name">
335
+ <span class="benchmark-title">AA-Briefcase</span>
336
+ </td>
337
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">917</span></td>
338
+ <td class="benchmark-value">–</td>
339
+ <td class="benchmark-value">–</td>
340
+ <td class="benchmark-value">–</td>
341
+ <td class="benchmark-value">833</td>
342
+ <td class="benchmark-value">870</td>
343
+ <td class="benchmark-value">839</td>
344
+ <td class="benchmark-value">612</td>
345
+ <td class="benchmark-value">–</td>
346
+ <td class="benchmark-value">–</td>
347
+ </tr>
348
+ </tbody>
349
+ <tbody class="benchmark-category-group">
350
+ <tr class="benchmark-category-row">
351
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Reasoning (general)</span></th>
352
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
353
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
354
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
355
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
356
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
357
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
358
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
359
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
360
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
361
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
362
+ </tr>
363
+ <tr>
364
+ <td class="benchmark-name">
365
+ <span class="benchmark-title">GPQA Diamond</span>
366
+ </td>
367
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">89.5%</span></td>
368
+ <td class="benchmark-value">89.3%</td>
369
+ <td class="benchmark-value">84.9%</td>
370
+ <td class="benchmark-value">87.4%</td>
371
+ <td class="benchmark-value">89.4%</td>
372
+ <td class="benchmark-value">86.7%</td>
373
+ <td class="benchmark-value">87.2%</td>
374
+ <td class="benchmark-value">67.2%</td>
375
+ <td class="benchmark-value">83.8%</td>
376
+ <td class="benchmark-value">89.5%</td>
377
+ </tr>
378
+ <tr>
379
+ <td class="benchmark-name">
380
+ <span class="benchmark-title">HLE</span> <span class="benchmark-subtitle">(text only)</span>
381
+ </td>
382
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">31.6%</span></td>
383
+ <td class="benchmark-value">27.3%</td>
384
+ <td class="benchmark-value">25.2%</td>
385
+ <td class="benchmark-value">28.1%</td>
386
+ <td class="benchmark-value">32.1%</td>
387
+ <td class="benchmark-value">26.6%</td>
388
+ <td class="benchmark-value">29.7%</td>
389
+ <td class="benchmark-value">9.7%</td>
390
+ <td class="benchmark-value">17.5%</td>
391
+ <td class="benchmark-value">35.6%</td>
392
+ </tr>
393
+ <tr>
394
+ <td class="benchmark-name">
395
+ <span class="benchmark-title">HLE</span> <span class="benchmark-subtitle">(with tools)</span>
396
+ </td>
397
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">47.8%</span></td>
398
+ <td class="benchmark-value">48.3%</td>
399
+ <td class="benchmark-value">40.0%</td>
400
+ <td class="benchmark-value">40.3%</td>
401
+ <td class="benchmark-value">45.1%</td>
402
+ <td class="benchmark-value">37.4%</td>
403
+ <td class="benchmark-value">46.0%</td>
404
+ <td class="benchmark-value">17.8%</td>
405
+ <td class="benchmark-value">42.5%</td>
406
+ <td class="benchmark-value">48.9%</td>
407
+ </tr>
408
+ <tr>
409
+ <td class="benchmark-name">
410
+ <span class="benchmark-title">AIME 2026</span>
411
+ </td>
412
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">95.5%</span></td>
413
+ <td class="benchmark-value">93.3%</td>
414
+ <td class="benchmark-value">93.6%</td>
415
+ <td class="benchmark-value">87.7%</td>
416
+ <td class="benchmark-value">95.8%</td>
417
+ <td class="benchmark-value">94.2%</td>
418
+ <td class="benchmark-value">97.1%</td>
419
+ <td class="benchmark-value">85.1%</td>
420
+ <td class="benchmark-value">82.2%</td>
421
+ <td class="benchmark-value">97.6%</td>
422
+ </tr>
423
+ <tr>
424
+ <td class="benchmark-name">
425
+ <span class="benchmark-title">HMMT Feb 2026</span>
426
+ </td>
427
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">90.2%</span></td>
428
+ <td class="benchmark-value">87.9%</td>
429
+ <td class="benchmark-value">82.6%</td>
430
+ <td class="benchmark-value">71.2%</td>
431
+ <td class="benchmark-value">93.9%</td>
432
+ <td class="benchmark-value">78.8%</td>
433
+ <td class="benchmark-value">86.3%</td>
434
+ <td class="benchmark-value">66.7%</td>
435
+ <td class="benchmark-value">63.6%</td>
436
+ <td class="benchmark-value">98.5%</td>
437
+ </tr>
438
+ <tr>
439
+ <td class="benchmark-name">
440
+ <span class="benchmark-title">CritPt</span>
441
+ </td>
442
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">8.3%</span></td>
443
+ <td class="benchmark-value">1.7%</td>
444
+ <td class="benchmark-value">3.7%</td>
445
+ <td class="benchmark-value">0.6%</td>
446
+ <td class="benchmark-value">7.1%</td>
447
+ <td class="benchmark-value">3.1%</td>
448
+ <td class="benchmark-value">5.4%</td>
449
+ <td class="benchmark-value">0.0%</td>
450
+ <td class="benchmark-value">0.0%</td>
451
+ <td class="benchmark-value">20.6%</td>
452
+ </tr>
453
+ </tbody>
454
+ <tbody class="benchmark-category-group">
455
+ <tr class="benchmark-category-row">
456
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Reasoning (abstract)</span></th>
457
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
458
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
459
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
460
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
461
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
462
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
463
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
464
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
465
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
466
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
467
+ </tr>
468
+ <tr>
469
+ <td class="benchmark-name">
470
+ <span class="benchmark-title">ARC-AGI-1</span>
471
+ </td>
472
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">84.0%</span></td>
473
+ <td class="benchmark-value">–</td>
474
+ <td class="benchmark-value">–</td>
475
+ <td class="benchmark-value">–</td>
476
+ <td class="benchmark-value">–</td>
477
+ <td class="benchmark-value">–</td>
478
+ <td class="benchmark-value">79.5%</td>
479
+ <td class="benchmark-value">47.7%</td>
480
+ <td class="benchmark-value">–</td>
481
+ <td class="benchmark-value">87.7%</td>
482
+ </tr>
483
+ <tr>
484
+ <td class="benchmark-name">
485
+ <span class="benchmark-title">ARC-AGI-2</span>
486
+ </td>
487
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">40.1%</span></td>
488
+ <td class="benchmark-value">–</td>
489
+ <td class="benchmark-value">–</td>
490
+ <td class="benchmark-value">–</td>
491
+ <td class="benchmark-value">–</td>
492
+ <td class="benchmark-value">–</td>
493
+ <td class="benchmark-value">36.5%</td>
494
+ <td class="benchmark-value">4.0%</td>
495
+ <td class="benchmark-value">–</td>
496
+ <td class="benchmark-value">47.6%</td>
497
+ </tr>
498
+ </tbody>
499
+ <tbody class="benchmark-category-group">
500
+ <tr class="benchmark-category-row">
501
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Factuality</span></th>
502
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
503
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
504
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
505
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
506
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
507
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
508
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
509
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
510
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
511
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
512
+ </tr>
513
+ <tr>
514
+ <td class="benchmark-name">
515
+ <span class="benchmark-title">SimpleQA Verified</span>
516
+ </td>
517
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">20.6%</span></td>
518
+ <td class="benchmark-value">26.0%</td>
519
+ <td class="benchmark-value">16.1%</td>
520
+ <td class="benchmark-value">13.5%</td>
521
+ <td class="benchmark-value">34.1%</td>
522
+ <td class="benchmark-value">32.4%</td>
523
+ <td class="benchmark-value">43.9%</td>
524
+ <td class="benchmark-value">5.9%</td>
525
+ <td class="benchmark-value">44.1%</td>
526
+ <td class="benchmark-value">41.7%</td>
527
+ </tr>
528
+ <tr>
529
+ <td class="benchmark-name">
530
+ <span class="benchmark-title">AA Omniscience</span> <span class="benchmark-subtitle">(index)</span>
531
+ </td>
532
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">-9.0</span></td>
533
+ <td class="benchmark-value">-29.8</td>
534
+ <td class="benchmark-value">-9.3</td>
535
+ <td class="benchmark-value">0.7</td>
536
+ <td class="benchmark-value">-22.9</td>
537
+ <td class="benchmark-value">-1.0</td>
538
+ <td class="benchmark-value">2.1</td>
539
+ <td class="benchmark-value">-4.2</td>
540
+ <td class="benchmark-value">6.9</td>
541
+ <td class="benchmark-value">-11.6</td>
542
+ </tr>
543
+ </tbody>
544
+ <tbody class="benchmark-category-group">
545
+ <tr class="benchmark-category-row">
546
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Chat</span></th>
547
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
548
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
549
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
550
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
551
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
552
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
553
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
554
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
555
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
556
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
557
+ </tr>
558
+ <tr>
559
+ <td class="benchmark-name">
560
+ <span class="benchmark-title">IFBench</span>
561
+ </td>
562
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">82.2%</span></td>
563
+ <td class="benchmark-value">78.8%</td>
564
+ <td class="benchmark-value">67.1%</td>
565
+ <td class="benchmark-value">75.7%</td>
566
+ <td class="benchmark-value">79.2%</td>
567
+ <td class="benchmark-value">81.4%</td>
568
+ <td class="benchmark-value">79.8%</td>
569
+ <td class="benchmark-value">54.3%</td>
570
+ <td class="benchmark-value">78.6%</td>
571
+ <td class="benchmark-value">67.3%</td>
572
+ </tr>
573
+ <tr>
574
+ <td class="benchmark-name">
575
+ <span class="benchmark-title">Global-MMLU-Lite</span>
576
+ </td>
577
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">86.7%</span></td>
578
+ <td class="benchmark-value">90.0%</td>
579
+ <td class="benchmark-value">83.5%</td>
580
+ <td class="benchmark-value">83.9%</td>
581
+ <td class="benchmark-value">88.4%</td>
582
+ <td class="benchmark-value">85.6%</td>
583
+ <td class="benchmark-value">88.7%</td>
584
+ <td class="benchmark-value">83.4%</td>
585
+ <td class="benchmark-value">89.4%</td>
586
+ <td class="benchmark-value">88.7%</td>
587
+ </tr>
588
+ </tbody>
589
+ <tbody class="benchmark-category-group">
590
+ <tr class="benchmark-category-row">
591
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Safety</span></th>
592
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
593
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
594
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
595
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
596
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
597
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
598
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
599
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
600
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
601
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
602
+ </tr>
603
+ <tr>
604
+ <td class="benchmark-name">
605
+ <span class="benchmark-title">StrongREJECT</span>
606
+ </td>
607
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">98.4%</span></td>
608
+ <td class="benchmark-value">99.4%</td>
609
+ <td class="benchmark-value">99.3%</td>
610
+ <td class="benchmark-value">99.4%</td>
611
+ <td class="benchmark-value">97.4%</td>
612
+ <td class="benchmark-value">98.7%</td>
613
+ <td class="benchmark-value">98.6%</td>
614
+ <td class="benchmark-value">98.6%</td>
615
+ <td class="benchmark-value">97.6%</td>
616
+ <td class="benchmark-value">98.7%</td>
617
+ </tr>
618
+ <tr>
619
+ <td class="benchmark-name">
620
+ <span class="benchmark-title">FORTRESS</span> <span class="benchmark-subtitle">(adversarial)</span>
621
+ </td>
622
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">71.6%</span></td>
623
+ <td class="benchmark-value">77.3%</td>
624
+ <td class="benchmark-value">64.8%</td>
625
+ <td class="benchmark-value">86.3%</td>
626
+ <td class="benchmark-value">32.0%</td>
627
+ <td class="benchmark-value">77.6%</td>
628
+ <td class="benchmark-value">78.0%</td>
629
+ <td class="benchmark-value">91.3%</td>
630
+ <td class="benchmark-value">70.7%</td>
631
+ <td class="benchmark-value">83.8%</td>
632
+ </tr>
633
+ <tr>
634
+ <td class="benchmark-name">
635
+ <span class="benchmark-title">FORTRESS</span> <span class="benchmark-subtitle">(benign)</span>
636
+ </td>
637
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">96.9%</span></td>
638
+ <td class="benchmark-value">95.4%</td>
639
+ <td class="benchmark-value">94.6%</td>
640
+ <td class="benchmark-value">90.1%</td>
641
+ <td class="benchmark-value">99.2%</td>
642
+ <td class="benchmark-value">90.6%</td>
643
+ <td class="benchmark-value">95.9%</td>
644
+ <td class="benchmark-value">94.1%</td>
645
+ <td class="benchmark-value">95.5%</td>
646
+ <td class="benchmark-value">97.8%</td>
647
+ </tr>
648
+ </tbody>
649
+ <tbody class="benchmark-category-group">
650
+ <tr class="benchmark-category-row">
651
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Vision</span></th>
652
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
653
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
654
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
655
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
656
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
657
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
658
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
659
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
660
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
661
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
662
+ </tr>
663
+ <tr>
664
+ <td class="benchmark-name">
665
+ <span class="benchmark-title">MMMU Pro</span> <span class="benchmark-subtitle">(Standard 10)</span>
666
+ </td>
667
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">74.0%</span></td>
668
+ <td class="benchmark-value">77.3%</td>
669
+ <td class="benchmark-value">75.4%</td>
670
+ <td class="benchmark-value">–</td>
671
+ <td class="benchmark-value">–</td>
672
+ <td class="benchmark-value">–</td>
673
+ <td class="benchmark-value">73.5%</td>
674
+ <td class="benchmark-value">58.6%</td>
675
+ <td class="benchmark-value">79.0%</td>
676
+ <td class="benchmark-value">78.6%</td>
677
+ </tr>
678
+ <tr>
679
+ <td class="benchmark-name">
680
+ <span class="benchmark-title">Charxiv RQ</span> <span class="benchmark-subtitle">(original / with python)</span>
681
+ </td>
682
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">77.4/81.3%</span></td>
683
+ <td class="benchmark-value">80.8%/–</td>
684
+ <td class="benchmark-value">81.0%/–</td>
685
+ <td class="benchmark-value">–</td>
686
+ <td class="benchmark-value">–</td>
687
+ <td class="benchmark-value">–</td>
688
+ <td class="benchmark-value">78.1/82.0%</td>
689
+ <td class="benchmark-value">57.4%/–</td>
690
+ <td class="benchmark-value">70.0%/–</td>
691
+ <td class="benchmark-value">81.4%/–</td>
692
+ </tr>
693
+ </tbody>
694
+ <tbody class="benchmark-category-group">
695
+ <tr class="benchmark-category-row">
696
+ <th class="benchmark-category-label" scope="rowgroup"><span class="benchmark-category-label-text">Audio</span></th>
697
+ <td class="benchmark-category-cell benchmark-value benchmark-instant-start" aria-hidden="true"></td>
698
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
699
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
700
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
701
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
702
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
703
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
704
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
705
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
706
+ <td class="benchmark-category-cell" aria-hidden="true"></td>
707
+ </tr>
708
+ <tr>
709
+ <td class="benchmark-name">
710
+ <span class="benchmark-title">Audio MC</span>
711
+ </td>
712
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">54.9%</span></td>
713
+ <td class="benchmark-value">–</td>
714
+ <td class="benchmark-value">30.4%</td>
715
+ <td class="benchmark-value">–</td>
716
+ <td class="benchmark-value">–</td>
717
+ <td class="benchmark-value">–</td>
718
+ <td class="benchmark-value">56.6%</td>
719
+ <td class="benchmark-value">–</td>
720
+ <td class="benchmark-value">33.6%</td>
721
+ <td class="benchmark-value">–</td>
722
+ </tr>
723
+ <tr>
724
+ <td class="benchmark-name">
725
+ <span class="benchmark-title">MMAU</span>
726
+ </td>
727
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">77.0%</span></td>
728
+ <td class="benchmark-value">–</td>
729
+ <td class="benchmark-value">73.6%</td>
730
+ <td class="benchmark-value">–</td>
731
+ <td class="benchmark-value">–</td>
732
+ <td class="benchmark-value">–</td>
733
+ <td class="benchmark-value">77.2%</td>
734
+ <td class="benchmark-value">–</td>
735
+ <td class="benchmark-value">75.2%</td>
736
+ <td class="benchmark-value">–</td>
737
+ </tr>
738
+ <tr>
739
+ <td class="benchmark-name">
740
+ <span class="benchmark-title">VoiceBench</span>
741
+ </td>
742
+ <td class="benchmark-value benchmark-instant-start"><span class="benchmark-lock-content">90.1%</span></td>
743
+ <td class="benchmark-value">–</td>
744
+ <td class="benchmark-value">86.4%</td>
745
+ <td class="benchmark-value">–</td>
746
+ <td class="benchmark-value">–</td>
747
+ <td class="benchmark-value">–</td>
748
+ <td class="benchmark-value">91.4%</td>
749
+ <td class="benchmark-value">–</td>
750
+ <td class="benchmark-value">85.9%</td>
751
+ <td class="benchmark-value">–</td>
752
+ </tr>
753
+ </tbody>
754
+ </table>
755
+
756
+ - <small>Inkling-Small against open-and closed-weights models across the full eval suite. Activated and total parameters are given for scale; a dash means the score was not available at the time of writing.</small>
757
+ - <small>SWEBench Verified: Inkling and Inkling-Small’s numbers are reported using a bash-only harness. We use self-reported numbers for external models.</small>
758
+ - <small>Terminal Bench 2.1: Inkling and Inkling-Small’s numbers are reported using an internal coding harness. A small number of solutions were found to be contaminated from web search and were assigned a score of 0. We use self-reported numbers for external models where available. Otherwise, we report performance using our internal harness.</small>
759
+ - <small>Audio MC: Other models were evaluated internally since they are not on the official leaderboard.</small>
760
+ - <small>VoiceBench: VoiceBench uses rule-based, hard-coded string matching for grading, making the evaluation sensitive to output-formatting differences. We therefore added a system message instructing models to follow the expected answer format.</small>
761
+ - <small>HLE with tools: We benchmarked Minimax M2.7, Claude 4.5 Haiku, Gemini 3.5 Flash-Lite, and GPT 5.6 Luna using our internal harness.</small>
762
+
763
+ ## 6. Safety
764
+
765
+ We conducted safety evaluations ahead of release, spanning both everyday human-AI interaction and dangerous-capability testing. Because Inkling-Small is multimodal, we paid attention to whether safety behavior held consistently across text, audio, and image inputs. We applied mitigations to reduce risks before release.
766
+
767
+ For everyday interaction, we evaluated sycophancy, harmful manipulation, and psychological-harm patterns like parasocial dependency and validation of delusional reasoning, including through multi-turn, open-ended external red-teaming designed to surface issues that only emerge over longer conversations. We also assessed whether the model refuses genuinely harmful requests without over-refusing benign ones. For CBRN and cyber, we assessed knowledge and procedural uplift through internal evaluations, external testing, and refusal-suppressed variants intended to estimate latent capability with safeguards removed. For loss of control, we evaluated agentic capability, strategic deception, and sabotage potential, benchmarked against public frontier models, and found the model materially below frontier capabilities.
768
+
769
+ Across all areas, we concluded that Inkling-Small did not present risk of material uplift beyond what's already available in the open-weight ecosystem.
770
+
771
+ The residual risks identified in our evaluations — specifically, Inkling-Small’s occasional tendency to comply with role-play and indirectly framed prompts concerning harmful topics — are consistent with what you would see from any open-weight model, and are best addressed with defense-in-depth rather than relying on the model's refusals alone. Common downstream moderation tools, such as Llama Guard, are compatible with Inkling-Small and can be layered around the model to catch jailbreak attempts, filter unsafe outputs, and enforce use-case-specific policies. We would encourage treating this kind of input/output classification as a part of your deployment stack, especially for consumer-facing or high-traffic applications where adversarial prompting is more likely.
772
+
773
+ ## 7. Bias, risks and limitations
774
+
775
+ Inkling-Small may exhibit general limitations common to foundation models, including hallucination (generating plausible but factually incorrect or unsupported content), occasional failures to follow instructions precisely, and degraded performance in long multi-turn conversations. As with other large-scale models trained on web-derived and synthetic data, Inkling-Small may reflect biases present in its training data, including demographic, cultural, or linguistic biases, and may perform unevenly across languages, dialects, or subject domains that were less represented during training.
776
+
777
+ Inkling-Small's knowledge is limited to information available as of its training cutoff, and it may not reflect events, developments, or changes that occurred afterward.
778
+
779
+ We recommend that downstream developers and deployers apply appropriate human oversight and review for outputs used in high-stakes or safety-critical contexts, rather than relying on Inkling-Small's outputs without verification.
780
+
781
+ - Conduct their own evaluation of Inkling-Small's performance, safety, and fairness for their specific use case, language, and population prior to deployment, particularly for applications involving vulnerable groups.
782
+ - Implement additional safeguards – such as content filtering, rate limiting, and monitoring – at the application layer, especially for open deployment contexts where Inkling-Small's built-in mitigations may not be sufficient on their own.
783
+ - Avoid deploying Inkling-Small in domains such as medical, legal, or safety-critical decision-making without additional fine-tuning, domain-specific validation, and human oversight
784
+
785
+ ## 8. Legal
786
+
787
+ [Training Data Documentation](https://thinkingmachines.ai/training-data-documentation/)
chat_template.jinja ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set effort_map = {"none": 0.0, "minimal": 0.1, "low": 0.2, "medium": 0.7, "high": 0.9, "max": 0.99} -%}
2
+ {%- set role_token = {"user": "<|message_user|>", "assistant": "<|message_model|>", "system": "<|message_system|>", "tool": "<|message_tool|>"} -%}
3
+
4
+ {%- macro emit_thinking_effort() -%}
5
+ {%- set eff = reasoning_effort if reasoning_effort is defined and reasoning_effort is not none else 0.9 -%}
6
+ {%- if eff is string -%}
7
+ {%- set key = eff | trim -%}
8
+ {%- if key not in effort_map -%}
9
+ {{- raise_exception("Unknown reasoning_effort: " ~ eff) -}}
10
+ {%- endif -%}
11
+ {%- set num = effort_map[key] -%}
12
+ {%- else -%}
13
+ {%- set num = eff | float -%}
14
+ {%- endif -%}
15
+ {%- if num < 0.0 or num > 0.99 -%}
16
+ {{- raise_exception("reasoning_effort must be in [0.0, 0.99]") -}}
17
+ {%- endif -%}
18
+ {{- "<|message_system|><|content_text|>Thinking effort level: " -}}
19
+ {%- if num == 0.0 -%}0{%- else -%}{{ num }}{%- endif -%}
20
+ {{- "<|end_message|>" -}}
21
+ {%- endmacro -%}
22
+
23
+ {%- if tools -%}
24
+ {%- set tool_state = namespace(specs=[]) -%}
25
+ {%- for tool in tools -%}
26
+ {%- set fn = tool.function if tool.function is defined else tool -%}
27
+ {%- set spec = {
28
+ "description": (fn.description if fn.description is defined and fn.description else ""),
29
+ "name": fn.name,
30
+ "parameters": (fn.parameters if fn.parameters is defined and fn.parameters else {}),
31
+ "type": (tool.type if tool.type is defined and tool.type else "function"),
32
+ } -%}
33
+ {%- set tool_state.specs = tool_state.specs + [spec] -%}
34
+ {%- endfor -%}
35
+ {{- "<|message_system|>tool_declare<|content_xml|>" -}}
36
+ {{- tool_state.specs | tojson(sort_keys=true, separators=(",", ":")) -}}
37
+ {{- "<|end_message|>" -}}
38
+ {%- endif -%}
39
+
40
+ {%- set state = namespace(effort_emitted=false) -%}
41
+ {%- for message in messages -%}
42
+ {%- if message.role not in role_token -%}
43
+ {{- raise_exception("Unknown message role: " ~ message.role) -}}
44
+ {%- endif -%}
45
+ {%- if not state.effort_emitted and message.role != "system" -%}
46
+ {{- emit_thinking_effort() -}}
47
+ {%- set state.effort_emitted = true -%}
48
+ {%- endif -%}
49
+
50
+ {%- set rtok = role_token[message.role] -%}
51
+
52
+ {%- if message.role == "tool" -%}
53
+ {%- set tool_name_state = namespace(name="") -%}
54
+ {%- if message.name is defined and message.name -%}
55
+ {%- set tool_name_state.name = message.name -%}
56
+ {%- elif message.tool_call_id is defined and message.tool_call_id -%}
57
+ {%- for prev in messages -%}
58
+ {%- if prev.role == "assistant" and prev.tool_calls -%}
59
+ {%- for tc in prev.tool_calls -%}
60
+ {%- if tc.id is defined and tc.id == message.tool_call_id and tc.function.name is defined -%}
61
+ {%- set tool_name_state.name = tc.function.name -%}
62
+ {%- endif -%}
63
+ {%- endfor -%}
64
+ {%- endif -%}
65
+ {%- endfor -%}
66
+ {%- endif -%}
67
+ {{- rtok -}}
68
+ {%- if tool_name_state.name -%}{{- tool_name_state.name -}}{%- endif -%}
69
+ {{- "<|content_text|>" -}}
70
+ {%- if message.content is string -%}{{- message.content -}}{%- endif -%}
71
+ {{- "<|end_message|>" -}}
72
+
73
+ {%- else -%}
74
+ {%- if message.role == "assistant" and message.reasoning_content is defined and message.reasoning_content -%}
75
+ {{- "<|message_model|><|content_thinking|>" ~ message.reasoning_content ~ "<|end_message|>" -}}
76
+ {%- endif -%}
77
+
78
+ {%- if message.content is string -%}
79
+ {{- rtok ~ "<|content_text|>" ~ message.content ~ "<|end_message|>" -}}
80
+ {%- elif message.content -%}
81
+ {%- for part in message.content -%}
82
+ {%- if part is string -%}
83
+ {{- rtok ~ "<|content_text|>" ~ part ~ "<|end_message|>" -}}
84
+ {%- elif part.type is not defined or part.type in ("text", "input_text") -%}
85
+ {%- set text_part = (part.text if part.text is defined and part.text is string else "") -%}
86
+ {{- rtok ~ "<|content_text|>" ~ text_part ~ "<|end_message|>" -}}
87
+ {%- elif part.type in ("image", "input_image", "image_url") -%}
88
+ {{- rtok ~ "<|content_image|><|unused_200054|><|end_message|>" -}}
89
+ {%- elif part.type in ("audio", "input_audio", "audio_url") -%}
90
+ {{- rtok ~ "<|content_audio_input|><|unused_200053|><|audio_end|><|end_message|>" -}}
91
+ {%- else -%}
92
+ {{- raise_exception("Unsupported content part type: " ~ part.type) -}}
93
+ {%- endif -%}
94
+ {%- endfor -%}
95
+ {%- endif -%}
96
+
97
+ {%- if message.role == "assistant" and message.tool_calls -%}
98
+ {%- for tc in message.tool_calls -%}
99
+ {%- set fn = tc.function -%}
100
+ {%- if fn.name is not defined or fn.name is not string -%}
101
+ {{- raise_exception("tool call function name must be a string") -}}
102
+ {%- endif -%}
103
+ {%- set args = fn.arguments if fn.arguments is defined and fn.arguments else {} -%}
104
+ {%- if args is string -%}
105
+ {{- raise_exception("tool call arguments must be a parsed object, not a JSON string; canonicalize upstream") -}}
106
+ {%- endif -%}
107
+ {%- if args is not mapping -%}
108
+ {{- raise_exception("tool call arguments must be an object") -}}
109
+ {%- endif -%}
110
+ {{- "<|message_model|>" ~ fn.name ~ "<|content_invoke_tool_json|>" -}}
111
+ {{- '{"name":' ~ (fn.name | tojson(sort_keys=true, separators=(",", ":"))) ~ ',"args":' -}}
112
+ {{- (args | tojson(sort_keys=true, separators=(",", ":"))) -}}
113
+ {{- "}<|end_message|>" -}}
114
+ {%- endfor -%}
115
+ {%- endif -%}
116
+
117
+ {%- if message.role == "assistant" -%}
118
+ {{- "<|content_model_end_sampling|>" -}}
119
+ {%- endif -%}
120
+ {%- endif -%}
121
+ {%- endfor -%}
122
+
123
+ {%- if not state.effort_emitted -%}
124
+ {{- emit_thinking_effort() -}}
125
+ {%- endif -%}
126
+
127
+ {%- if add_generation_prompt -%}
128
+ {{- "<|message_model|>" -}}
129
+ {%- endif -%}
config.json ADDED
@@ -0,0 +1,2077 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "InklingForConditionalGeneration"
4
+ ],
5
+ "model_type": "inkling_mm_model",
6
+ "eos_token_id": 200006,
7
+ "text_config": {
8
+ "model_max_length": 1048576,
9
+ "torch_dtype": "bfloat16",
10
+ "hidden_size": 4096,
11
+ "num_hidden_layers": 42,
12
+ "vocab_size": 201024,
13
+ "num_attention_heads": 32,
14
+ "num_key_value_heads": 8,
15
+ "head_dim": 128,
16
+ "d_rel": 16,
17
+ "rel_extent": 1024,
18
+ "q_bias": false,
19
+ "o_bias": false,
20
+ "log_scaling_n_floor": 128000,
21
+ "log_scaling_alpha": 0.1,
22
+ "rms_norm_eps": 1e-06,
23
+ "use_embed_norm": true,
24
+ "local_layer_ids": [
25
+ 0,
26
+ 1,
27
+ 2,
28
+ 3,
29
+ 4,
30
+ 6,
31
+ 7,
32
+ 8,
33
+ 9,
34
+ 10,
35
+ 12,
36
+ 13,
37
+ 14,
38
+ 15,
39
+ 16,
40
+ 18,
41
+ 19,
42
+ 20,
43
+ 21,
44
+ 22,
45
+ 24,
46
+ 25,
47
+ 26,
48
+ 27,
49
+ 28,
50
+ 30,
51
+ 31,
52
+ 32,
53
+ 33,
54
+ 34,
55
+ 36,
56
+ 37,
57
+ 38,
58
+ 39,
59
+ 40
60
+ ],
61
+ "dense_mlp_idx": 2,
62
+ "use_sconv": true,
63
+ "sconv_kernel_size": 4,
64
+ "unpadded_vocab_size": 200058,
65
+ "logits_mup_width_multiplier": 16.0,
66
+ "final_logit_softcapping": null,
67
+ "swa_head_dim": 128,
68
+ "swa_num_attention_heads": 32,
69
+ "swa_num_key_value_heads": 8,
70
+ "sliding_window_size": 512,
71
+ "n_routed_experts": 256,
72
+ "num_experts_per_tok": 6,
73
+ "n_shared_experts": 2,
74
+ "shared_expert_sink": true,
75
+ "dense_intermediate_size": 16384,
76
+ "intermediate_size": 2048,
77
+ "route_scale": 8.0,
78
+ "use_gate_bias": true,
79
+ "gate_activation": "sigmoid",
80
+ "norm_after_topk": true,
81
+ "use_global_scale": true
82
+ },
83
+ "audio_config": {
84
+ "decoder_dmodel": 4096,
85
+ "n_mel_bins": 80,
86
+ "mel_vocab_size": 16,
87
+ "bias": false,
88
+ "dmel_min_value": -7.0,
89
+ "dmel_max_value": 2.0,
90
+ "use_audio_norm": true,
91
+ "audio_mode": "dmel"
92
+ },
93
+ "vision_config": {
94
+ "vision_encoder_type": "hmlp",
95
+ "decoder_dmodel": 4096,
96
+ "patch_size": 40,
97
+ "temporal_patch_size": 2,
98
+ "n_channels": 3,
99
+ "n_layers": 4,
100
+ "use_vision_norm": true
101
+ },
102
+ "mtp_config": {
103
+ "num_nextn_predict_layers": 8,
104
+ "chain_hidden_post_norm": false,
105
+ "local_layer_ids": [
106
+ 0,
107
+ 2,
108
+ 4,
109
+ 5,
110
+ 6,
111
+ 7
112
+ ]
113
+ },
114
+ "quantization_config": {
115
+ "quant_method": "auto-round",
116
+ "packing_format": "auto_round:llm_compressor",
117
+ "bits": 4,
118
+ "group_size": 32,
119
+ "sym": true,
120
+ "data_type": "mx_fp",
121
+ "iters": 0,
122
+ "model_free": true,
123
+ "autoround_version": "0.15.0",
124
+ "enable_quanted_input": false,
125
+ "block_name_to_quantize": "model.llm.layers",
126
+ "act_bits": 16,
127
+ "act_data_type": "fp",
128
+ "act_dynamic": true,
129
+ "act_group_size": 128,
130
+ "act_sym": true,
131
+ "extra_config": {
132
+ "mlp.experts": {
133
+ "act_bits": 4,
134
+ "act_group_size": 32,
135
+ "act_data_type": "mx_fp"
136
+ },
137
+ "model.audio.encoder": {
138
+ "bits": 16,
139
+ "data_type": "float",
140
+ "act_bits": 16,
141
+ "act_data_type": "float"
142
+ },
143
+ "model.llm.layers.13.attn.wr_du": {
144
+ "bits": 16,
145
+ "data_type": "float",
146
+ "act_bits": 16,
147
+ "act_data_type": "float"
148
+ },
149
+ "model.llm.layers.3.mlp.gate": {
150
+ "bits": 16,
151
+ "data_type": "float",
152
+ "act_bits": 16,
153
+ "act_data_type": "float"
154
+ },
155
+ "model.llm.layers.34.attn.wk_dv": {
156
+ "bits": 16,
157
+ "data_type": "float",
158
+ "act_bits": 16,
159
+ "act_data_type": "float"
160
+ },
161
+ "model.llm.layers.37.attn.wk_dv": {
162
+ "bits": 16,
163
+ "data_type": "float",
164
+ "act_bits": 16,
165
+ "act_data_type": "float"
166
+ },
167
+ "model.llm.layers.37.attn.wo_ud": {
168
+ "bits": 16,
169
+ "data_type": "float",
170
+ "act_bits": 16,
171
+ "act_data_type": "float"
172
+ },
173
+ "model.llm.layers.11.mlp.gate": {
174
+ "bits": 16,
175
+ "data_type": "float",
176
+ "act_bits": 16,
177
+ "act_data_type": "float"
178
+ },
179
+ "model.llm.layers.13.attn.wk_dv": {
180
+ "bits": 16,
181
+ "data_type": "float",
182
+ "act_bits": 16,
183
+ "act_data_type": "float"
184
+ },
185
+ "model.llm.layers.18.attn.wq_du": {
186
+ "bits": 16,
187
+ "data_type": "float",
188
+ "act_bits": 16,
189
+ "act_data_type": "float"
190
+ },
191
+ "model.llm.layers.19.attn.wo_ud": {
192
+ "bits": 16,
193
+ "data_type": "float",
194
+ "act_bits": 16,
195
+ "act_data_type": "float"
196
+ },
197
+ "model.llm.layers.2.attn.wv_dv": {
198
+ "bits": 16,
199
+ "data_type": "float",
200
+ "act_bits": 16,
201
+ "act_data_type": "float"
202
+ },
203
+ "model.llm.layers.2.mlp.gate": {
204
+ "bits": 16,
205
+ "data_type": "float",
206
+ "act_bits": 16,
207
+ "act_data_type": "float"
208
+ },
209
+ "model.llm.layers.20.attn.wk_dv": {
210
+ "bits": 16,
211
+ "data_type": "float",
212
+ "act_bits": 16,
213
+ "act_data_type": "float"
214
+ },
215
+ "model.llm.layers.21.attn.wk_dv": {
216
+ "bits": 16,
217
+ "data_type": "float",
218
+ "act_bits": 16,
219
+ "act_data_type": "float"
220
+ },
221
+ "model.llm.layers.22.attn.wq_du": {
222
+ "bits": 16,
223
+ "data_type": "float",
224
+ "act_bits": 16,
225
+ "act_data_type": "float"
226
+ },
227
+ "model.llm.layers.23.attn.wr_du": {
228
+ "bits": 16,
229
+ "data_type": "float",
230
+ "act_bits": 16,
231
+ "act_data_type": "float"
232
+ },
233
+ "model.llm.layers.25.attn.wo_ud": {
234
+ "bits": 16,
235
+ "data_type": "float",
236
+ "act_bits": 16,
237
+ "act_data_type": "float"
238
+ },
239
+ "model.llm.layers.27.attn.wk_dv": {
240
+ "bits": 16,
241
+ "data_type": "float",
242
+ "act_bits": 16,
243
+ "act_data_type": "float"
244
+ },
245
+ "model.llm.layers.29.attn.wv_dv": {
246
+ "bits": 16,
247
+ "data_type": "float",
248
+ "act_bits": 16,
249
+ "act_data_type": "float"
250
+ },
251
+ "model.llm.layers.3.attn.wo_ud": {
252
+ "bits": 16,
253
+ "data_type": "float",
254
+ "act_bits": 16,
255
+ "act_data_type": "float"
256
+ },
257
+ "model.llm.layers.3.attn.wr_du": {
258
+ "bits": 16,
259
+ "data_type": "float",
260
+ "act_bits": 16,
261
+ "act_data_type": "float"
262
+ },
263
+ "model.llm.layers.32.attn.wo_ud": {
264
+ "bits": 16,
265
+ "data_type": "float",
266
+ "act_bits": 16,
267
+ "act_data_type": "float"
268
+ },
269
+ "model.llm.layers.32.attn.wq_du": {
270
+ "bits": 16,
271
+ "data_type": "float",
272
+ "act_bits": 16,
273
+ "act_data_type": "float"
274
+ },
275
+ "model.llm.layers.37.attn.wr_du": {
276
+ "bits": 16,
277
+ "data_type": "float",
278
+ "act_bits": 16,
279
+ "act_data_type": "float"
280
+ },
281
+ "model.llm.layers.4.attn.wk_dv": {
282
+ "bits": 16,
283
+ "data_type": "float",
284
+ "act_bits": 16,
285
+ "act_data_type": "float"
286
+ },
287
+ "model.llm.layers.4.attn.wq_du": {
288
+ "bits": 16,
289
+ "data_type": "float",
290
+ "act_bits": 16,
291
+ "act_data_type": "float"
292
+ },
293
+ "model.llm.layers.40.attn.wr_du": {
294
+ "bits": 16,
295
+ "data_type": "float",
296
+ "act_bits": 16,
297
+ "act_data_type": "float"
298
+ },
299
+ "model.llm.layers.15.mlp.gate": {
300
+ "bits": 16,
301
+ "data_type": "float",
302
+ "act_bits": 16,
303
+ "act_data_type": "float"
304
+ },
305
+ "model.llm.layers.23.attn.wk_dv": {
306
+ "bits": 16,
307
+ "data_type": "float",
308
+ "act_bits": 16,
309
+ "act_data_type": "float"
310
+ },
311
+ "model.llm.layers.30.attn.wr_du": {
312
+ "bits": 16,
313
+ "data_type": "float",
314
+ "act_bits": 16,
315
+ "act_data_type": "float"
316
+ },
317
+ "model.llm.layers.34.attn.wv_dv": {
318
+ "bits": 16,
319
+ "data_type": "float",
320
+ "act_bits": 16,
321
+ "act_data_type": "float"
322
+ },
323
+ "model.llm.layers.36.attn.wr_du": {
324
+ "bits": 16,
325
+ "data_type": "float",
326
+ "act_bits": 16,
327
+ "act_data_type": "float"
328
+ },
329
+ "model.llm.layers.4.mlp.gate": {
330
+ "bits": 16,
331
+ "data_type": "float",
332
+ "act_bits": 16,
333
+ "act_data_type": "float"
334
+ },
335
+ "model.llm.layers.6.attn.wr_du": {
336
+ "bits": 16,
337
+ "data_type": "float",
338
+ "act_bits": 16,
339
+ "act_data_type": "float"
340
+ },
341
+ "model.llm.layers.8.attn.wr_du": {
342
+ "bits": 16,
343
+ "data_type": "float",
344
+ "act_bits": 16,
345
+ "act_data_type": "float"
346
+ },
347
+ "model.llm.layers.9.attn.wr_du": {
348
+ "bits": 16,
349
+ "data_type": "float",
350
+ "act_bits": 16,
351
+ "act_data_type": "float"
352
+ },
353
+ "model.llm.layers.18.mlp.gate": {
354
+ "bits": 16,
355
+ "data_type": "float",
356
+ "act_bits": 16,
357
+ "act_data_type": "float"
358
+ },
359
+ "model.llm.layers.27.mlp.gate": {
360
+ "bits": 16,
361
+ "data_type": "float",
362
+ "act_bits": 16,
363
+ "act_data_type": "float"
364
+ },
365
+ "model.llm.layers.30.attn.wq_du": {
366
+ "bits": 16,
367
+ "data_type": "float",
368
+ "act_bits": 16,
369
+ "act_data_type": "float"
370
+ },
371
+ "model.llm.layers.0.attn.wq_du": {
372
+ "bits": 16,
373
+ "data_type": "float",
374
+ "act_bits": 16,
375
+ "act_data_type": "float"
376
+ },
377
+ "model.llm.layers.0.mlp.w13_dn": {
378
+ "bits": 16,
379
+ "data_type": "float",
380
+ "act_bits": 16,
381
+ "act_data_type": "float"
382
+ },
383
+ "model.llm.layers.14.mlp.gate": {
384
+ "bits": 16,
385
+ "data_type": "float",
386
+ "act_bits": 16,
387
+ "act_data_type": "float"
388
+ },
389
+ "model.llm.layers.20.attn.wv_dv": {
390
+ "bits": 16,
391
+ "data_type": "float",
392
+ "act_bits": 16,
393
+ "act_data_type": "float"
394
+ },
395
+ "model.llm.layers.23.attn.wo_ud": {
396
+ "bits": 16,
397
+ "data_type": "float",
398
+ "act_bits": 16,
399
+ "act_data_type": "float"
400
+ },
401
+ "model.llm.layers.23.attn.wq_du": {
402
+ "bits": 16,
403
+ "data_type": "float",
404
+ "act_bits": 16,
405
+ "act_data_type": "float"
406
+ },
407
+ "model.llm.layers.23.attn.wv_dv": {
408
+ "bits": 16,
409
+ "data_type": "float",
410
+ "act_bits": 16,
411
+ "act_data_type": "float"
412
+ },
413
+ "model.llm.layers.24.attn.wq_du": {
414
+ "bits": 16,
415
+ "data_type": "float",
416
+ "act_bits": 16,
417
+ "act_data_type": "float"
418
+ },
419
+ "model.llm.layers.26.mlp.gate": {
420
+ "bits": 16,
421
+ "data_type": "float",
422
+ "act_bits": 16,
423
+ "act_data_type": "float"
424
+ },
425
+ "model.llm.layers.27.attn.wq_du": {
426
+ "bits": 16,
427
+ "data_type": "float",
428
+ "act_bits": 16,
429
+ "act_data_type": "float"
430
+ },
431
+ "model.llm.layers.35.attn.wk_dv": {
432
+ "bits": 16,
433
+ "data_type": "float",
434
+ "act_bits": 16,
435
+ "act_data_type": "float"
436
+ },
437
+ "model.llm.layers.36.attn.wo_ud": {
438
+ "bits": 16,
439
+ "data_type": "float",
440
+ "act_bits": 16,
441
+ "act_data_type": "float"
442
+ },
443
+ "model.llm.layers.37.mlp.gate": {
444
+ "bits": 16,
445
+ "data_type": "float",
446
+ "act_bits": 16,
447
+ "act_data_type": "float"
448
+ },
449
+ "model.llm.layers.41.mlp.gate": {
450
+ "bits": 16,
451
+ "data_type": "float",
452
+ "act_bits": 16,
453
+ "act_data_type": "float"
454
+ },
455
+ "model.llm.layers.5.mlp.gate": {
456
+ "bits": 16,
457
+ "data_type": "float",
458
+ "act_bits": 16,
459
+ "act_data_type": "float"
460
+ },
461
+ "model.llm.layers.7.attn.wr_du": {
462
+ "bits": 16,
463
+ "data_type": "float",
464
+ "act_bits": 16,
465
+ "act_data_type": "float"
466
+ },
467
+ "model.visual.layers.linear_2": {
468
+ "bits": 16,
469
+ "data_type": "float",
470
+ "act_bits": 16,
471
+ "act_data_type": "float"
472
+ },
473
+ "model.llm.layers.12.attn.wo_ud": {
474
+ "bits": 16,
475
+ "data_type": "float",
476
+ "act_bits": 16,
477
+ "act_data_type": "float"
478
+ },
479
+ "model.llm.layers.14.attn.wq_du": {
480
+ "bits": 16,
481
+ "data_type": "float",
482
+ "act_bits": 16,
483
+ "act_data_type": "float"
484
+ },
485
+ "model.llm.layers.28.attn.wq_du": {
486
+ "bits": 16,
487
+ "data_type": "float",
488
+ "act_bits": 16,
489
+ "act_data_type": "float"
490
+ },
491
+ "model.llm.layers.31.attn.wr_du": {
492
+ "bits": 16,
493
+ "data_type": "float",
494
+ "act_bits": 16,
495
+ "act_data_type": "float"
496
+ },
497
+ "model.llm.layers.35.attn.wr_du": {
498
+ "bits": 16,
499
+ "data_type": "float",
500
+ "act_bits": 16,
501
+ "act_data_type": "float"
502
+ },
503
+ "model.llm.layers.38.attn.wq_du": {
504
+ "bits": 16,
505
+ "data_type": "float",
506
+ "act_bits": 16,
507
+ "act_data_type": "float"
508
+ },
509
+ "model.llm.layers.8.attn.wo_ud": {
510
+ "bits": 16,
511
+ "data_type": "float",
512
+ "act_bits": 16,
513
+ "act_data_type": "float"
514
+ },
515
+ "model.llm.layers.9.attn.wk_dv": {
516
+ "bits": 16,
517
+ "data_type": "float",
518
+ "act_bits": 16,
519
+ "act_data_type": "float"
520
+ },
521
+ "model.llm.layers.11.attn.wk_dv": {
522
+ "bits": 16,
523
+ "data_type": "float",
524
+ "act_bits": 16,
525
+ "act_data_type": "float"
526
+ },
527
+ "model.llm.layers.12.attn.wq_du": {
528
+ "bits": 16,
529
+ "data_type": "float",
530
+ "act_bits": 16,
531
+ "act_data_type": "float"
532
+ },
533
+ "model.llm.layers.14.attn.wo_ud": {
534
+ "bits": 16,
535
+ "data_type": "float",
536
+ "act_bits": 16,
537
+ "act_data_type": "float"
538
+ },
539
+ "model.llm.layers.15.attn.wk_dv": {
540
+ "bits": 16,
541
+ "data_type": "float",
542
+ "act_bits": 16,
543
+ "act_data_type": "float"
544
+ },
545
+ "model.llm.layers.19.mlp.gate": {
546
+ "bits": 16,
547
+ "data_type": "float",
548
+ "act_bits": 16,
549
+ "act_data_type": "float"
550
+ },
551
+ "model.llm.layers.2.attn.wo_ud": {
552
+ "bits": 16,
553
+ "data_type": "float",
554
+ "act_bits": 16,
555
+ "act_data_type": "float"
556
+ },
557
+ "model.llm.layers.24.mlp.gate": {
558
+ "bits": 16,
559
+ "data_type": "float",
560
+ "act_bits": 16,
561
+ "act_data_type": "float"
562
+ },
563
+ "model.llm.layers.28.attn.wo_ud": {
564
+ "bits": 16,
565
+ "data_type": "float",
566
+ "act_bits": 16,
567
+ "act_data_type": "float"
568
+ },
569
+ "model.llm.layers.3.attn.wk_dv": {
570
+ "bits": 16,
571
+ "data_type": "float",
572
+ "act_bits": 16,
573
+ "act_data_type": "float"
574
+ },
575
+ "model.llm.layers.33.attn.wr_du": {
576
+ "bits": 16,
577
+ "data_type": "float",
578
+ "act_bits": 16,
579
+ "act_data_type": "float"
580
+ },
581
+ "model.llm.layers.34.attn.wr_du": {
582
+ "bits": 16,
583
+ "data_type": "float",
584
+ "act_bits": 16,
585
+ "act_data_type": "float"
586
+ },
587
+ "model.llm.layers.35.mlp.gate": {
588
+ "bits": 16,
589
+ "data_type": "float",
590
+ "act_bits": 16,
591
+ "act_data_type": "float"
592
+ },
593
+ "model.llm.layers.6.attn.wo_ud": {
594
+ "bits": 16,
595
+ "data_type": "float",
596
+ "act_bits": 16,
597
+ "act_data_type": "float"
598
+ },
599
+ "model.llm.layers.0.attn.wk_dv": {
600
+ "bits": 16,
601
+ "data_type": "float",
602
+ "act_bits": 16,
603
+ "act_data_type": "float"
604
+ },
605
+ "model.llm.layers.10.attn.wr_du": {
606
+ "bits": 16,
607
+ "data_type": "float",
608
+ "act_bits": 16,
609
+ "act_data_type": "float"
610
+ },
611
+ "model.llm.layers.13.attn.wv_dv": {
612
+ "bits": 16,
613
+ "data_type": "float",
614
+ "act_bits": 16,
615
+ "act_data_type": "float"
616
+ },
617
+ "model.llm.layers.15.attn.wr_du": {
618
+ "bits": 16,
619
+ "data_type": "float",
620
+ "act_bits": 16,
621
+ "act_data_type": "float"
622
+ },
623
+ "model.llm.layers.17.attn.wo_ud": {
624
+ "bits": 16,
625
+ "data_type": "float",
626
+ "act_bits": 16,
627
+ "act_data_type": "float"
628
+ },
629
+ "model.llm.layers.20.attn.wr_du": {
630
+ "bits": 16,
631
+ "data_type": "float",
632
+ "act_bits": 16,
633
+ "act_data_type": "float"
634
+ },
635
+ "model.llm.layers.36.attn.wv_dv": {
636
+ "bits": 16,
637
+ "data_type": "float",
638
+ "act_bits": 16,
639
+ "act_data_type": "float"
640
+ },
641
+ "model.llm.layers.38.attn.wk_dv": {
642
+ "bits": 16,
643
+ "data_type": "float",
644
+ "act_bits": 16,
645
+ "act_data_type": "float"
646
+ },
647
+ "model.llm.layers.41.attn.wo_ud": {
648
+ "bits": 16,
649
+ "data_type": "float",
650
+ "act_bits": 16,
651
+ "act_data_type": "float"
652
+ },
653
+ "model.llm.layers.9.attn.wo_ud": {
654
+ "bits": 16,
655
+ "data_type": "float",
656
+ "act_bits": 16,
657
+ "act_data_type": "float"
658
+ },
659
+ "model.llm.layers.9.attn.wv_dv": {
660
+ "bits": 16,
661
+ "data_type": "float",
662
+ "act_bits": 16,
663
+ "act_data_type": "float"
664
+ },
665
+ "model.llm.layers.0.attn.wo_ud": {
666
+ "bits": 16,
667
+ "data_type": "float",
668
+ "act_bits": 16,
669
+ "act_data_type": "float"
670
+ },
671
+ "model.llm.layers.10.attn.wq_du": {
672
+ "bits": 16,
673
+ "data_type": "float",
674
+ "act_bits": 16,
675
+ "act_data_type": "float"
676
+ },
677
+ "model.llm.layers.10.mlp.gate": {
678
+ "bits": 16,
679
+ "data_type": "float",
680
+ "act_bits": 16,
681
+ "act_data_type": "float"
682
+ },
683
+ "model.llm.layers.19.attn.wv_dv": {
684
+ "bits": 16,
685
+ "data_type": "float",
686
+ "act_bits": 16,
687
+ "act_data_type": "float"
688
+ },
689
+ "model.llm.layers.21.attn.wo_ud": {
690
+ "bits": 16,
691
+ "data_type": "float",
692
+ "act_bits": 16,
693
+ "act_data_type": "float"
694
+ },
695
+ "model.llm.layers.28.attn.wr_du": {
696
+ "bits": 16,
697
+ "data_type": "float",
698
+ "act_bits": 16,
699
+ "act_data_type": "float"
700
+ },
701
+ "model.llm.layers.30.attn.wo_ud": {
702
+ "bits": 16,
703
+ "data_type": "float",
704
+ "act_bits": 16,
705
+ "act_data_type": "float"
706
+ },
707
+ "model.llm.layers.35.attn.wq_du": {
708
+ "bits": 16,
709
+ "data_type": "float",
710
+ "act_bits": 16,
711
+ "act_data_type": "float"
712
+ },
713
+ "model.llm.layers.40.attn.wo_ud": {
714
+ "bits": 16,
715
+ "data_type": "float",
716
+ "act_bits": 16,
717
+ "act_data_type": "float"
718
+ },
719
+ "model.llm.layers.8.attn.wq_du": {
720
+ "bits": 16,
721
+ "data_type": "float",
722
+ "act_bits": 16,
723
+ "act_data_type": "float"
724
+ },
725
+ "model.llm.layers.0.attn.wr_du": {
726
+ "bits": 16,
727
+ "data_type": "float",
728
+ "act_bits": 16,
729
+ "act_data_type": "float"
730
+ },
731
+ "model.llm.layers.0.attn.wv_dv": {
732
+ "bits": 16,
733
+ "data_type": "float",
734
+ "act_bits": 16,
735
+ "act_data_type": "float"
736
+ },
737
+ "model.llm.layers.1.attn.wo_ud": {
738
+ "bits": 16,
739
+ "data_type": "float",
740
+ "act_bits": 16,
741
+ "act_data_type": "float"
742
+ },
743
+ "model.llm.layers.1.attn.wv_dv": {
744
+ "bits": 16,
745
+ "data_type": "float",
746
+ "act_bits": 16,
747
+ "act_data_type": "float"
748
+ },
749
+ "model.llm.layers.16.attn.wr_du": {
750
+ "bits": 16,
751
+ "data_type": "float",
752
+ "act_bits": 16,
753
+ "act_data_type": "float"
754
+ },
755
+ "model.llm.layers.21.attn.wq_du": {
756
+ "bits": 16,
757
+ "data_type": "float",
758
+ "act_bits": 16,
759
+ "act_data_type": "float"
760
+ },
761
+ "model.llm.layers.21.attn.wv_dv": {
762
+ "bits": 16,
763
+ "data_type": "float",
764
+ "act_bits": 16,
765
+ "act_data_type": "float"
766
+ },
767
+ "model.llm.layers.23.mlp.gate": {
768
+ "bits": 16,
769
+ "data_type": "float",
770
+ "act_bits": 16,
771
+ "act_data_type": "float"
772
+ },
773
+ "model.llm.layers.24.attn.wv_dv": {
774
+ "bits": 16,
775
+ "data_type": "float",
776
+ "act_bits": 16,
777
+ "act_data_type": "float"
778
+ },
779
+ "model.llm.layers.25.attn.wq_du": {
780
+ "bits": 16,
781
+ "data_type": "float",
782
+ "act_bits": 16,
783
+ "act_data_type": "float"
784
+ },
785
+ "model.llm.layers.26.attn.wk_dv": {
786
+ "bits": 16,
787
+ "data_type": "float",
788
+ "act_bits": 16,
789
+ "act_data_type": "float"
790
+ },
791
+ "model.llm.layers.27.attn.wo_ud": {
792
+ "bits": 16,
793
+ "data_type": "float",
794
+ "act_bits": 16,
795
+ "act_data_type": "float"
796
+ },
797
+ "model.llm.layers.29.attn.wq_du": {
798
+ "bits": 16,
799
+ "data_type": "float",
800
+ "act_bits": 16,
801
+ "act_data_type": "float"
802
+ },
803
+ "model.llm.layers.33.attn.wv_dv": {
804
+ "bits": 16,
805
+ "data_type": "float",
806
+ "act_bits": 16,
807
+ "act_data_type": "float"
808
+ },
809
+ "model.llm.layers.36.attn.wk_dv": {
810
+ "bits": 16,
811
+ "data_type": "float",
812
+ "act_bits": 16,
813
+ "act_data_type": "float"
814
+ },
815
+ "model.llm.layers.39.attn.wr_du": {
816
+ "bits": 16,
817
+ "data_type": "float",
818
+ "act_bits": 16,
819
+ "act_data_type": "float"
820
+ },
821
+ "model.llm.layers.13.attn.wq_du": {
822
+ "bits": 16,
823
+ "data_type": "float",
824
+ "act_bits": 16,
825
+ "act_data_type": "float"
826
+ },
827
+ "model.llm.layers.15.attn.wq_du": {
828
+ "bits": 16,
829
+ "data_type": "float",
830
+ "act_bits": 16,
831
+ "act_data_type": "float"
832
+ },
833
+ "model.llm.layers.16.attn.wq_du": {
834
+ "bits": 16,
835
+ "data_type": "float",
836
+ "act_bits": 16,
837
+ "act_data_type": "float"
838
+ },
839
+ "model.llm.layers.18.attn.wk_dv": {
840
+ "bits": 16,
841
+ "data_type": "float",
842
+ "act_bits": 16,
843
+ "act_data_type": "float"
844
+ },
845
+ "model.llm.layers.20.attn.wo_ud": {
846
+ "bits": 16,
847
+ "data_type": "float",
848
+ "act_bits": 16,
849
+ "act_data_type": "float"
850
+ },
851
+ "model.llm.layers.22.attn.wo_ud": {
852
+ "bits": 16,
853
+ "data_type": "float",
854
+ "act_bits": 16,
855
+ "act_data_type": "float"
856
+ },
857
+ "model.llm.layers.27.attn.wr_du": {
858
+ "bits": 16,
859
+ "data_type": "float",
860
+ "act_bits": 16,
861
+ "act_data_type": "float"
862
+ },
863
+ "model.llm.layers.29.mlp.gate": {
864
+ "bits": 16,
865
+ "data_type": "float",
866
+ "act_bits": 16,
867
+ "act_data_type": "float"
868
+ },
869
+ "model.llm.layers.16.attn.wv_dv": {
870
+ "bits": 16,
871
+ "data_type": "float",
872
+ "act_bits": 16,
873
+ "act_data_type": "float"
874
+ },
875
+ "model.llm.layers.17.attn.wq_du": {
876
+ "bits": 16,
877
+ "data_type": "float",
878
+ "act_bits": 16,
879
+ "act_data_type": "float"
880
+ },
881
+ "model.llm.layers.19.attn.wq_du": {
882
+ "bits": 16,
883
+ "data_type": "float",
884
+ "act_bits": 16,
885
+ "act_data_type": "float"
886
+ },
887
+ "model.llm.layers.27.attn.wv_dv": {
888
+ "bits": 16,
889
+ "data_type": "float",
890
+ "act_bits": 16,
891
+ "act_data_type": "float"
892
+ },
893
+ "model.llm.layers.29.attn.wo_ud": {
894
+ "bits": 16,
895
+ "data_type": "float",
896
+ "act_bits": 16,
897
+ "act_data_type": "float"
898
+ },
899
+ "model.llm.layers.38.attn.wv_dv": {
900
+ "bits": 16,
901
+ "data_type": "float",
902
+ "act_bits": 16,
903
+ "act_data_type": "float"
904
+ },
905
+ "model.llm.layers.39.attn.wk_dv": {
906
+ "bits": 16,
907
+ "data_type": "float",
908
+ "act_bits": 16,
909
+ "act_data_type": "float"
910
+ },
911
+ "model.llm.layers.4.attn.wo_ud": {
912
+ "bits": 16,
913
+ "data_type": "float",
914
+ "act_bits": 16,
915
+ "act_data_type": "float"
916
+ },
917
+ "model.llm.layers.14.attn.wv_dv": {
918
+ "bits": 16,
919
+ "data_type": "float",
920
+ "act_bits": 16,
921
+ "act_data_type": "float"
922
+ },
923
+ "model.llm.layers.16.attn.wk_dv": {
924
+ "bits": 16,
925
+ "data_type": "float",
926
+ "act_bits": 16,
927
+ "act_data_type": "float"
928
+ },
929
+ "model.llm.layers.18.attn.wo_ud": {
930
+ "bits": 16,
931
+ "data_type": "float",
932
+ "act_bits": 16,
933
+ "act_data_type": "float"
934
+ },
935
+ "model.llm.layers.22.attn.wv_dv": {
936
+ "bits": 16,
937
+ "data_type": "float",
938
+ "act_bits": 16,
939
+ "act_data_type": "float"
940
+ },
941
+ "model.llm.layers.28.mlp.gate": {
942
+ "bits": 16,
943
+ "data_type": "float",
944
+ "act_bits": 16,
945
+ "act_data_type": "float"
946
+ },
947
+ "model.llm.layers.31.mlp.gate": {
948
+ "bits": 16,
949
+ "data_type": "float",
950
+ "act_bits": 16,
951
+ "act_data_type": "float"
952
+ },
953
+ "model.llm.layers.40.attn.wk_dv": {
954
+ "bits": 16,
955
+ "data_type": "float",
956
+ "act_bits": 16,
957
+ "act_data_type": "float"
958
+ },
959
+ "model.llm.layers.5.attn.wv_dv": {
960
+ "bits": 16,
961
+ "data_type": "float",
962
+ "act_bits": 16,
963
+ "act_data_type": "float"
964
+ },
965
+ "model.llm.layers.7.attn.wo_ud": {
966
+ "bits": 16,
967
+ "data_type": "float",
968
+ "act_bits": 16,
969
+ "act_data_type": "float"
970
+ },
971
+ "model.llm.layers.8.mlp.gate": {
972
+ "bits": 16,
973
+ "data_type": "float",
974
+ "act_bits": 16,
975
+ "act_data_type": "float"
976
+ },
977
+ "model.visual.layers.linear_1": {
978
+ "bits": 16,
979
+ "data_type": "float",
980
+ "act_bits": 16,
981
+ "act_data_type": "float"
982
+ },
983
+ "model.visual.layers.linear_3": {
984
+ "bits": 16,
985
+ "data_type": "float",
986
+ "act_bits": 16,
987
+ "act_data_type": "float"
988
+ },
989
+ "model.llm.layers.0.mlp.w2_md": {
990
+ "bits": 16,
991
+ "data_type": "float",
992
+ "act_bits": 16,
993
+ "act_data_type": "float"
994
+ },
995
+ "model.llm.layers.18.attn.wv_dv": {
996
+ "bits": 16,
997
+ "data_type": "float",
998
+ "act_bits": 16,
999
+ "act_data_type": "float"
1000
+ },
1001
+ "model.llm.layers.20.attn.wq_du": {
1002
+ "bits": 16,
1003
+ "data_type": "float",
1004
+ "act_bits": 16,
1005
+ "act_data_type": "float"
1006
+ },
1007
+ "model.llm.layers.24.attn.wr_du": {
1008
+ "bits": 16,
1009
+ "data_type": "float",
1010
+ "act_bits": 16,
1011
+ "act_data_type": "float"
1012
+ },
1013
+ "model.llm.layers.25.attn.wk_dv": {
1014
+ "bits": 16,
1015
+ "data_type": "float",
1016
+ "act_bits": 16,
1017
+ "act_data_type": "float"
1018
+ },
1019
+ "model.llm.layers.3.attn.wq_du": {
1020
+ "bits": 16,
1021
+ "data_type": "float",
1022
+ "act_bits": 16,
1023
+ "act_data_type": "float"
1024
+ },
1025
+ "model.llm.layers.33.attn.wq_du": {
1026
+ "bits": 16,
1027
+ "data_type": "float",
1028
+ "act_bits": 16,
1029
+ "act_data_type": "float"
1030
+ },
1031
+ "model.llm.layers.33.mlp.gate": {
1032
+ "bits": 16,
1033
+ "data_type": "float",
1034
+ "act_bits": 16,
1035
+ "act_data_type": "float"
1036
+ },
1037
+ "model.llm.layers.39.attn.wo_ud": {
1038
+ "bits": 16,
1039
+ "data_type": "float",
1040
+ "act_bits": 16,
1041
+ "act_data_type": "float"
1042
+ },
1043
+ "model.llm.layers.12.attn.wk_dv": {
1044
+ "bits": 16,
1045
+ "data_type": "float",
1046
+ "act_bits": 16,
1047
+ "act_data_type": "float"
1048
+ },
1049
+ "model.llm.layers.13.attn.wo_ud": {
1050
+ "bits": 16,
1051
+ "data_type": "float",
1052
+ "act_bits": 16,
1053
+ "act_data_type": "float"
1054
+ },
1055
+ "model.llm.layers.16.attn.wo_ud": {
1056
+ "bits": 16,
1057
+ "data_type": "float",
1058
+ "act_bits": 16,
1059
+ "act_data_type": "float"
1060
+ },
1061
+ "model.llm.layers.21.mlp.gate": {
1062
+ "bits": 16,
1063
+ "data_type": "float",
1064
+ "act_bits": 16,
1065
+ "act_data_type": "float"
1066
+ },
1067
+ "model.llm.layers.28.attn.wk_dv": {
1068
+ "bits": 16,
1069
+ "data_type": "float",
1070
+ "act_bits": 16,
1071
+ "act_data_type": "float"
1072
+ },
1073
+ "model.llm.layers.29.attn.wr_du": {
1074
+ "bits": 16,
1075
+ "data_type": "float",
1076
+ "act_bits": 16,
1077
+ "act_data_type": "float"
1078
+ },
1079
+ "model.llm.layers.34.attn.wo_ud": {
1080
+ "bits": 16,
1081
+ "data_type": "float",
1082
+ "act_bits": 16,
1083
+ "act_data_type": "float"
1084
+ },
1085
+ "model.llm.layers.38.attn.wo_ud": {
1086
+ "bits": 16,
1087
+ "data_type": "float",
1088
+ "act_bits": 16,
1089
+ "act_data_type": "float"
1090
+ },
1091
+ "model.llm.layers.38.attn.wr_du": {
1092
+ "bits": 16,
1093
+ "data_type": "float",
1094
+ "act_bits": 16,
1095
+ "act_data_type": "float"
1096
+ },
1097
+ "model.llm.layers.4.attn.wr_du": {
1098
+ "bits": 16,
1099
+ "data_type": "float",
1100
+ "act_bits": 16,
1101
+ "act_data_type": "float"
1102
+ },
1103
+ "model.llm.layers.4.attn.wv_dv": {
1104
+ "bits": 16,
1105
+ "data_type": "float",
1106
+ "act_bits": 16,
1107
+ "act_data_type": "float"
1108
+ },
1109
+ "model.llm.layers.7.attn.wv_dv": {
1110
+ "bits": 16,
1111
+ "data_type": "float",
1112
+ "act_bits": 16,
1113
+ "act_data_type": "float"
1114
+ },
1115
+ "model.llm.layers.1.attn.wr_du": {
1116
+ "bits": 16,
1117
+ "data_type": "float",
1118
+ "act_bits": 16,
1119
+ "act_data_type": "float"
1120
+ },
1121
+ "model.llm.layers.11.attn.wo_ud": {
1122
+ "bits": 16,
1123
+ "data_type": "float",
1124
+ "act_bits": 16,
1125
+ "act_data_type": "float"
1126
+ },
1127
+ "model.llm.layers.16.mlp.gate": {
1128
+ "bits": 16,
1129
+ "data_type": "float",
1130
+ "act_bits": 16,
1131
+ "act_data_type": "float"
1132
+ },
1133
+ "model.llm.layers.18.attn.wr_du": {
1134
+ "bits": 16,
1135
+ "data_type": "float",
1136
+ "act_bits": 16,
1137
+ "act_data_type": "float"
1138
+ },
1139
+ "model.llm.layers.26.attn.wo_ud": {
1140
+ "bits": 16,
1141
+ "data_type": "float",
1142
+ "act_bits": 16,
1143
+ "act_data_type": "float"
1144
+ },
1145
+ "model.llm.layers.26.attn.wv_dv": {
1146
+ "bits": 16,
1147
+ "data_type": "float",
1148
+ "act_bits": 16,
1149
+ "act_data_type": "float"
1150
+ },
1151
+ "model.llm.layers.31.attn.wq_du": {
1152
+ "bits": 16,
1153
+ "data_type": "float",
1154
+ "act_bits": 16,
1155
+ "act_data_type": "float"
1156
+ },
1157
+ "model.llm.layers.36.attn.wq_du": {
1158
+ "bits": 16,
1159
+ "data_type": "float",
1160
+ "act_bits": 16,
1161
+ "act_data_type": "float"
1162
+ },
1163
+ "model.llm.layers.37.attn.wq_du": {
1164
+ "bits": 16,
1165
+ "data_type": "float",
1166
+ "act_bits": 16,
1167
+ "act_data_type": "float"
1168
+ },
1169
+ "model.llm.layers.39.attn.wq_du": {
1170
+ "bits": 16,
1171
+ "data_type": "float",
1172
+ "act_bits": 16,
1173
+ "act_data_type": "float"
1174
+ },
1175
+ "model.llm.layers.40.attn.wq_du": {
1176
+ "bits": 16,
1177
+ "data_type": "float",
1178
+ "act_bits": 16,
1179
+ "act_data_type": "float"
1180
+ },
1181
+ "model.llm.layers.40.mlp.gate": {
1182
+ "bits": 16,
1183
+ "data_type": "float",
1184
+ "act_bits": 16,
1185
+ "act_data_type": "float"
1186
+ },
1187
+ "model.llm.layers.5.attn.wr_du": {
1188
+ "bits": 16,
1189
+ "data_type": "float",
1190
+ "act_bits": 16,
1191
+ "act_data_type": "float"
1192
+ },
1193
+ "model.llm.layers.8.attn.wv_dv": {
1194
+ "bits": 16,
1195
+ "data_type": "float",
1196
+ "act_bits": 16,
1197
+ "act_data_type": "float"
1198
+ },
1199
+ "model.llm.layers.9.mlp.gate": {
1200
+ "bits": 16,
1201
+ "data_type": "float",
1202
+ "act_bits": 16,
1203
+ "act_data_type": "float"
1204
+ },
1205
+ "model.llm.layers.1.attn.wk_dv": {
1206
+ "bits": 16,
1207
+ "data_type": "float",
1208
+ "act_bits": 16,
1209
+ "act_data_type": "float"
1210
+ },
1211
+ "model.llm.layers.1.mlp.w2_md": {
1212
+ "bits": 16,
1213
+ "data_type": "float",
1214
+ "act_bits": 16,
1215
+ "act_data_type": "float"
1216
+ },
1217
+ "model.llm.layers.10.attn.wk_dv": {
1218
+ "bits": 16,
1219
+ "data_type": "float",
1220
+ "act_bits": 16,
1221
+ "act_data_type": "float"
1222
+ },
1223
+ "model.llm.layers.17.attn.wr_du": {
1224
+ "bits": 16,
1225
+ "data_type": "float",
1226
+ "act_bits": 16,
1227
+ "act_data_type": "float"
1228
+ },
1229
+ "model.llm.layers.22.attn.wk_dv": {
1230
+ "bits": 16,
1231
+ "data_type": "float",
1232
+ "act_bits": 16,
1233
+ "act_data_type": "float"
1234
+ },
1235
+ "model.llm.layers.38.mlp.gate": {
1236
+ "bits": 16,
1237
+ "data_type": "float",
1238
+ "act_bits": 16,
1239
+ "act_data_type": "float"
1240
+ },
1241
+ "model.llm.layers.41.attn.wr_du": {
1242
+ "bits": 16,
1243
+ "data_type": "float",
1244
+ "act_bits": 16,
1245
+ "act_data_type": "float"
1246
+ },
1247
+ "model.llm.layers.41.attn.wv_dv": {
1248
+ "bits": 16,
1249
+ "data_type": "float",
1250
+ "act_bits": 16,
1251
+ "act_data_type": "float"
1252
+ },
1253
+ "model.llm.layers.8.attn.wk_dv": {
1254
+ "bits": 16,
1255
+ "data_type": "float",
1256
+ "act_bits": 16,
1257
+ "act_data_type": "float"
1258
+ },
1259
+ "model.visual.layers.linear_0": {
1260
+ "bits": 16,
1261
+ "data_type": "float",
1262
+ "act_bits": 16,
1263
+ "act_data_type": "float"
1264
+ },
1265
+ "model.llm.layers.31.attn.wk_dv": {
1266
+ "bits": 16,
1267
+ "data_type": "float",
1268
+ "act_bits": 16,
1269
+ "act_data_type": "float"
1270
+ },
1271
+ "model.llm.layers.31.attn.wv_dv": {
1272
+ "bits": 16,
1273
+ "data_type": "float",
1274
+ "act_bits": 16,
1275
+ "act_data_type": "float"
1276
+ },
1277
+ "model.llm.layers.32.attn.wk_dv": {
1278
+ "bits": 16,
1279
+ "data_type": "float",
1280
+ "act_bits": 16,
1281
+ "act_data_type": "float"
1282
+ },
1283
+ "model.llm.layers.1.attn.wq_du": {
1284
+ "bits": 16,
1285
+ "data_type": "float",
1286
+ "act_bits": 16,
1287
+ "act_data_type": "float"
1288
+ },
1289
+ "model.llm.layers.20.mlp.gate": {
1290
+ "bits": 16,
1291
+ "data_type": "float",
1292
+ "act_bits": 16,
1293
+ "act_data_type": "float"
1294
+ },
1295
+ "model.llm.layers.24.attn.wo_ud": {
1296
+ "bits": 16,
1297
+ "data_type": "float",
1298
+ "act_bits": 16,
1299
+ "act_data_type": "float"
1300
+ },
1301
+ "model.llm.layers.25.mlp.gate": {
1302
+ "bits": 16,
1303
+ "data_type": "float",
1304
+ "act_bits": 16,
1305
+ "act_data_type": "float"
1306
+ },
1307
+ "model.llm.layers.17.attn.wk_dv": {
1308
+ "bits": 16,
1309
+ "data_type": "float",
1310
+ "act_bits": 16,
1311
+ "act_data_type": "float"
1312
+ },
1313
+ "model.llm.layers.17.attn.wv_dv": {
1314
+ "bits": 16,
1315
+ "data_type": "float",
1316
+ "act_bits": 16,
1317
+ "act_data_type": "float"
1318
+ },
1319
+ "model.llm.layers.22.attn.wr_du": {
1320
+ "bits": 16,
1321
+ "data_type": "float",
1322
+ "act_bits": 16,
1323
+ "act_data_type": "float"
1324
+ },
1325
+ "model.llm.layers.1.mlp.w13_dn": {
1326
+ "bits": 16,
1327
+ "data_type": "float",
1328
+ "act_bits": 16,
1329
+ "act_data_type": "float"
1330
+ },
1331
+ "model.llm.layers.12.attn.wr_du": {
1332
+ "bits": 16,
1333
+ "data_type": "float",
1334
+ "act_bits": 16,
1335
+ "act_data_type": "float"
1336
+ },
1337
+ "model.llm.layers.25.attn.wr_du": {
1338
+ "bits": 16,
1339
+ "data_type": "float",
1340
+ "act_bits": 16,
1341
+ "act_data_type": "float"
1342
+ },
1343
+ "model.llm.layers.35.attn.wv_dv": {
1344
+ "bits": 16,
1345
+ "data_type": "float",
1346
+ "act_bits": 16,
1347
+ "act_data_type": "float"
1348
+ },
1349
+ "model.llm.layers.36.mlp.gate": {
1350
+ "bits": 16,
1351
+ "data_type": "float",
1352
+ "act_bits": 16,
1353
+ "act_data_type": "float"
1354
+ },
1355
+ "model.llm.layers.10.attn.wo_ud": {
1356
+ "bits": 16,
1357
+ "data_type": "float",
1358
+ "act_bits": 16,
1359
+ "act_data_type": "float"
1360
+ },
1361
+ "model.llm.layers.14.attn.wr_du": {
1362
+ "bits": 16,
1363
+ "data_type": "float",
1364
+ "act_bits": 16,
1365
+ "act_data_type": "float"
1366
+ },
1367
+ "model.llm.layers.15.attn.wv_dv": {
1368
+ "bits": 16,
1369
+ "data_type": "float",
1370
+ "act_bits": 16,
1371
+ "act_data_type": "float"
1372
+ },
1373
+ "model.llm.layers.2.attn.wk_dv": {
1374
+ "bits": 16,
1375
+ "data_type": "float",
1376
+ "act_bits": 16,
1377
+ "act_data_type": "float"
1378
+ },
1379
+ "model.llm.layers.24.attn.wk_dv": {
1380
+ "bits": 16,
1381
+ "data_type": "float",
1382
+ "act_bits": 16,
1383
+ "act_data_type": "float"
1384
+ },
1385
+ "model.llm.layers.25.attn.wv_dv": {
1386
+ "bits": 16,
1387
+ "data_type": "float",
1388
+ "act_bits": 16,
1389
+ "act_data_type": "float"
1390
+ },
1391
+ "model.llm.layers.26.attn.wr_du": {
1392
+ "bits": 16,
1393
+ "data_type": "float",
1394
+ "act_bits": 16,
1395
+ "act_data_type": "float"
1396
+ },
1397
+ "model.llm.layers.29.attn.wk_dv": {
1398
+ "bits": 16,
1399
+ "data_type": "float",
1400
+ "act_bits": 16,
1401
+ "act_data_type": "float"
1402
+ },
1403
+ "model.llm.layers.32.attn.wr_du": {
1404
+ "bits": 16,
1405
+ "data_type": "float",
1406
+ "act_bits": 16,
1407
+ "act_data_type": "float"
1408
+ },
1409
+ "model.llm.layers.32.attn.wv_dv": {
1410
+ "bits": 16,
1411
+ "data_type": "float",
1412
+ "act_bits": 16,
1413
+ "act_data_type": "float"
1414
+ },
1415
+ "model.llm.layers.35.attn.wo_ud": {
1416
+ "bits": 16,
1417
+ "data_type": "float",
1418
+ "act_bits": 16,
1419
+ "act_data_type": "float"
1420
+ },
1421
+ "model.llm.layers.5.attn.wk_dv": {
1422
+ "bits": 16,
1423
+ "data_type": "float",
1424
+ "act_bits": 16,
1425
+ "act_data_type": "float"
1426
+ },
1427
+ "model.llm.layers.6.attn.wk_dv": {
1428
+ "bits": 16,
1429
+ "data_type": "float",
1430
+ "act_bits": 16,
1431
+ "act_data_type": "float"
1432
+ },
1433
+ "model.llm.layers.6.attn.wq_du": {
1434
+ "bits": 16,
1435
+ "data_type": "float",
1436
+ "act_bits": 16,
1437
+ "act_data_type": "float"
1438
+ },
1439
+ "model.llm.layers.6.attn.wv_dv": {
1440
+ "bits": 16,
1441
+ "data_type": "float",
1442
+ "act_bits": 16,
1443
+ "act_data_type": "float"
1444
+ },
1445
+ "model.llm.layers.9.attn.wq_du": {
1446
+ "bits": 16,
1447
+ "data_type": "float",
1448
+ "act_bits": 16,
1449
+ "act_data_type": "float"
1450
+ },
1451
+ "model.llm.layers.10.attn.wv_dv": {
1452
+ "bits": 16,
1453
+ "data_type": "float",
1454
+ "act_bits": 16,
1455
+ "act_data_type": "float"
1456
+ },
1457
+ "model.llm.layers.11.attn.wr_du": {
1458
+ "bits": 16,
1459
+ "data_type": "float",
1460
+ "act_bits": 16,
1461
+ "act_data_type": "float"
1462
+ },
1463
+ "model.llm.layers.12.attn.wv_dv": {
1464
+ "bits": 16,
1465
+ "data_type": "float",
1466
+ "act_bits": 16,
1467
+ "act_data_type": "float"
1468
+ },
1469
+ "model.llm.layers.21.attn.wr_du": {
1470
+ "bits": 16,
1471
+ "data_type": "float",
1472
+ "act_bits": 16,
1473
+ "act_data_type": "float"
1474
+ },
1475
+ "model.llm.layers.22.mlp.gate": {
1476
+ "bits": 16,
1477
+ "data_type": "float",
1478
+ "act_bits": 16,
1479
+ "act_data_type": "float"
1480
+ },
1481
+ "model.llm.layers.3.attn.wv_dv": {
1482
+ "bits": 16,
1483
+ "data_type": "float",
1484
+ "act_bits": 16,
1485
+ "act_data_type": "float"
1486
+ },
1487
+ "model.llm.layers.6.mlp.gate": {
1488
+ "bits": 16,
1489
+ "data_type": "float",
1490
+ "act_bits": 16,
1491
+ "act_data_type": "float"
1492
+ },
1493
+ "model.llm.layers.7.attn.wk_dv": {
1494
+ "bits": 16,
1495
+ "data_type": "float",
1496
+ "act_bits": 16,
1497
+ "act_data_type": "float"
1498
+ },
1499
+ "model.llm.layers.11.attn.wv_dv": {
1500
+ "bits": 16,
1501
+ "data_type": "float",
1502
+ "act_bits": 16,
1503
+ "act_data_type": "float"
1504
+ },
1505
+ "model.llm.layers.12.mlp.gate": {
1506
+ "bits": 16,
1507
+ "data_type": "float",
1508
+ "act_bits": 16,
1509
+ "act_data_type": "float"
1510
+ },
1511
+ "model.llm.layers.17.mlp.gate": {
1512
+ "bits": 16,
1513
+ "data_type": "float",
1514
+ "act_bits": 16,
1515
+ "act_data_type": "float"
1516
+ },
1517
+ "model.llm.layers.19.attn.wk_dv": {
1518
+ "bits": 16,
1519
+ "data_type": "float",
1520
+ "act_bits": 16,
1521
+ "act_data_type": "float"
1522
+ },
1523
+ "model.llm.layers.19.attn.wr_du": {
1524
+ "bits": 16,
1525
+ "data_type": "float",
1526
+ "act_bits": 16,
1527
+ "act_data_type": "float"
1528
+ },
1529
+ "model.llm.layers.28.attn.wv_dv": {
1530
+ "bits": 16,
1531
+ "data_type": "float",
1532
+ "act_bits": 16,
1533
+ "act_data_type": "float"
1534
+ },
1535
+ "model.llm.layers.33.attn.wk_dv": {
1536
+ "bits": 16,
1537
+ "data_type": "float",
1538
+ "act_bits": 16,
1539
+ "act_data_type": "float"
1540
+ },
1541
+ "model.llm.layers.33.attn.wo_ud": {
1542
+ "bits": 16,
1543
+ "data_type": "float",
1544
+ "act_bits": 16,
1545
+ "act_data_type": "float"
1546
+ },
1547
+ "model.llm.layers.40.attn.wv_dv": {
1548
+ "bits": 16,
1549
+ "data_type": "float",
1550
+ "act_bits": 16,
1551
+ "act_data_type": "float"
1552
+ },
1553
+ "model.llm.layers.15.attn.wo_ud": {
1554
+ "bits": 16,
1555
+ "data_type": "float",
1556
+ "act_bits": 16,
1557
+ "act_data_type": "float"
1558
+ },
1559
+ "model.llm.layers.30.mlp.gate": {
1560
+ "bits": 16,
1561
+ "data_type": "float",
1562
+ "act_bits": 16,
1563
+ "act_data_type": "float"
1564
+ },
1565
+ "model.llm.layers.31.attn.wo_ud": {
1566
+ "bits": 16,
1567
+ "data_type": "float",
1568
+ "act_bits": 16,
1569
+ "act_data_type": "float"
1570
+ },
1571
+ "model.mtp.layers.0.input_proj": {
1572
+ "bits": 16,
1573
+ "data_type": "float",
1574
+ "act_bits": 16,
1575
+ "act_data_type": "float"
1576
+ },
1577
+ "model.mtp.layers.0.transformer_block.attn.wk_dv": {
1578
+ "bits": 16,
1579
+ "data_type": "float",
1580
+ "act_bits": 16,
1581
+ "act_data_type": "float"
1582
+ },
1583
+ "model.mtp.layers.0.transformer_block.attn.wo_ud": {
1584
+ "bits": 16,
1585
+ "data_type": "float",
1586
+ "act_bits": 16,
1587
+ "act_data_type": "float"
1588
+ },
1589
+ "model.mtp.layers.0.transformer_block.attn.wq_du": {
1590
+ "bits": 16,
1591
+ "data_type": "float",
1592
+ "act_bits": 16,
1593
+ "act_data_type": "float"
1594
+ },
1595
+ "model.mtp.layers.0.transformer_block.attn.wr_du": {
1596
+ "bits": 16,
1597
+ "data_type": "float",
1598
+ "act_bits": 16,
1599
+ "act_data_type": "float"
1600
+ },
1601
+ "model.mtp.layers.0.transformer_block.attn.wv_dv": {
1602
+ "bits": 16,
1603
+ "data_type": "float",
1604
+ "act_bits": 16,
1605
+ "act_data_type": "float"
1606
+ },
1607
+ "model.mtp.layers.0.transformer_block.mlp.w13_dn": {
1608
+ "bits": 16,
1609
+ "data_type": "float",
1610
+ "act_bits": 16,
1611
+ "act_data_type": "float"
1612
+ },
1613
+ "model.mtp.layers.0.transformer_block.mlp.w2_md": {
1614
+ "bits": 16,
1615
+ "data_type": "float",
1616
+ "act_bits": 16,
1617
+ "act_data_type": "float"
1618
+ },
1619
+ "model.mtp.layers.1.input_proj": {
1620
+ "bits": 16,
1621
+ "data_type": "float",
1622
+ "act_bits": 16,
1623
+ "act_data_type": "float"
1624
+ },
1625
+ "model.mtp.layers.1.transformer_block.attn.wk_dv": {
1626
+ "bits": 16,
1627
+ "data_type": "float",
1628
+ "act_bits": 16,
1629
+ "act_data_type": "float"
1630
+ },
1631
+ "model.mtp.layers.1.transformer_block.attn.wo_ud": {
1632
+ "bits": 16,
1633
+ "data_type": "float",
1634
+ "act_bits": 16,
1635
+ "act_data_type": "float"
1636
+ },
1637
+ "model.mtp.layers.1.transformer_block.attn.wq_du": {
1638
+ "bits": 16,
1639
+ "data_type": "float",
1640
+ "act_bits": 16,
1641
+ "act_data_type": "float"
1642
+ },
1643
+ "model.mtp.layers.1.transformer_block.attn.wr_du": {
1644
+ "bits": 16,
1645
+ "data_type": "float",
1646
+ "act_bits": 16,
1647
+ "act_data_type": "float"
1648
+ },
1649
+ "model.mtp.layers.1.transformer_block.attn.wv_dv": {
1650
+ "bits": 16,
1651
+ "data_type": "float",
1652
+ "act_bits": 16,
1653
+ "act_data_type": "float"
1654
+ },
1655
+ "model.mtp.layers.1.transformer_block.mlp.w13_dn": {
1656
+ "bits": 16,
1657
+ "data_type": "float",
1658
+ "act_bits": 16,
1659
+ "act_data_type": "float"
1660
+ },
1661
+ "model.mtp.layers.1.transformer_block.mlp.w2_md": {
1662
+ "bits": 16,
1663
+ "data_type": "float",
1664
+ "act_bits": 16,
1665
+ "act_data_type": "float"
1666
+ },
1667
+ "model.mtp.layers.2.input_proj": {
1668
+ "bits": 16,
1669
+ "data_type": "float",
1670
+ "act_bits": 16,
1671
+ "act_data_type": "float"
1672
+ },
1673
+ "model.mtp.layers.2.transformer_block.attn.wk_dv": {
1674
+ "bits": 16,
1675
+ "data_type": "float",
1676
+ "act_bits": 16,
1677
+ "act_data_type": "float"
1678
+ },
1679
+ "model.mtp.layers.2.transformer_block.attn.wo_ud": {
1680
+ "bits": 16,
1681
+ "data_type": "float",
1682
+ "act_bits": 16,
1683
+ "act_data_type": "float"
1684
+ },
1685
+ "model.mtp.layers.2.transformer_block.attn.wq_du": {
1686
+ "bits": 16,
1687
+ "data_type": "float",
1688
+ "act_bits": 16,
1689
+ "act_data_type": "float"
1690
+ },
1691
+ "model.mtp.layers.2.transformer_block.attn.wr_du": {
1692
+ "bits": 16,
1693
+ "data_type": "float",
1694
+ "act_bits": 16,
1695
+ "act_data_type": "float"
1696
+ },
1697
+ "model.mtp.layers.2.transformer_block.attn.wv_dv": {
1698
+ "bits": 16,
1699
+ "data_type": "float",
1700
+ "act_bits": 16,
1701
+ "act_data_type": "float"
1702
+ },
1703
+ "model.mtp.layers.2.transformer_block.mlp.w13_dn": {
1704
+ "bits": 16,
1705
+ "data_type": "float",
1706
+ "act_bits": 16,
1707
+ "act_data_type": "float"
1708
+ },
1709
+ "model.mtp.layers.2.transformer_block.mlp.w2_md": {
1710
+ "bits": 16,
1711
+ "data_type": "float",
1712
+ "act_bits": 16,
1713
+ "act_data_type": "float"
1714
+ },
1715
+ "model.mtp.layers.3.input_proj": {
1716
+ "bits": 16,
1717
+ "data_type": "float",
1718
+ "act_bits": 16,
1719
+ "act_data_type": "float"
1720
+ },
1721
+ "model.mtp.layers.3.transformer_block.attn.wk_dv": {
1722
+ "bits": 16,
1723
+ "data_type": "float",
1724
+ "act_bits": 16,
1725
+ "act_data_type": "float"
1726
+ },
1727
+ "model.mtp.layers.3.transformer_block.attn.wo_ud": {
1728
+ "bits": 16,
1729
+ "data_type": "float",
1730
+ "act_bits": 16,
1731
+ "act_data_type": "float"
1732
+ },
1733
+ "model.mtp.layers.3.transformer_block.attn.wq_du": {
1734
+ "bits": 16,
1735
+ "data_type": "float",
1736
+ "act_bits": 16,
1737
+ "act_data_type": "float"
1738
+ },
1739
+ "model.mtp.layers.3.transformer_block.attn.wr_du": {
1740
+ "bits": 16,
1741
+ "data_type": "float",
1742
+ "act_bits": 16,
1743
+ "act_data_type": "float"
1744
+ },
1745
+ "model.mtp.layers.3.transformer_block.attn.wv_dv": {
1746
+ "bits": 16,
1747
+ "data_type": "float",
1748
+ "act_bits": 16,
1749
+ "act_data_type": "float"
1750
+ },
1751
+ "model.mtp.layers.3.transformer_block.mlp.w13_dn": {
1752
+ "bits": 16,
1753
+ "data_type": "float",
1754
+ "act_bits": 16,
1755
+ "act_data_type": "float"
1756
+ },
1757
+ "model.mtp.layers.3.transformer_block.mlp.w2_md": {
1758
+ "bits": 16,
1759
+ "data_type": "float",
1760
+ "act_bits": 16,
1761
+ "act_data_type": "float"
1762
+ },
1763
+ "model.mtp.layers.4.input_proj": {
1764
+ "bits": 16,
1765
+ "data_type": "float",
1766
+ "act_bits": 16,
1767
+ "act_data_type": "float"
1768
+ },
1769
+ "model.mtp.layers.4.transformer_block.attn.wk_dv": {
1770
+ "bits": 16,
1771
+ "data_type": "float",
1772
+ "act_bits": 16,
1773
+ "act_data_type": "float"
1774
+ },
1775
+ "model.mtp.layers.4.transformer_block.attn.wo_ud": {
1776
+ "bits": 16,
1777
+ "data_type": "float",
1778
+ "act_bits": 16,
1779
+ "act_data_type": "float"
1780
+ },
1781
+ "model.mtp.layers.4.transformer_block.attn.wq_du": {
1782
+ "bits": 16,
1783
+ "data_type": "float",
1784
+ "act_bits": 16,
1785
+ "act_data_type": "float"
1786
+ },
1787
+ "model.mtp.layers.4.transformer_block.attn.wr_du": {
1788
+ "bits": 16,
1789
+ "data_type": "float",
1790
+ "act_bits": 16,
1791
+ "act_data_type": "float"
1792
+ },
1793
+ "model.mtp.layers.4.transformer_block.attn.wv_dv": {
1794
+ "bits": 16,
1795
+ "data_type": "float",
1796
+ "act_bits": 16,
1797
+ "act_data_type": "float"
1798
+ },
1799
+ "model.mtp.layers.4.transformer_block.mlp.w13_dn": {
1800
+ "bits": 16,
1801
+ "data_type": "float",
1802
+ "act_bits": 16,
1803
+ "act_data_type": "float"
1804
+ },
1805
+ "model.mtp.layers.4.transformer_block.mlp.w2_md": {
1806
+ "bits": 16,
1807
+ "data_type": "float",
1808
+ "act_bits": 16,
1809
+ "act_data_type": "float"
1810
+ },
1811
+ "model.mtp.layers.5.input_proj": {
1812
+ "bits": 16,
1813
+ "data_type": "float",
1814
+ "act_bits": 16,
1815
+ "act_data_type": "float"
1816
+ },
1817
+ "model.mtp.layers.5.transformer_block.attn.wk_dv": {
1818
+ "bits": 16,
1819
+ "data_type": "float",
1820
+ "act_bits": 16,
1821
+ "act_data_type": "float"
1822
+ },
1823
+ "model.mtp.layers.5.transformer_block.attn.wo_ud": {
1824
+ "bits": 16,
1825
+ "data_type": "float",
1826
+ "act_bits": 16,
1827
+ "act_data_type": "float"
1828
+ },
1829
+ "model.mtp.layers.5.transformer_block.attn.wq_du": {
1830
+ "bits": 16,
1831
+ "data_type": "float",
1832
+ "act_bits": 16,
1833
+ "act_data_type": "float"
1834
+ },
1835
+ "model.mtp.layers.5.transformer_block.attn.wr_du": {
1836
+ "bits": 16,
1837
+ "data_type": "float",
1838
+ "act_bits": 16,
1839
+ "act_data_type": "float"
1840
+ },
1841
+ "model.mtp.layers.5.transformer_block.attn.wv_dv": {
1842
+ "bits": 16,
1843
+ "data_type": "float",
1844
+ "act_bits": 16,
1845
+ "act_data_type": "float"
1846
+ },
1847
+ "model.mtp.layers.5.transformer_block.mlp.w13_dn": {
1848
+ "bits": 16,
1849
+ "data_type": "float",
1850
+ "act_bits": 16,
1851
+ "act_data_type": "float"
1852
+ },
1853
+ "model.mtp.layers.5.transformer_block.mlp.w2_md": {
1854
+ "bits": 16,
1855
+ "data_type": "float",
1856
+ "act_bits": 16,
1857
+ "act_data_type": "float"
1858
+ },
1859
+ "model.mtp.layers.6.input_proj": {
1860
+ "bits": 16,
1861
+ "data_type": "float",
1862
+ "act_bits": 16,
1863
+ "act_data_type": "float"
1864
+ },
1865
+ "model.mtp.layers.6.transformer_block.attn.wk_dv": {
1866
+ "bits": 16,
1867
+ "data_type": "float",
1868
+ "act_bits": 16,
1869
+ "act_data_type": "float"
1870
+ },
1871
+ "model.mtp.layers.6.transformer_block.attn.wo_ud": {
1872
+ "bits": 16,
1873
+ "data_type": "float",
1874
+ "act_bits": 16,
1875
+ "act_data_type": "float"
1876
+ },
1877
+ "model.mtp.layers.6.transformer_block.attn.wq_du": {
1878
+ "bits": 16,
1879
+ "data_type": "float",
1880
+ "act_bits": 16,
1881
+ "act_data_type": "float"
1882
+ },
1883
+ "model.mtp.layers.6.transformer_block.attn.wr_du": {
1884
+ "bits": 16,
1885
+ "data_type": "float",
1886
+ "act_bits": 16,
1887
+ "act_data_type": "float"
1888
+ },
1889
+ "model.mtp.layers.6.transformer_block.attn.wv_dv": {
1890
+ "bits": 16,
1891
+ "data_type": "float",
1892
+ "act_bits": 16,
1893
+ "act_data_type": "float"
1894
+ },
1895
+ "model.mtp.layers.6.transformer_block.mlp.w13_dn": {
1896
+ "bits": 16,
1897
+ "data_type": "float",
1898
+ "act_bits": 16,
1899
+ "act_data_type": "float"
1900
+ },
1901
+ "model.mtp.layers.6.transformer_block.mlp.w2_md": {
1902
+ "bits": 16,
1903
+ "data_type": "float",
1904
+ "act_bits": 16,
1905
+ "act_data_type": "float"
1906
+ },
1907
+ "model.mtp.layers.7.input_proj": {
1908
+ "bits": 16,
1909
+ "data_type": "float",
1910
+ "act_bits": 16,
1911
+ "act_data_type": "float"
1912
+ },
1913
+ "model.mtp.layers.7.transformer_block.attn.wk_dv": {
1914
+ "bits": 16,
1915
+ "data_type": "float",
1916
+ "act_bits": 16,
1917
+ "act_data_type": "float"
1918
+ },
1919
+ "model.mtp.layers.7.transformer_block.attn.wo_ud": {
1920
+ "bits": 16,
1921
+ "data_type": "float",
1922
+ "act_bits": 16,
1923
+ "act_data_type": "float"
1924
+ },
1925
+ "model.mtp.layers.7.transformer_block.attn.wq_du": {
1926
+ "bits": 16,
1927
+ "data_type": "float",
1928
+ "act_bits": 16,
1929
+ "act_data_type": "float"
1930
+ },
1931
+ "model.mtp.layers.7.transformer_block.attn.wr_du": {
1932
+ "bits": 16,
1933
+ "data_type": "float",
1934
+ "act_bits": 16,
1935
+ "act_data_type": "float"
1936
+ },
1937
+ "model.mtp.layers.7.transformer_block.attn.wv_dv": {
1938
+ "bits": 16,
1939
+ "data_type": "float",
1940
+ "act_bits": 16,
1941
+ "act_data_type": "float"
1942
+ },
1943
+ "model.mtp.layers.7.transformer_block.mlp.w13_dn": {
1944
+ "bits": 16,
1945
+ "data_type": "float",
1946
+ "act_bits": 16,
1947
+ "act_data_type": "float"
1948
+ },
1949
+ "model.mtp.layers.7.transformer_block.mlp.w2_md": {
1950
+ "bits": 16,
1951
+ "data_type": "float",
1952
+ "act_bits": 16,
1953
+ "act_data_type": "float"
1954
+ },
1955
+ "model.llm.layers.14.attn.wk_dv": {
1956
+ "bits": 16,
1957
+ "data_type": "float",
1958
+ "act_bits": 16,
1959
+ "act_data_type": "float"
1960
+ },
1961
+ "model.llm.layers.2.attn.wr_du": {
1962
+ "bits": 16,
1963
+ "data_type": "float",
1964
+ "act_bits": 16,
1965
+ "act_data_type": "float"
1966
+ },
1967
+ "model.llm.layers.30.attn.wv_dv": {
1968
+ "bits": 16,
1969
+ "data_type": "float",
1970
+ "act_bits": 16,
1971
+ "act_data_type": "float"
1972
+ },
1973
+ "model.llm.layers.39.attn.wv_dv": {
1974
+ "bits": 16,
1975
+ "data_type": "float",
1976
+ "act_bits": 16,
1977
+ "act_data_type": "float"
1978
+ },
1979
+ "model.llm.layers.39.mlp.gate": {
1980
+ "bits": 16,
1981
+ "data_type": "float",
1982
+ "act_bits": 16,
1983
+ "act_data_type": "float"
1984
+ },
1985
+ "model.llm.layers.7.mlp.gate": {
1986
+ "bits": 16,
1987
+ "data_type": "float",
1988
+ "act_bits": 16,
1989
+ "act_data_type": "float"
1990
+ },
1991
+ "model.llm.layers.13.mlp.gate": {
1992
+ "bits": 16,
1993
+ "data_type": "float",
1994
+ "act_bits": 16,
1995
+ "act_data_type": "float"
1996
+ },
1997
+ "model.llm.layers.30.attn.wk_dv": {
1998
+ "bits": 16,
1999
+ "data_type": "float",
2000
+ "act_bits": 16,
2001
+ "act_data_type": "float"
2002
+ },
2003
+ "model.llm.layers.34.attn.wq_du": {
2004
+ "bits": 16,
2005
+ "data_type": "float",
2006
+ "act_bits": 16,
2007
+ "act_data_type": "float"
2008
+ },
2009
+ "model.llm.layers.41.attn.wq_du": {
2010
+ "bits": 16,
2011
+ "data_type": "float",
2012
+ "act_bits": 16,
2013
+ "act_data_type": "float"
2014
+ },
2015
+ "model.llm.layers.5.attn.wq_du": {
2016
+ "bits": 16,
2017
+ "data_type": "float",
2018
+ "act_bits": 16,
2019
+ "act_data_type": "float"
2020
+ },
2021
+ "model.llm.layers.7.attn.wq_du": {
2022
+ "bits": 16,
2023
+ "data_type": "float",
2024
+ "act_bits": 16,
2025
+ "act_data_type": "float"
2026
+ },
2027
+ "model.llm.layers.11.attn.wq_du": {
2028
+ "bits": 16,
2029
+ "data_type": "float",
2030
+ "act_bits": 16,
2031
+ "act_data_type": "float"
2032
+ },
2033
+ "model.llm.layers.37.attn.wv_dv": {
2034
+ "bits": 16,
2035
+ "data_type": "float",
2036
+ "act_bits": 16,
2037
+ "act_data_type": "float"
2038
+ },
2039
+ "model.llm.layers.41.attn.wk_dv": {
2040
+ "bits": 16,
2041
+ "data_type": "float",
2042
+ "act_bits": 16,
2043
+ "act_data_type": "float"
2044
+ },
2045
+ "model.llm.layers.26.attn.wq_du": {
2046
+ "bits": 16,
2047
+ "data_type": "float",
2048
+ "act_bits": 16,
2049
+ "act_data_type": "float"
2050
+ },
2051
+ "model.llm.layers.32.mlp.gate": {
2052
+ "bits": 16,
2053
+ "data_type": "float",
2054
+ "act_bits": 16,
2055
+ "act_data_type": "float"
2056
+ },
2057
+ "model.llm.layers.5.attn.wo_ud": {
2058
+ "bits": 16,
2059
+ "data_type": "float",
2060
+ "act_bits": 16,
2061
+ "act_data_type": "float"
2062
+ },
2063
+ "model.llm.layers.2.attn.wq_du": {
2064
+ "bits": 16,
2065
+ "data_type": "float",
2066
+ "act_bits": 16,
2067
+ "act_data_type": "float"
2068
+ },
2069
+ "model.llm.layers.34.mlp.gate": {
2070
+ "bits": 16,
2071
+ "data_type": "float",
2072
+ "act_bits": 16,
2073
+ "act_data_type": "float"
2074
+ }
2075
+ }
2076
+ }
2077
+ }
model-00001-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:220896ff2c157a8f1ab3aab571101b7ad85e8cee98969bdea32e62c935474ad6
3
+ size 5136558332
model-00002-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26d415d1f62c7c3274a1c032f412b2fd9c01f52a189d15d270c12e137c07fffb
3
+ size 5281036664
model-00003-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ce895ebb1f8f01c0feec547fb46f9ec1199a86164f8f3b0d4dd01f8816428a0
3
+ size 6248246264
model-00004-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7f9c0ee179e34ff85a7f166e4e22143c9321400822d843faacc6c7ee1fbcc48e
3
+ size 5189658660
model-00005-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15981a82efa092ddda33accefe1d6193d23abe03326ed7c5fafe3e66923be6fe
3
+ size 3632409948
model-00006-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0def8931f6090b84e557090fece26affd298974c1bacdce89a108a55be3566aa
3
+ size 4815146544
model-00007-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26dfa2e1120f054d4040756587b537816b656bfd7e331e0ffaf1b8523366cf9c
3
+ size 4972656182
model-00008-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27d9f00ca87295c3ae96e9a69c67439d747a9131ef16e4cda6b58387b6ef3cda
3
+ size 4852912548
model-00009-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:562a015c331c585cd05f10d37bf2cb8968b4356286fcaff2e7906f853f77efb1
3
+ size 5081635456
model-00010-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33713271b1ec1c1a057c133e5386a539a396233dbb17a713dc1d117269f79e12
3
+ size 4972678420
model-00011-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91fff27e59784cc24f5a6f01293a7511a5fa7b3387269f004be7a7557826c022
3
+ size 4947387284
model-00012-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:191ff934efbd6fd2350aecea7a130e367a7eb6fe0f19b82fc557d222fa0cd6d3
3
+ size 4731326832
model-00013-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b43fbdffa99bf3f689a6d63aaace4b7de4818b9e43893319f90b43c3e1dc8cff
3
+ size 5022781176
model-00014-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:add22ccf098179917146b9550212a3dbb6bf2a348700ebd5245e0948b78c0ce6
3
+ size 4913846332
model-00015-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d34035b5b4edf6303cfd7029d7f8759fdf7c8c28385e00419bc87b2957c0d2b5
3
+ size 3851986884
model-00016-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:950b1820b86f0da9d935c4bde10f32657f77b4c9b319d37b52226d3cc7676d3a
3
+ size 3961855356
model-00017-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fa4346facdacf80ff3b2c115f554e3024098673e2de3738bac0318ca2fa5902
3
+ size 3676559044
model-00018-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4516c277e22b0d9f82b1e7209e3342265165a48eb66c1e1abad4eadaef869885
3
+ size 4701975196
model-00019-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e44e60153c4e565f20d650e868e2e2ee78796fb2ee7402843812fbd4868257c
3
+ size 3615534910
model-00020-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:051f2cdff57b0073185d5575ddc3de979adde4bc09fcce7e99e84b6f7897f7eb
3
+ size 4850864624
model-00021-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aafe833b58d2551ee5f3f7bdbdee592bcdee970216c88ec8221b6aaa58656061
3
+ size 3477138356
model-00022-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4974d4637964f78a6a5db513c2bf050cb2d6955d312a9ece711a617a385af68a
3
+ size 5020791324
model-00023-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d3799de2e8cacace1df03abe109c90503a8c3bdb9d85fecf6fea5a6f13d9703
3
+ size 4844582312
model-00024-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e3fc079f1bbf748f805a8d40f916747787d53973806f6909553c02d1acd12872
3
+ size 4915946096
model-00025-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e7c5365f2e576e377b27e0dc543458979b2700e66830dbc6e5ac7628a4b383a
3
+ size 4632691884
model-00026-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:18b79b2646860ec912fa31521e6d935bfd50525d97f4cac909834accdac8d266
3
+ size 4571810092
model-00027-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe3966ec194474b6404cba78d0e10e5f1ed94b3508df4a5428aec124371e9509
3
+ size 4775343344
model-00028-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:410744ffd6f07956ba68daf95ca156304434c6584dbaabf8e29d2a7d5f0893d3
3
+ size 4723077176
model-00029-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3117ec491126a500a87e08b83305a372d428ac741c4a89f0d66b5a60d106dd5
3
+ size 4714473684
model-00030-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05bbf88e6ad7f0574e4fc4f7d70ced033c758aa4176a2462415fc699b2498c69
3
+ size 4699802708
model-00031-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5714f0b41166665e807205658bd33ad07be6ae0716f521e4c6d20297a6e1713b
3
+ size 4596957956
model-00032-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:98ebe21a387a182095082526c89836b11df099a68e0c840596e724b55b27b778
3
+ size 3525521192
model-00033-of-00033.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aac3b906b63c529ef7b0c9fc724f7860b5d0de58af21d39bd2013580dac7d7a5
3
+ size 4463845360
model.safetensors.index.json ADDED
@@ -0,0 +1,1135 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 0
4
+ },
5
+ "weight_map": {
6
+ "model.audio.encoder.weight": "model-00001-of-00033.safetensors",
7
+ "model.llm.layers.13.attn.wr_du.weight": "model-00001-of-00033.safetensors",
8
+ "model.llm.layers.14.attn.k_norm.weight": "model-00001-of-00033.safetensors",
9
+ "model.llm.layers.16.mlp.experts.w13_weight.weight_packed": "model-00001-of-00033.safetensors",
10
+ "model.llm.layers.16.mlp.experts.w13_weight.weight_scale": "model-00001-of-00033.safetensors",
11
+ "model.llm.layers.20.mlp_norm.weight": "model-00001-of-00033.safetensors",
12
+ "model.llm.layers.30.mlp.experts.w2_weight.weight_packed": "model-00001-of-00033.safetensors",
13
+ "model.llm.layers.30.mlp.experts.w2_weight.weight_scale": "model-00001-of-00033.safetensors",
14
+ "model.llm.layers.34.attn.wk_dv.weight": "model-00001-of-00033.safetensors",
15
+ "model.llm.layers.37.attn.wk_dv.weight": "model-00001-of-00033.safetensors",
16
+ "model.llm.layers.37.attn.wo_ud.weight": "model-00001-of-00033.safetensors",
17
+ "model.llm.layers.38.mlp.gate.global_scale": "model-00001-of-00033.safetensors",
18
+ "model.llm.embed_norm.weight": "model-00001-of-00033.safetensors",
19
+ "model.llm.layers.16.mlp_sconv.weight": "model-00001-of-00033.safetensors",
20
+ "model.llm.layers.24.attn.k_sconv.weight": "model-00001-of-00033.safetensors",
21
+ "model.llm.layers.29.attn_sconv.weight": "model-00001-of-00033.safetensors",
22
+ "model.llm.layers.3.mlp.gate.weight": "model-00001-of-00033.safetensors",
23
+ "model.llm.unembed.weight": "model-00001-of-00033.safetensors",
24
+ "model.audio.final_norm.weight": "model-00002-of-00033.safetensors",
25
+ "model.llm.layers.10.attn.q_norm.weight": "model-00002-of-00033.safetensors",
26
+ "model.llm.layers.11.mlp.shared_experts.shared_w13_weight": "model-00002-of-00033.safetensors",
27
+ "model.llm.layers.13.attn.q_norm.weight": "model-00002-of-00033.safetensors",
28
+ "model.llm.layers.13.attn.wk_dv.weight": "model-00002-of-00033.safetensors",
29
+ "model.llm.layers.14.attn.q_norm.weight": "model-00002-of-00033.safetensors",
30
+ "model.llm.layers.15.mlp.gate.bias": "model-00002-of-00033.safetensors",
31
+ "model.llm.layers.15.mlp.shared_experts.shared_w13_weight": "model-00002-of-00033.safetensors",
32
+ "model.llm.layers.17.mlp.shared_experts.shared_w2_weight": "model-00002-of-00033.safetensors",
33
+ "model.llm.layers.18.attn.q_norm.weight": "model-00002-of-00033.safetensors",
34
+ "model.llm.layers.18.attn.wq_du.weight": "model-00002-of-00033.safetensors",
35
+ "model.llm.layers.18.mlp.shared_experts.shared_w2_weight": "model-00002-of-00033.safetensors",
36
+ "model.llm.layers.19.attn.wo_ud.weight": "model-00002-of-00033.safetensors",
37
+ "model.llm.layers.19.mlp.experts.w2_weight.weight_packed": "model-00002-of-00033.safetensors",
38
+ "model.llm.layers.19.mlp.experts.w2_weight.weight_scale": "model-00002-of-00033.safetensors",
39
+ "model.llm.layers.2.attn.wv_dv.weight": "model-00002-of-00033.safetensors",
40
+ "model.llm.layers.20.attn.rel_logits_proj.proj": "model-00002-of-00033.safetensors",
41
+ "model.llm.layers.20.attn.wk_dv.weight": "model-00002-of-00033.safetensors",
42
+ "model.llm.layers.21.attn.k_norm.weight": "model-00002-of-00033.safetensors",
43
+ "model.llm.layers.21.attn.wk_dv.weight": "model-00002-of-00033.safetensors",
44
+ "model.llm.layers.21.mlp_norm.weight": "model-00002-of-00033.safetensors",
45
+ "model.llm.layers.22.attn.k_norm.weight": "model-00002-of-00033.safetensors",
46
+ "model.llm.layers.22.attn.wq_du.weight": "model-00002-of-00033.safetensors",
47
+ "model.llm.layers.23.attn.wr_du.weight": "model-00002-of-00033.safetensors",
48
+ "model.llm.layers.23.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
49
+ "model.llm.layers.24.mlp.gate.bias": "model-00002-of-00033.safetensors",
50
+ "model.llm.layers.25.attn.wo_ud.weight": "model-00002-of-00033.safetensors",
51
+ "model.llm.layers.25.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
52
+ "model.llm.layers.25.mlp_norm.weight": "model-00002-of-00033.safetensors",
53
+ "model.llm.layers.26.attn_norm.weight": "model-00002-of-00033.safetensors",
54
+ "model.llm.layers.27.attn.wk_dv.weight": "model-00002-of-00033.safetensors",
55
+ "model.llm.layers.27.attn_norm.weight": "model-00002-of-00033.safetensors",
56
+ "model.llm.layers.27.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
57
+ "model.llm.layers.27.mlp.shared_experts.shared_w2_weight": "model-00002-of-00033.safetensors",
58
+ "model.llm.layers.29.attn.k_norm.weight": "model-00002-of-00033.safetensors",
59
+ "model.llm.layers.29.attn.wv_dv.weight": "model-00002-of-00033.safetensors",
60
+ "model.llm.layers.3.attn.wo_ud.weight": "model-00002-of-00033.safetensors",
61
+ "model.llm.layers.3.attn.wr_du.weight": "model-00002-of-00033.safetensors",
62
+ "model.llm.layers.31.attn.rel_logits_proj.proj": "model-00002-of-00033.safetensors",
63
+ "model.llm.layers.32.attn.wo_ud.weight": "model-00002-of-00033.safetensors",
64
+ "model.llm.layers.32.attn.wq_du.weight": "model-00002-of-00033.safetensors",
65
+ "model.llm.layers.32.mlp.gate.bias": "model-00002-of-00033.safetensors",
66
+ "model.llm.layers.33.attn.k_norm.weight": "model-00002-of-00033.safetensors",
67
+ "model.llm.layers.33.attn.rel_logits_proj.proj": "model-00002-of-00033.safetensors",
68
+ "model.llm.layers.33.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
69
+ "model.llm.layers.34.attn_norm.weight": "model-00002-of-00033.safetensors",
70
+ "model.llm.layers.34.mlp.experts.w13_weight.weight_packed": "model-00002-of-00033.safetensors",
71
+ "model.llm.layers.34.mlp.experts.w13_weight.weight_scale": "model-00002-of-00033.safetensors",
72
+ "model.llm.layers.34.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
73
+ "model.llm.layers.36.mlp.gate.bias": "model-00002-of-00033.safetensors",
74
+ "model.llm.layers.37.attn.wr_du.weight": "model-00002-of-00033.safetensors",
75
+ "model.llm.layers.37.mlp.shared_experts.shared_w2_weight": "model-00002-of-00033.safetensors",
76
+ "model.llm.layers.38.attn.rel_logits_proj.proj": "model-00002-of-00033.safetensors",
77
+ "model.llm.layers.4.attn.wk_dv.weight": "model-00002-of-00033.safetensors",
78
+ "model.llm.layers.4.attn.wq_du.weight": "model-00002-of-00033.safetensors",
79
+ "model.llm.layers.40.attn.wr_du.weight": "model-00002-of-00033.safetensors",
80
+ "model.llm.layers.41.attn_norm.weight": "model-00002-of-00033.safetensors",
81
+ "model.llm.layers.5.mlp.gate.global_scale": "model-00002-of-00033.safetensors",
82
+ "model.llm.layers.5.mlp.shared_experts.shared_w13_weight": "model-00002-of-00033.safetensors",
83
+ "model.llm.layers.5.mlp.shared_experts.shared_w2_weight": "model-00002-of-00033.safetensors",
84
+ "model.llm.layers.7.mlp.experts.w2_weight.weight_packed": "model-00002-of-00033.safetensors",
85
+ "model.llm.layers.7.mlp.experts.w2_weight.weight_scale": "model-00002-of-00033.safetensors",
86
+ "model.llm.layers.11.mlp.gate.weight": "model-00002-of-00033.safetensors",
87
+ "model.llm.layers.18.attn.v_sconv.weight": "model-00002-of-00033.safetensors",
88
+ "model.llm.layers.2.mlp.gate.weight": "model-00002-of-00033.safetensors",
89
+ "model.llm.layers.25.mlp_sconv.weight": "model-00002-of-00033.safetensors",
90
+ "model.llm.layers.3.attn.k_sconv.weight": "model-00002-of-00033.safetensors",
91
+ "model.llm.layers.33.attn_sconv.weight": "model-00002-of-00033.safetensors",
92
+ "model.llm.layers.34.attn.k_sconv.weight": "model-00002-of-00033.safetensors",
93
+ "model.llm.layers.35.attn.v_sconv.weight": "model-00002-of-00033.safetensors",
94
+ "model.llm.layers.36.attn.k_sconv.weight": "model-00002-of-00033.safetensors",
95
+ "model.llm.layers.36.mlp_sconv.weight": "model-00002-of-00033.safetensors",
96
+ "model.llm.layers.38.mlp_sconv.weight": "model-00002-of-00033.safetensors",
97
+ "model.llm.layers.8.attn_sconv.weight": "model-00002-of-00033.safetensors",
98
+ "model.llm.layers.9.mlp_sconv.weight": "model-00002-of-00033.safetensors",
99
+ "model.llm.layers.0.attn.q_norm.weight": "model-00005-of-00033.safetensors",
100
+ "model.llm.layers.12.mlp_norm.weight": "model-00005-of-00033.safetensors",
101
+ "model.llm.layers.13.mlp_norm.weight": "model-00005-of-00033.safetensors",
102
+ "model.llm.layers.14.mlp.shared_experts.shared_w13_weight": "model-00005-of-00033.safetensors",
103
+ "model.llm.layers.16.mlp.shared_experts.shared_w2_weight": "model-00005-of-00033.safetensors",
104
+ "model.llm.layers.19.attn.q_norm.weight": "model-00005-of-00033.safetensors",
105
+ "model.llm.layers.22.mlp.gate.global_scale": "model-00005-of-00033.safetensors",
106
+ "model.llm.layers.23.attn.wk_dv.weight": "model-00005-of-00033.safetensors",
107
+ "model.llm.layers.26.attn.rel_logits_proj.proj": "model-00005-of-00033.safetensors",
108
+ "model.llm.layers.27.mlp.experts.w2_weight.weight_packed": "model-00005-of-00033.safetensors",
109
+ "model.llm.layers.27.mlp.experts.w2_weight.weight_scale": "model-00005-of-00033.safetensors",
110
+ "model.llm.layers.30.attn.wr_du.weight": "model-00005-of-00033.safetensors",
111
+ "model.llm.layers.34.attn.wv_dv.weight": "model-00005-of-00033.safetensors",
112
+ "model.llm.layers.36.attn.wr_du.weight": "model-00005-of-00033.safetensors",
113
+ "model.llm.layers.38.mlp.experts.w13_weight.weight_packed": "model-00005-of-00033.safetensors",
114
+ "model.llm.layers.38.mlp.experts.w13_weight.weight_scale": "model-00005-of-00033.safetensors",
115
+ "model.llm.layers.39.mlp.shared_experts.shared_w13_weight": "model-00005-of-00033.safetensors",
116
+ "model.llm.layers.6.attn.wr_du.weight": "model-00005-of-00033.safetensors",
117
+ "model.llm.layers.8.attn.wr_du.weight": "model-00005-of-00033.safetensors",
118
+ "model.llm.layers.9.attn.wr_du.weight": "model-00005-of-00033.safetensors",
119
+ "model.llm.layers.15.mlp.gate.weight": "model-00005-of-00033.safetensors",
120
+ "model.llm.layers.25.attn.k_sconv.weight": "model-00005-of-00033.safetensors",
121
+ "model.llm.layers.37.mlp_sconv.weight": "model-00005-of-00033.safetensors",
122
+ "model.llm.layers.4.mlp.gate.weight": "model-00005-of-00033.safetensors",
123
+ "model.llm.layers.6.mlp_sconv.weight": "model-00005-of-00033.safetensors",
124
+ "model.llm.layers.11.mlp.experts.w13_weight.weight_packed": "model-00003-of-00033.safetensors",
125
+ "model.llm.layers.11.mlp.experts.w13_weight.weight_scale": "model-00003-of-00033.safetensors",
126
+ "model.llm.layers.18.attn.rel_logits_proj.proj": "model-00003-of-00033.safetensors",
127
+ "model.llm.layers.21.attn.rel_logits_proj.proj": "model-00003-of-00033.safetensors",
128
+ "model.llm.layers.3.attn_norm.weight": "model-00003-of-00033.safetensors",
129
+ "model.llm.layers.30.attn.wq_du.weight": "model-00003-of-00033.safetensors",
130
+ "model.llm.layers.31.mlp.experts.w13_weight.weight_packed": "model-00003-of-00033.safetensors",
131
+ "model.llm.layers.31.mlp.experts.w13_weight.weight_scale": "model-00003-of-00033.safetensors",
132
+ "model.llm.layers.32.mlp_norm.weight": "model-00003-of-00033.safetensors",
133
+ "model.llm.layers.36.attn.rel_logits_proj.proj": "model-00003-of-00033.safetensors",
134
+ "model.llm.layers.39.attn_norm.weight": "model-00003-of-00033.safetensors",
135
+ "model.llm.layers.4.attn_norm.weight": "model-00003-of-00033.safetensors",
136
+ "model.llm.embed.weight": "model-00003-of-00033.safetensors",
137
+ "model.llm.layers.0.attn.k_sconv.weight": "model-00003-of-00033.safetensors",
138
+ "model.llm.layers.1.attn.v_sconv.weight": "model-00003-of-00033.safetensors",
139
+ "model.llm.layers.10.attn.v_sconv.weight": "model-00003-of-00033.safetensors",
140
+ "model.llm.layers.12.mlp_sconv.weight": "model-00003-of-00033.safetensors",
141
+ "model.llm.layers.14.attn.k_sconv.weight": "model-00003-of-00033.safetensors",
142
+ "model.llm.layers.18.mlp.gate.weight": "model-00003-of-00033.safetensors",
143
+ "model.llm.layers.22.attn.k_sconv.weight": "model-00003-of-00033.safetensors",
144
+ "model.llm.layers.26.attn.k_sconv.weight": "model-00003-of-00033.safetensors",
145
+ "model.llm.layers.26.attn_sconv.weight": "model-00003-of-00033.safetensors",
146
+ "model.llm.layers.27.mlp.gate.weight": "model-00003-of-00033.safetensors",
147
+ "model.llm.layers.34.attn_sconv.weight": "model-00003-of-00033.safetensors",
148
+ "model.llm.layers.36.attn.v_sconv.weight": "model-00003-of-00033.safetensors",
149
+ "model.llm.layers.9.attn_sconv.weight": "model-00003-of-00033.safetensors",
150
+ "model.llm.layers.0.attn.k_norm.weight": "model-00004-of-00033.safetensors",
151
+ "model.llm.layers.0.attn.wq_du.weight": "model-00004-of-00033.safetensors",
152
+ "model.llm.layers.0.mlp.w13_dn.weight": "model-00004-of-00033.safetensors",
153
+ "model.llm.layers.13.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
154
+ "model.llm.layers.13.mlp.gate.bias": "model-00004-of-00033.safetensors",
155
+ "model.llm.layers.15.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
156
+ "model.llm.layers.15.mlp.shared_experts.shared_w2_weight": "model-00004-of-00033.safetensors",
157
+ "model.llm.layers.16.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
158
+ "model.llm.layers.16.mlp.gate.global_scale": "model-00004-of-00033.safetensors",
159
+ "model.llm.layers.17.attn.k_norm.weight": "model-00004-of-00033.safetensors",
160
+ "model.llm.layers.19.mlp.gate.bias": "model-00004-of-00033.safetensors",
161
+ "model.llm.layers.20.attn.wv_dv.weight": "model-00004-of-00033.safetensors",
162
+ "model.llm.layers.21.mlp.gate.global_scale": "model-00004-of-00033.safetensors",
163
+ "model.llm.layers.22.mlp_norm.weight": "model-00004-of-00033.safetensors",
164
+ "model.llm.layers.23.attn.wo_ud.weight": "model-00004-of-00033.safetensors",
165
+ "model.llm.layers.23.attn.wq_du.weight": "model-00004-of-00033.safetensors",
166
+ "model.llm.layers.23.attn.wv_dv.weight": "model-00004-of-00033.safetensors",
167
+ "model.llm.layers.24.attn.wq_du.weight": "model-00004-of-00033.safetensors",
168
+ "model.llm.layers.26.attn.q_norm.weight": "model-00004-of-00033.safetensors",
169
+ "model.llm.layers.27.attn.wq_du.weight": "model-00004-of-00033.safetensors",
170
+ "model.llm.layers.28.mlp.experts.w2_weight.weight_packed": "model-00004-of-00033.safetensors",
171
+ "model.llm.layers.28.mlp.experts.w2_weight.weight_scale": "model-00004-of-00033.safetensors",
172
+ "model.llm.layers.32.attn_norm.weight": "model-00004-of-00033.safetensors",
173
+ "model.llm.layers.32.mlp.gate.global_scale": "model-00004-of-00033.safetensors",
174
+ "model.llm.layers.34.mlp.shared_experts.shared_w2_weight": "model-00004-of-00033.safetensors",
175
+ "model.llm.layers.35.attn.wk_dv.weight": "model-00004-of-00033.safetensors",
176
+ "model.llm.layers.36.attn.wo_ud.weight": "model-00004-of-00033.safetensors",
177
+ "model.llm.layers.39.attn.q_norm.weight": "model-00004-of-00033.safetensors",
178
+ "model.llm.layers.4.mlp.experts.w13_weight.weight_packed": "model-00004-of-00033.safetensors",
179
+ "model.llm.layers.4.mlp.experts.w13_weight.weight_scale": "model-00004-of-00033.safetensors",
180
+ "model.llm.layers.4.mlp.gate.bias": "model-00004-of-00033.safetensors",
181
+ "model.llm.layers.41.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
182
+ "model.llm.layers.41.mlp.experts.w2_weight.weight_packed": "model-00004-of-00033.safetensors",
183
+ "model.llm.layers.41.mlp.experts.w2_weight.weight_scale": "model-00004-of-00033.safetensors",
184
+ "model.llm.layers.5.mlp_norm.weight": "model-00004-of-00033.safetensors",
185
+ "model.llm.layers.7.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
186
+ "model.llm.layers.7.attn.wr_du.weight": "model-00004-of-00033.safetensors",
187
+ "model.llm.layers.9.attn.rel_logits_proj.proj": "model-00004-of-00033.safetensors",
188
+ "model.visual.layers.linear_2.weight": "model-00004-of-00033.safetensors",
189
+ "model.llm.layers.1.attn_sconv.weight": "model-00004-of-00033.safetensors",
190
+ "model.llm.layers.11.attn.k_sconv.weight": "model-00004-of-00033.safetensors",
191
+ "model.llm.layers.14.mlp.gate.weight": "model-00004-of-00033.safetensors",
192
+ "model.llm.layers.18.attn_sconv.weight": "model-00004-of-00033.safetensors",
193
+ "model.llm.layers.26.mlp.gate.weight": "model-00004-of-00033.safetensors",
194
+ "model.llm.layers.28.attn.k_sconv.weight": "model-00004-of-00033.safetensors",
195
+ "model.llm.layers.29.mlp_sconv.weight": "model-00004-of-00033.safetensors",
196
+ "model.llm.layers.34.attn.v_sconv.weight": "model-00004-of-00033.safetensors",
197
+ "model.llm.layers.37.attn.v_sconv.weight": "model-00004-of-00033.safetensors",
198
+ "model.llm.layers.37.mlp.gate.weight": "model-00004-of-00033.safetensors",
199
+ "model.llm.layers.38.attn.v_sconv.weight": "model-00004-of-00033.safetensors",
200
+ "model.llm.layers.41.mlp.gate.weight": "model-00004-of-00033.safetensors",
201
+ "model.llm.layers.5.attn.k_sconv.weight": "model-00004-of-00033.safetensors",
202
+ "model.llm.layers.5.mlp.gate.weight": "model-00004-of-00033.safetensors",
203
+ "model.llm.layers.8.attn.k_sconv.weight": "model-00004-of-00033.safetensors",
204
+ "model.llm.layers.0.attn.rel_logits_proj.proj": "model-00006-of-00033.safetensors",
205
+ "model.llm.layers.12.attn.wo_ud.weight": "model-00006-of-00033.safetensors",
206
+ "model.llm.layers.14.attn.wq_du.weight": "model-00006-of-00033.safetensors",
207
+ "model.llm.layers.16.attn.k_norm.weight": "model-00006-of-00033.safetensors",
208
+ "model.llm.layers.25.mlp.experts.w2_weight.weight_packed": "model-00006-of-00033.safetensors",
209
+ "model.llm.layers.25.mlp.experts.w2_weight.weight_scale": "model-00006-of-00033.safetensors",
210
+ "model.llm.layers.27.attn.q_norm.weight": "model-00006-of-00033.safetensors",
211
+ "model.llm.layers.28.attn.rel_logits_proj.proj": "model-00006-of-00033.safetensors",
212
+ "model.llm.layers.28.attn.wq_du.weight": "model-00006-of-00033.safetensors",
213
+ "model.llm.layers.31.attn.wr_du.weight": "model-00006-of-00033.safetensors",
214
+ "model.llm.layers.34.mlp_norm.weight": "model-00006-of-00033.safetensors",
215
+ "model.llm.layers.35.attn.wr_du.weight": "model-00006-of-00033.safetensors",
216
+ "model.llm.layers.36.mlp.experts.w2_weight.weight_packed": "model-00006-of-00033.safetensors",
217
+ "model.llm.layers.36.mlp.experts.w2_weight.weight_scale": "model-00006-of-00033.safetensors",
218
+ "model.llm.layers.36.mlp.shared_experts.shared_w2_weight": "model-00006-of-00033.safetensors",
219
+ "model.llm.layers.38.attn.wq_du.weight": "model-00006-of-00033.safetensors",
220
+ "model.llm.layers.41.mlp.experts.w13_weight.weight_packed": "model-00006-of-00033.safetensors",
221
+ "model.llm.layers.41.mlp.experts.w13_weight.weight_scale": "model-00006-of-00033.safetensors",
222
+ "model.llm.layers.6.mlp.shared_experts.shared_w2_weight": "model-00006-of-00033.safetensors",
223
+ "model.llm.layers.8.attn.q_norm.weight": "model-00006-of-00033.safetensors",
224
+ "model.llm.layers.8.attn.wo_ud.weight": "model-00006-of-00033.safetensors",
225
+ "model.llm.layers.9.attn.wk_dv.weight": "model-00006-of-00033.safetensors",
226
+ "model.llm.layers.25.attn.v_sconv.weight": "model-00006-of-00033.safetensors",
227
+ "model.llm.layers.3.attn_sconv.weight": "model-00006-of-00033.safetensors",
228
+ "model.llm.layers.0.mlp.global_scale": "model-00007-of-00033.safetensors",
229
+ "model.llm.layers.11.attn.wk_dv.weight": "model-00007-of-00033.safetensors",
230
+ "model.llm.layers.11.mlp.experts.w2_weight.weight_packed": "model-00007-of-00033.safetensors",
231
+ "model.llm.layers.11.mlp.experts.w2_weight.weight_scale": "model-00007-of-00033.safetensors",
232
+ "model.llm.layers.11.mlp_norm.weight": "model-00007-of-00033.safetensors",
233
+ "model.llm.layers.12.attn.wq_du.weight": "model-00007-of-00033.safetensors",
234
+ "model.llm.layers.14.attn.wo_ud.weight": "model-00007-of-00033.safetensors",
235
+ "model.llm.layers.15.attn.wk_dv.weight": "model-00007-of-00033.safetensors",
236
+ "model.llm.layers.17.attn.rel_logits_proj.proj": "model-00007-of-00033.safetensors",
237
+ "model.llm.layers.17.mlp.shared_experts.shared_w13_weight": "model-00007-of-00033.safetensors",
238
+ "model.llm.layers.2.attn.wo_ud.weight": "model-00007-of-00033.safetensors",
239
+ "model.llm.layers.21.mlp.gate.bias": "model-00007-of-00033.safetensors",
240
+ "model.llm.layers.25.mlp.experts.w13_weight.weight_packed": "model-00007-of-00033.safetensors",
241
+ "model.llm.layers.25.mlp.experts.w13_weight.weight_scale": "model-00007-of-00033.safetensors",
242
+ "model.llm.layers.28.attn.wo_ud.weight": "model-00007-of-00033.safetensors",
243
+ "model.llm.layers.28.mlp.gate.global_scale": "model-00007-of-00033.safetensors",
244
+ "model.llm.layers.3.attn.wk_dv.weight": "model-00007-of-00033.safetensors",
245
+ "model.llm.layers.3.mlp_norm.weight": "model-00007-of-00033.safetensors",
246
+ "model.llm.layers.30.attn_norm.weight": "model-00007-of-00033.safetensors",
247
+ "model.llm.layers.30.mlp.shared_experts.shared_w13_weight": "model-00007-of-00033.safetensors",
248
+ "model.llm.layers.32.attn.rel_logits_proj.proj": "model-00007-of-00033.safetensors",
249
+ "model.llm.layers.33.attn.wr_du.weight": "model-00007-of-00033.safetensors",
250
+ "model.llm.layers.34.attn.wr_du.weight": "model-00007-of-00033.safetensors",
251
+ "model.llm.layers.36.mlp.shared_experts.shared_w13_weight": "model-00007-of-00033.safetensors",
252
+ "model.llm.layers.38.attn.k_norm.weight": "model-00007-of-00033.safetensors",
253
+ "model.llm.layers.6.attn.wo_ud.weight": "model-00007-of-00033.safetensors",
254
+ "model.llm.layers.8.mlp.experts.w2_weight.weight_packed": "model-00007-of-00033.safetensors",
255
+ "model.llm.layers.8.mlp.experts.w2_weight.weight_scale": "model-00007-of-00033.safetensors",
256
+ "model.llm.layers.0.attn.v_sconv.weight": "model-00007-of-00033.safetensors",
257
+ "model.llm.layers.13.attn.v_sconv.weight": "model-00007-of-00033.safetensors",
258
+ "model.llm.layers.15.mlp_sconv.weight": "model-00007-of-00033.safetensors",
259
+ "model.llm.layers.19.mlp.gate.weight": "model-00007-of-00033.safetensors",
260
+ "model.llm.layers.2.mlp_sconv.weight": "model-00007-of-00033.safetensors",
261
+ "model.llm.layers.24.mlp.gate.weight": "model-00007-of-00033.safetensors",
262
+ "model.llm.layers.32.mlp_sconv.weight": "model-00007-of-00033.safetensors",
263
+ "model.llm.layers.35.attn_sconv.weight": "model-00007-of-00033.safetensors",
264
+ "model.llm.layers.35.mlp.gate.weight": "model-00007-of-00033.safetensors",
265
+ "model.llm.layers.4.mlp_sconv.weight": "model-00007-of-00033.safetensors",
266
+ "model.llm.layers.0.attn.wk_dv.weight": "model-00008-of-00033.safetensors",
267
+ "model.llm.layers.10.attn.wr_du.weight": "model-00008-of-00033.safetensors",
268
+ "model.llm.layers.13.attn.wv_dv.weight": "model-00008-of-00033.safetensors",
269
+ "model.llm.layers.15.attn.wr_du.weight": "model-00008-of-00033.safetensors",
270
+ "model.llm.layers.15.mlp_norm.weight": "model-00008-of-00033.safetensors",
271
+ "model.llm.layers.16.mlp.experts.w2_weight.weight_packed": "model-00008-of-00033.safetensors",
272
+ "model.llm.layers.16.mlp.experts.w2_weight.weight_scale": "model-00008-of-00033.safetensors",
273
+ "model.llm.layers.17.attn.wo_ud.weight": "model-00008-of-00033.safetensors",
274
+ "model.llm.layers.19.mlp.shared_experts.shared_w13_weight": "model-00008-of-00033.safetensors",
275
+ "model.llm.layers.19.mlp.shared_experts.shared_w2_weight": "model-00008-of-00033.safetensors",
276
+ "model.llm.layers.2.attn.k_norm.weight": "model-00008-of-00033.safetensors",
277
+ "model.llm.layers.2.attn.q_norm.weight": "model-00008-of-00033.safetensors",
278
+ "model.llm.layers.2.mlp.shared_experts.shared_w2_weight": "model-00008-of-00033.safetensors",
279
+ "model.llm.layers.20.attn.wr_du.weight": "model-00008-of-00033.safetensors",
280
+ "model.llm.layers.23.mlp_norm.weight": "model-00008-of-00033.safetensors",
281
+ "model.llm.layers.24.mlp_norm.weight": "model-00008-of-00033.safetensors",
282
+ "model.llm.layers.30.mlp.gate.global_scale": "model-00008-of-00033.safetensors",
283
+ "model.llm.layers.31.attn.q_norm.weight": "model-00008-of-00033.safetensors",
284
+ "model.llm.layers.31.mlp.experts.w2_weight.weight_packed": "model-00008-of-00033.safetensors",
285
+ "model.llm.layers.31.mlp.experts.w2_weight.weight_scale": "model-00008-of-00033.safetensors",
286
+ "model.llm.layers.36.attn.wv_dv.weight": "model-00008-of-00033.safetensors",
287
+ "model.llm.layers.37.mlp.experts.w13_weight.weight_packed": "model-00008-of-00033.safetensors",
288
+ "model.llm.layers.37.mlp.experts.w13_weight.weight_scale": "model-00008-of-00033.safetensors",
289
+ "model.llm.layers.38.attn.wk_dv.weight": "model-00008-of-00033.safetensors",
290
+ "model.llm.layers.41.attn.wo_ud.weight": "model-00008-of-00033.safetensors",
291
+ "model.llm.layers.5.attn.rel_logits_proj.proj": "model-00008-of-00033.safetensors",
292
+ "model.llm.layers.6.attn.rel_logits_proj.proj": "model-00008-of-00033.safetensors",
293
+ "model.llm.layers.9.attn.wo_ud.weight": "model-00008-of-00033.safetensors",
294
+ "model.llm.layers.9.attn.wv_dv.weight": "model-00008-of-00033.safetensors",
295
+ "model.llm.norm.weight": "model-00008-of-00033.safetensors",
296
+ "model.llm.layers.2.attn.v_sconv.weight": "model-00008-of-00033.safetensors",
297
+ "model.llm.layers.32.attn.v_sconv.weight": "model-00008-of-00033.safetensors",
298
+ "model.llm.layers.0.attn.wo_ud.weight": "model-00009-of-00033.safetensors",
299
+ "model.llm.layers.1.attn_norm.weight": "model-00009-of-00033.safetensors",
300
+ "model.llm.layers.10.attn.wq_du.weight": "model-00009-of-00033.safetensors",
301
+ "model.llm.layers.12.attn.rel_logits_proj.proj": "model-00009-of-00033.safetensors",
302
+ "model.llm.layers.17.mlp.experts.w2_weight.weight_packed": "model-00009-of-00033.safetensors",
303
+ "model.llm.layers.17.mlp.experts.w2_weight.weight_scale": "model-00009-of-00033.safetensors",
304
+ "model.llm.layers.19.attn.wv_dv.weight": "model-00009-of-00033.safetensors",
305
+ "model.llm.layers.20.mlp.shared_experts.shared_w13_weight": "model-00009-of-00033.safetensors",
306
+ "model.llm.layers.21.attn.wo_ud.weight": "model-00009-of-00033.safetensors",
307
+ "model.llm.layers.24.mlp.shared_experts.shared_w13_weight": "model-00009-of-00033.safetensors",
308
+ "model.llm.layers.25.attn.rel_logits_proj.proj": "model-00009-of-00033.safetensors",
309
+ "model.llm.layers.28.attn.k_norm.weight": "model-00009-of-00033.safetensors",
310
+ "model.llm.layers.28.attn.wr_du.weight": "model-00009-of-00033.safetensors",
311
+ "model.llm.layers.3.mlp.shared_experts.shared_w13_weight": "model-00009-of-00033.safetensors",
312
+ "model.llm.layers.30.attn.wo_ud.weight": "model-00009-of-00033.safetensors",
313
+ "model.llm.layers.30.mlp.gate.bias": "model-00009-of-00033.safetensors",
314
+ "model.llm.layers.30.mlp_norm.weight": "model-00009-of-00033.safetensors",
315
+ "model.llm.layers.32.mlp.experts.w2_weight.weight_packed": "model-00009-of-00033.safetensors",
316
+ "model.llm.layers.32.mlp.experts.w2_weight.weight_scale": "model-00009-of-00033.safetensors",
317
+ "model.llm.layers.35.attn.wq_du.weight": "model-00009-of-00033.safetensors",
318
+ "model.llm.layers.35.mlp.experts.w13_weight.weight_packed": "model-00009-of-00033.safetensors",
319
+ "model.llm.layers.35.mlp.experts.w13_weight.weight_scale": "model-00009-of-00033.safetensors",
320
+ "model.llm.layers.35.mlp_norm.weight": "model-00009-of-00033.safetensors",
321
+ "model.llm.layers.37.mlp.shared_experts.shared_w13_weight": "model-00009-of-00033.safetensors",
322
+ "model.llm.layers.40.attn.wo_ud.weight": "model-00009-of-00033.safetensors",
323
+ "model.llm.layers.7.attn.q_norm.weight": "model-00009-of-00033.safetensors",
324
+ "model.llm.layers.8.attn.wq_du.weight": "model-00009-of-00033.safetensors",
325
+ "model.llm.layers.9.mlp.gate.bias": "model-00009-of-00033.safetensors",
326
+ "model.llm.layers.10.mlp.gate.weight": "model-00009-of-00033.safetensors",
327
+ "model.llm.layers.13.attn_sconv.weight": "model-00009-of-00033.safetensors",
328
+ "model.llm.layers.17.mlp_sconv.weight": "model-00009-of-00033.safetensors",
329
+ "model.llm.layers.18.attn.k_sconv.weight": "model-00009-of-00033.safetensors",
330
+ "model.llm.layers.19.attn.k_sconv.weight": "model-00009-of-00033.safetensors",
331
+ "model.llm.layers.27.attn_sconv.weight": "model-00009-of-00033.safetensors",
332
+ "model.llm.layers.31.attn.v_sconv.weight": "model-00009-of-00033.safetensors",
333
+ "model.llm.layers.31.mlp_sconv.weight": "model-00009-of-00033.safetensors",
334
+ "model.llm.layers.0.attn.wr_du.weight": "model-00010-of-00033.safetensors",
335
+ "model.llm.layers.0.attn.wv_dv.weight": "model-00010-of-00033.safetensors",
336
+ "model.llm.layers.1.attn.wo_ud.weight": "model-00010-of-00033.safetensors",
337
+ "model.llm.layers.1.attn.wv_dv.weight": "model-00010-of-00033.safetensors",
338
+ "model.llm.layers.1.mlp_norm.weight": "model-00010-of-00033.safetensors",
339
+ "model.llm.layers.12.mlp.shared_experts.shared_w2_weight": "model-00010-of-00033.safetensors",
340
+ "model.llm.layers.14.attn_norm.weight": "model-00010-of-00033.safetensors",
341
+ "model.llm.layers.14.mlp.experts.w2_weight.weight_packed": "model-00010-of-00033.safetensors",
342
+ "model.llm.layers.14.mlp.experts.w2_weight.weight_scale": "model-00010-of-00033.safetensors",
343
+ "model.llm.layers.16.attn.wr_du.weight": "model-00010-of-00033.safetensors",
344
+ "model.llm.layers.17.attn_norm.weight": "model-00010-of-00033.safetensors",
345
+ "model.llm.layers.18.attn.k_norm.weight": "model-00010-of-00033.safetensors",
346
+ "model.llm.layers.18.attn_norm.weight": "model-00010-of-00033.safetensors",
347
+ "model.llm.layers.19.attn_norm.weight": "model-00010-of-00033.safetensors",
348
+ "model.llm.layers.19.mlp_norm.weight": "model-00010-of-00033.safetensors",
349
+ "model.llm.layers.21.attn.wq_du.weight": "model-00010-of-00033.safetensors",
350
+ "model.llm.layers.21.attn.wv_dv.weight": "model-00010-of-00033.safetensors",
351
+ "model.llm.layers.23.attn.q_norm.weight": "model-00010-of-00033.safetensors",
352
+ "model.llm.layers.24.attn.rel_logits_proj.proj": "model-00010-of-00033.safetensors",
353
+ "model.llm.layers.24.attn.wv_dv.weight": "model-00010-of-00033.safetensors",
354
+ "model.llm.layers.25.attn.wq_du.weight": "model-00010-of-00033.safetensors",
355
+ "model.llm.layers.26.attn.wk_dv.weight": "model-00010-of-00033.safetensors",
356
+ "model.llm.layers.27.attn.wo_ud.weight": "model-00010-of-00033.safetensors",
357
+ "model.llm.layers.28.mlp.shared_experts.shared_w13_weight": "model-00010-of-00033.safetensors",
358
+ "model.llm.layers.29.attn.wq_du.weight": "model-00010-of-00033.safetensors",
359
+ "model.llm.layers.3.attn.k_norm.weight": "model-00010-of-00033.safetensors",
360
+ "model.llm.layers.30.attn.k_norm.weight": "model-00010-of-00033.safetensors",
361
+ "model.llm.layers.30.attn.q_norm.weight": "model-00010-of-00033.safetensors",
362
+ "model.llm.layers.31.attn.k_norm.weight": "model-00010-of-00033.safetensors",
363
+ "model.llm.layers.31.mlp_norm.weight": "model-00010-of-00033.safetensors",
364
+ "model.llm.layers.33.attn.wv_dv.weight": "model-00010-of-00033.safetensors",
365
+ "model.llm.layers.33.mlp.experts.w2_weight.weight_packed": "model-00010-of-00033.safetensors",
366
+ "model.llm.layers.33.mlp.experts.w2_weight.weight_scale": "model-00010-of-00033.safetensors",
367
+ "model.llm.layers.36.attn.wk_dv.weight": "model-00010-of-00033.safetensors",
368
+ "model.llm.layers.39.attn.rel_logits_proj.proj": "model-00010-of-00033.safetensors",
369
+ "model.llm.layers.39.attn.wr_du.weight": "model-00010-of-00033.safetensors",
370
+ "model.llm.layers.39.mlp.shared_experts.shared_w2_weight": "model-00010-of-00033.safetensors",
371
+ "model.llm.layers.4.attn.rel_logits_proj.proj": "model-00010-of-00033.safetensors",
372
+ "model.llm.layers.4.mlp.gate.global_scale": "model-00010-of-00033.safetensors",
373
+ "model.llm.layers.5.attn.q_norm.weight": "model-00010-of-00033.safetensors",
374
+ "model.llm.layers.5.mlp.experts.w13_weight.weight_packed": "model-00010-of-00033.safetensors",
375
+ "model.llm.layers.5.mlp.experts.w13_weight.weight_scale": "model-00010-of-00033.safetensors",
376
+ "model.llm.layers.5.mlp.gate.bias": "model-00010-of-00033.safetensors",
377
+ "model.llm.layers.8.mlp.shared_experts.shared_w2_weight": "model-00010-of-00033.safetensors",
378
+ "model.visual.layers.norm_2.weight": "model-00010-of-00033.safetensors",
379
+ "model.llm.layers.10.mlp_sconv.weight": "model-00010-of-00033.safetensors",
380
+ "model.llm.layers.12.attn_sconv.weight": "model-00010-of-00033.safetensors",
381
+ "model.llm.layers.14.attn_sconv.weight": "model-00010-of-00033.safetensors",
382
+ "model.llm.layers.19.attn.v_sconv.weight": "model-00010-of-00033.safetensors",
383
+ "model.llm.layers.23.mlp.gate.bias": "model-00010-of-00033.safetensors",
384
+ "model.llm.layers.23.mlp.gate.weight": "model-00010-of-00033.safetensors",
385
+ "model.llm.layers.28.attn.v_sconv.weight": "model-00010-of-00033.safetensors",
386
+ "model.llm.layers.30.mlp_sconv.weight": "model-00010-of-00033.safetensors",
387
+ "model.llm.layers.41.attn_sconv.weight": "model-00010-of-00033.safetensors",
388
+ "model.llm.layers.7.attn.k_sconv.weight": "model-00010-of-00033.safetensors",
389
+ "model.llm.layers.0.attn_norm.weight": "model-00011-of-00033.safetensors",
390
+ "model.llm.layers.1.attn.k_norm.weight": "model-00011-of-00033.safetensors",
391
+ "model.llm.layers.13.attn.wq_du.weight": "model-00011-of-00033.safetensors",
392
+ "model.llm.layers.14.attn.rel_logits_proj.proj": "model-00011-of-00033.safetensors",
393
+ "model.llm.layers.14.mlp.shared_experts.shared_w2_weight": "model-00011-of-00033.safetensors",
394
+ "model.llm.layers.15.attn.wq_du.weight": "model-00011-of-00033.safetensors",
395
+ "model.llm.layers.16.attn.wq_du.weight": "model-00011-of-00033.safetensors",
396
+ "model.llm.layers.16.mlp.gate.bias": "model-00011-of-00033.safetensors",
397
+ "model.llm.layers.18.attn.wk_dv.weight": "model-00011-of-00033.safetensors",
398
+ "model.llm.layers.2.mlp_norm.weight": "model-00011-of-00033.safetensors",
399
+ "model.llm.layers.20.attn.wo_ud.weight": "model-00011-of-00033.safetensors",
400
+ "model.llm.layers.20.mlp.gate.bias": "model-00011-of-00033.safetensors",
401
+ "model.llm.layers.20.mlp.gate.global_scale": "model-00011-of-00033.safetensors",
402
+ "model.llm.layers.22.attn.wo_ud.weight": "model-00011-of-00033.safetensors",
403
+ "model.llm.layers.25.mlp.shared_experts.shared_w13_weight": "model-00011-of-00033.safetensors",
404
+ "model.llm.layers.26.mlp.experts.w13_weight.weight_packed": "model-00011-of-00033.safetensors",
405
+ "model.llm.layers.26.mlp.experts.w13_weight.weight_scale": "model-00011-of-00033.safetensors",
406
+ "model.llm.layers.27.attn.wr_du.weight": "model-00011-of-00033.safetensors",
407
+ "model.llm.layers.27.mlp.gate.bias": "model-00011-of-00033.safetensors",
408
+ "model.llm.layers.3.mlp.shared_experts.shared_w2_weight": "model-00011-of-00033.safetensors",
409
+ "model.llm.layers.35.attn.q_norm.weight": "model-00011-of-00033.safetensors",
410
+ "model.llm.layers.38.attn.q_norm.weight": "model-00011-of-00033.safetensors",
411
+ "model.llm.layers.39.mlp.experts.w2_weight.weight_packed": "model-00011-of-00033.safetensors",
412
+ "model.llm.layers.39.mlp.experts.w2_weight.weight_scale": "model-00011-of-00033.safetensors",
413
+ "model.llm.layers.4.mlp.shared_experts.shared_w13_weight": "model-00011-of-00033.safetensors",
414
+ "model.llm.layers.40.attn.k_norm.weight": "model-00011-of-00033.safetensors",
415
+ "model.llm.layers.6.mlp.experts.w2_weight.weight_packed": "model-00011-of-00033.safetensors",
416
+ "model.llm.layers.6.mlp.experts.w2_weight.weight_scale": "model-00011-of-00033.safetensors",
417
+ "model.llm.layers.8.attn.k_norm.weight": "model-00011-of-00033.safetensors",
418
+ "model.llm.layers.8.attn.rel_logits_proj.proj": "model-00011-of-00033.safetensors",
419
+ "model.visual.layers.norm_0.weight": "model-00011-of-00033.safetensors",
420
+ "model.llm.layers.12.attn.k_sconv.weight": "model-00011-of-00033.safetensors",
421
+ "model.llm.layers.20.mlp_sconv.weight": "model-00011-of-00033.safetensors",
422
+ "model.llm.layers.21.attn_sconv.weight": "model-00011-of-00033.safetensors",
423
+ "model.llm.layers.29.mlp.gate.weight": "model-00011-of-00033.safetensors",
424
+ "model.llm.layers.32.attn.k_sconv.weight": "model-00011-of-00033.safetensors",
425
+ "model.llm.layers.35.mlp_sconv.weight": "model-00011-of-00033.safetensors",
426
+ "model.llm.layers.41.attn.v_sconv.weight": "model-00011-of-00033.safetensors",
427
+ "model.llm.layers.7.attn.v_sconv.weight": "model-00011-of-00033.safetensors",
428
+ "model.llm.layers.15.mlp.experts.w13_weight.weight_packed": "model-00012-of-00033.safetensors",
429
+ "model.llm.layers.15.mlp.experts.w13_weight.weight_scale": "model-00012-of-00033.safetensors",
430
+ "model.llm.layers.16.attn.wv_dv.weight": "model-00012-of-00033.safetensors",
431
+ "model.llm.layers.17.attn.wq_du.weight": "model-00012-of-00033.safetensors",
432
+ "model.llm.layers.19.attn.wq_du.weight": "model-00012-of-00033.safetensors",
433
+ "model.llm.layers.2.mlp.gate.global_scale": "model-00012-of-00033.safetensors",
434
+ "model.llm.layers.20.attn.k_norm.weight": "model-00012-of-00033.safetensors",
435
+ "model.llm.layers.24.mlp.experts.w13_weight.weight_packed": "model-00012-of-00033.safetensors",
436
+ "model.llm.layers.24.mlp.experts.w13_weight.weight_scale": "model-00012-of-00033.safetensors",
437
+ "model.llm.layers.25.attn_norm.weight": "model-00012-of-00033.safetensors",
438
+ "model.llm.layers.27.attn.k_norm.weight": "model-00012-of-00033.safetensors",
439
+ "model.llm.layers.27.attn.wv_dv.weight": "model-00012-of-00033.safetensors",
440
+ "model.llm.layers.29.attn.wo_ud.weight": "model-00012-of-00033.safetensors",
441
+ "model.llm.layers.33.mlp.gate.bias": "model-00012-of-00033.safetensors",
442
+ "model.llm.layers.33.mlp_norm.weight": "model-00012-of-00033.safetensors",
443
+ "model.llm.layers.36.mlp.gate.global_scale": "model-00012-of-00033.safetensors",
444
+ "model.llm.layers.37.attn_norm.weight": "model-00012-of-00033.safetensors",
445
+ "model.llm.layers.38.attn.wv_dv.weight": "model-00012-of-00033.safetensors",
446
+ "model.llm.layers.39.attn.wk_dv.weight": "model-00012-of-00033.safetensors",
447
+ "model.llm.layers.4.attn.wo_ud.weight": "model-00012-of-00033.safetensors",
448
+ "model.llm.layers.5.attn_norm.weight": "model-00012-of-00033.safetensors",
449
+ "model.llm.layers.0.attn_sconv.weight": "model-00012-of-00033.safetensors",
450
+ "model.llm.layers.15.attn.k_sconv.weight": "model-00012-of-00033.safetensors",
451
+ "model.llm.layers.22.mlp_sconv.weight": "model-00012-of-00033.safetensors",
452
+ "model.llm.layers.33.attn.v_sconv.weight": "model-00012-of-00033.safetensors",
453
+ "model.llm.layers.4.attn_sconv.weight": "model-00012-of-00033.safetensors",
454
+ "model.llm.layers.12.mlp.gate.bias": "model-00015-of-00033.safetensors",
455
+ "model.llm.layers.14.attn.wv_dv.weight": "model-00015-of-00033.safetensors",
456
+ "model.llm.layers.16.attn.wk_dv.weight": "model-00015-of-00033.safetensors",
457
+ "model.llm.layers.17.attn.q_norm.weight": "model-00015-of-00033.safetensors",
458
+ "model.llm.layers.18.attn.wo_ud.weight": "model-00015-of-00033.safetensors",
459
+ "model.llm.layers.18.mlp.gate.bias": "model-00015-of-00033.safetensors",
460
+ "model.llm.layers.19.attn.k_norm.weight": "model-00015-of-00033.safetensors",
461
+ "model.llm.layers.21.mlp.experts.w2_weight.weight_packed": "model-00015-of-00033.safetensors",
462
+ "model.llm.layers.21.mlp.experts.w2_weight.weight_scale": "model-00015-of-00033.safetensors",
463
+ "model.llm.layers.22.attn.wv_dv.weight": "model-00015-of-00033.safetensors",
464
+ "model.llm.layers.22.mlp.shared_experts.shared_w13_weight": "model-00015-of-00033.safetensors",
465
+ "model.llm.layers.24.mlp.shared_experts.shared_w2_weight": "model-00015-of-00033.safetensors",
466
+ "model.llm.layers.26.attn.k_norm.weight": "model-00015-of-00033.safetensors",
467
+ "model.llm.layers.27.mlp.shared_experts.shared_w13_weight": "model-00015-of-00033.safetensors",
468
+ "model.llm.layers.28.attn.q_norm.weight": "model-00015-of-00033.safetensors",
469
+ "model.llm.layers.3.mlp.experts.w2_weight.weight_packed": "model-00015-of-00033.safetensors",
470
+ "model.llm.layers.3.mlp.experts.w2_weight.weight_scale": "model-00015-of-00033.safetensors",
471
+ "model.llm.layers.40.attn.q_norm.weight": "model-00015-of-00033.safetensors",
472
+ "model.llm.layers.40.attn.wk_dv.weight": "model-00015-of-00033.safetensors",
473
+ "model.llm.layers.41.mlp.gate.global_scale": "model-00015-of-00033.safetensors",
474
+ "model.llm.layers.41.mlp_norm.weight": "model-00015-of-00033.safetensors",
475
+ "model.llm.layers.5.attn.wv_dv.weight": "model-00015-of-00033.safetensors",
476
+ "model.llm.layers.6.mlp.gate.bias": "model-00015-of-00033.safetensors",
477
+ "model.llm.layers.7.attn.wo_ud.weight": "model-00015-of-00033.safetensors",
478
+ "model.llm.layers.9.attn.k_norm.weight": "model-00015-of-00033.safetensors",
479
+ "model.llm.layers.9.mlp.experts.w2_weight.weight_packed": "model-00015-of-00033.safetensors",
480
+ "model.llm.layers.9.mlp.experts.w2_weight.weight_scale": "model-00015-of-00033.safetensors",
481
+ "model.llm.layers.9.mlp.shared_experts.shared_w13_weight": "model-00015-of-00033.safetensors",
482
+ "model.llm.layers.9.mlp_norm.weight": "model-00015-of-00033.safetensors",
483
+ "model.visual.layers.linear_1.weight": "model-00015-of-00033.safetensors",
484
+ "model.visual.layers.linear_3.weight": "model-00015-of-00033.safetensors",
485
+ "model.visual.layers.norm_1.weight": "model-00015-of-00033.safetensors",
486
+ "model.llm.layers.0.mlp_sconv.weight": "model-00015-of-00033.safetensors",
487
+ "model.llm.layers.23.attn_sconv.weight": "model-00015-of-00033.safetensors",
488
+ "model.llm.layers.24.attn.v_sconv.weight": "model-00015-of-00033.safetensors",
489
+ "model.llm.layers.28.mlp.gate.weight": "model-00015-of-00033.safetensors",
490
+ "model.llm.layers.29.attn.v_sconv.weight": "model-00015-of-00033.safetensors",
491
+ "model.llm.layers.30.attn_sconv.weight": "model-00015-of-00033.safetensors",
492
+ "model.llm.layers.31.mlp.gate.weight": "model-00015-of-00033.safetensors",
493
+ "model.llm.layers.39.attn.v_sconv.weight": "model-00015-of-00033.safetensors",
494
+ "model.llm.layers.39.attn_sconv.weight": "model-00015-of-00033.safetensors",
495
+ "model.llm.layers.40.attn.v_sconv.weight": "model-00015-of-00033.safetensors",
496
+ "model.llm.layers.8.mlp.gate.weight": "model-00015-of-00033.safetensors",
497
+ "model.llm.layers.0.mlp.w2_md.weight": "model-00013-of-00033.safetensors",
498
+ "model.llm.layers.1.attn.rel_logits_proj.proj": "model-00013-of-00033.safetensors",
499
+ "model.llm.layers.10.attn.k_norm.weight": "model-00013-of-00033.safetensors",
500
+ "model.llm.layers.12.mlp.experts.w13_weight.weight_packed": "model-00013-of-00033.safetensors",
501
+ "model.llm.layers.12.mlp.experts.w13_weight.weight_scale": "model-00013-of-00033.safetensors",
502
+ "model.llm.layers.18.attn.wv_dv.weight": "model-00013-of-00033.safetensors",
503
+ "model.llm.layers.20.attn.wq_du.weight": "model-00013-of-00033.safetensors",
504
+ "model.llm.layers.20.mlp.shared_experts.shared_w2_weight": "model-00013-of-00033.safetensors",
505
+ "model.llm.layers.24.attn.wr_du.weight": "model-00013-of-00033.safetensors",
506
+ "model.llm.layers.25.attn.wk_dv.weight": "model-00013-of-00033.safetensors",
507
+ "model.llm.layers.26.mlp.shared_experts.shared_w13_weight": "model-00013-of-00033.safetensors",
508
+ "model.llm.layers.29.mlp.experts.w2_weight.weight_packed": "model-00013-of-00033.safetensors",
509
+ "model.llm.layers.29.mlp.experts.w2_weight.weight_scale": "model-00013-of-00033.safetensors",
510
+ "model.llm.layers.3.attn.wq_du.weight": "model-00013-of-00033.safetensors",
511
+ "model.llm.layers.33.attn.wq_du.weight": "model-00013-of-00033.safetensors",
512
+ "model.llm.layers.37.mlp.gate.global_scale": "model-00013-of-00033.safetensors",
513
+ "model.llm.layers.38.attn_norm.weight": "model-00013-of-00033.safetensors",
514
+ "model.llm.layers.39.attn.wo_ud.weight": "model-00013-of-00033.safetensors",
515
+ "model.llm.layers.40.mlp.experts.w2_weight.weight_packed": "model-00013-of-00033.safetensors",
516
+ "model.llm.layers.40.mlp.experts.w2_weight.weight_scale": "model-00013-of-00033.safetensors",
517
+ "model.llm.layers.40.mlp.shared_experts.shared_w13_weight": "model-00013-of-00033.safetensors",
518
+ "model.llm.layers.6.attn.q_norm.weight": "model-00013-of-00033.safetensors",
519
+ "model.llm.layers.6.mlp.gate.global_scale": "model-00013-of-00033.safetensors",
520
+ "model.llm.layers.9.attn_norm.weight": "model-00013-of-00033.safetensors",
521
+ "model.llm.layers.14.attn.v_sconv.weight": "model-00013-of-00033.safetensors",
522
+ "model.llm.layers.33.mlp.gate.weight": "model-00013-of-00033.safetensors",
523
+ "model.llm.layers.4.attn.v_sconv.weight": "model-00013-of-00033.safetensors",
524
+ "model.llm.layers.6.attn_sconv.weight": "model-00013-of-00033.safetensors",
525
+ "model.llm.layers.0.mlp_norm.weight": "model-00014-of-00033.safetensors",
526
+ "model.llm.layers.1.attn.q_norm.weight": "model-00014-of-00033.safetensors",
527
+ "model.llm.layers.10.mlp.gate.bias": "model-00014-of-00033.safetensors",
528
+ "model.llm.layers.12.attn.wk_dv.weight": "model-00014-of-00033.safetensors",
529
+ "model.llm.layers.13.attn.wo_ud.weight": "model-00014-of-00033.safetensors",
530
+ "model.llm.layers.14.mlp.experts.w13_weight.weight_packed": "model-00014-of-00033.safetensors",
531
+ "model.llm.layers.14.mlp.experts.w13_weight.weight_scale": "model-00014-of-00033.safetensors",
532
+ "model.llm.layers.16.attn.wo_ud.weight": "model-00014-of-00033.safetensors",
533
+ "model.llm.layers.2.mlp.experts.w2_weight.weight_packed": "model-00014-of-00033.safetensors",
534
+ "model.llm.layers.2.mlp.experts.w2_weight.weight_scale": "model-00014-of-00033.safetensors",
535
+ "model.llm.layers.20.attn.q_norm.weight": "model-00014-of-00033.safetensors",
536
+ "model.llm.layers.23.attn.k_norm.weight": "model-00014-of-00033.safetensors",
537
+ "model.llm.layers.25.mlp.shared_experts.shared_w2_weight": "model-00014-of-00033.safetensors",
538
+ "model.llm.layers.28.attn.wk_dv.weight": "model-00014-of-00033.safetensors",
539
+ "model.llm.layers.29.attn.wr_du.weight": "model-00014-of-00033.safetensors",
540
+ "model.llm.layers.3.attn.rel_logits_proj.proj": "model-00014-of-00033.safetensors",
541
+ "model.llm.layers.31.mlp.shared_experts.shared_w13_weight": "model-00014-of-00033.safetensors",
542
+ "model.llm.layers.34.attn.wo_ud.weight": "model-00014-of-00033.safetensors",
543
+ "model.llm.layers.36.attn_norm.weight": "model-00014-of-00033.safetensors",
544
+ "model.llm.layers.38.attn.wo_ud.weight": "model-00014-of-00033.safetensors",
545
+ "model.llm.layers.38.attn.wr_du.weight": "model-00014-of-00033.safetensors",
546
+ "model.llm.layers.38.mlp.experts.w2_weight.weight_packed": "model-00014-of-00033.safetensors",
547
+ "model.llm.layers.38.mlp.experts.w2_weight.weight_scale": "model-00014-of-00033.safetensors",
548
+ "model.llm.layers.38.mlp.shared_experts.shared_w2_weight": "model-00014-of-00033.safetensors",
549
+ "model.llm.layers.4.attn.wr_du.weight": "model-00014-of-00033.safetensors",
550
+ "model.llm.layers.4.attn.wv_dv.weight": "model-00014-of-00033.safetensors",
551
+ "model.llm.layers.7.attn.wv_dv.weight": "model-00014-of-00033.safetensors",
552
+ "model.llm.layers.8.mlp.gate.global_scale": "model-00014-of-00033.safetensors",
553
+ "model.llm.layers.9.mlp.shared_experts.shared_w2_weight": "model-00014-of-00033.safetensors",
554
+ "model.llm.layers.1.mlp_sconv.weight": "model-00014-of-00033.safetensors",
555
+ "model.llm.layers.12.attn.v_sconv.weight": "model-00014-of-00033.safetensors",
556
+ "model.llm.layers.19.mlp_sconv.weight": "model-00014-of-00033.safetensors",
557
+ "model.llm.layers.20.attn.v_sconv.weight": "model-00014-of-00033.safetensors",
558
+ "model.llm.layers.21.mlp.gate.weight": "model-00014-of-00033.safetensors",
559
+ "model.llm.layers.23.mlp_sconv.weight": "model-00014-of-00033.safetensors",
560
+ "model.llm.layers.28.attn_sconv.weight": "model-00014-of-00033.safetensors",
561
+ "model.llm.layers.41.attn.k_sconv.weight": "model-00014-of-00033.safetensors",
562
+ "model.llm.layers.6.attn.v_sconv.weight": "model-00014-of-00033.safetensors",
563
+ "model.llm.layers.1.attn.wr_du.weight": "model-00016-of-00033.safetensors",
564
+ "model.llm.layers.11.attn.wo_ud.weight": "model-00016-of-00033.safetensors",
565
+ "model.llm.layers.13.attn.k_norm.weight": "model-00016-of-00033.safetensors",
566
+ "model.llm.layers.16.mlp_norm.weight": "model-00016-of-00033.safetensors",
567
+ "model.llm.layers.18.attn.wr_du.weight": "model-00016-of-00033.safetensors",
568
+ "model.llm.layers.2.attn.rel_logits_proj.proj": "model-00016-of-00033.safetensors",
569
+ "model.llm.layers.21.attn.q_norm.weight": "model-00016-of-00033.safetensors",
570
+ "model.llm.layers.21.mlp.shared_experts.shared_w2_weight": "model-00016-of-00033.safetensors",
571
+ "model.llm.layers.23.attn_norm.weight": "model-00016-of-00033.safetensors",
572
+ "model.llm.layers.23.mlp.experts.w2_weight.weight_packed": "model-00016-of-00033.safetensors",
573
+ "model.llm.layers.23.mlp.experts.w2_weight.weight_scale": "model-00016-of-00033.safetensors",
574
+ "model.llm.layers.23.mlp.shared_experts.shared_w13_weight": "model-00016-of-00033.safetensors",
575
+ "model.llm.layers.24.attn.q_norm.weight": "model-00016-of-00033.safetensors",
576
+ "model.llm.layers.26.attn.wo_ud.weight": "model-00016-of-00033.safetensors",
577
+ "model.llm.layers.26.attn.wv_dv.weight": "model-00016-of-00033.safetensors",
578
+ "model.llm.layers.27.mlp_norm.weight": "model-00016-of-00033.safetensors",
579
+ "model.llm.layers.29.attn_norm.weight": "model-00016-of-00033.safetensors",
580
+ "model.llm.layers.30.mlp.experts.w13_weight.weight_packed": "model-00016-of-00033.safetensors",
581
+ "model.llm.layers.30.mlp.experts.w13_weight.weight_scale": "model-00016-of-00033.safetensors",
582
+ "model.llm.layers.31.attn.wq_du.weight": "model-00016-of-00033.safetensors",
583
+ "model.llm.layers.34.mlp.shared_experts.shared_w13_weight": "model-00016-of-00033.safetensors",
584
+ "model.llm.layers.35.attn.rel_logits_proj.proj": "model-00016-of-00033.safetensors",
585
+ "model.llm.layers.36.attn.wq_du.weight": "model-00016-of-00033.safetensors",
586
+ "model.llm.layers.37.attn.wq_du.weight": "model-00016-of-00033.safetensors",
587
+ "model.llm.layers.38.mlp.shared_experts.shared_w13_weight": "model-00016-of-00033.safetensors",
588
+ "model.llm.layers.39.attn.k_norm.weight": "model-00016-of-00033.safetensors",
589
+ "model.llm.layers.39.attn.wq_du.weight": "model-00016-of-00033.safetensors",
590
+ "model.llm.layers.4.attn.q_norm.weight": "model-00016-of-00033.safetensors",
591
+ "model.llm.layers.4.mlp.shared_experts.shared_w2_weight": "model-00016-of-00033.safetensors",
592
+ "model.llm.layers.40.attn.wq_du.weight": "model-00016-of-00033.safetensors",
593
+ "model.llm.layers.5.attn.wr_du.weight": "model-00016-of-00033.safetensors",
594
+ "model.llm.layers.6.attn.k_norm.weight": "model-00016-of-00033.safetensors",
595
+ "model.llm.layers.6.mlp_norm.weight": "model-00016-of-00033.safetensors",
596
+ "model.llm.layers.7.attn.k_norm.weight": "model-00016-of-00033.safetensors",
597
+ "model.llm.layers.8.attn.wv_dv.weight": "model-00016-of-00033.safetensors",
598
+ "model.llm.layers.1.attn.k_sconv.weight": "model-00016-of-00033.safetensors",
599
+ "model.llm.layers.16.mlp.gate.weight": "model-00016-of-00033.safetensors",
600
+ "model.llm.layers.18.mlp_sconv.weight": "model-00016-of-00033.safetensors",
601
+ "model.llm.layers.19.attn_sconv.weight": "model-00016-of-00033.safetensors",
602
+ "model.llm.layers.20.attn.k_sconv.weight": "model-00016-of-00033.safetensors",
603
+ "model.llm.layers.27.attn.k_sconv.weight": "model-00016-of-00033.safetensors",
604
+ "model.llm.layers.27.attn.v_sconv.weight": "model-00016-of-00033.safetensors",
605
+ "model.llm.layers.28.mlp_sconv.weight": "model-00016-of-00033.safetensors",
606
+ "model.llm.layers.3.attn.v_sconv.weight": "model-00016-of-00033.safetensors",
607
+ "model.llm.layers.33.mlp_sconv.weight": "model-00016-of-00033.safetensors",
608
+ "model.llm.layers.40.attn.k_sconv.weight": "model-00016-of-00033.safetensors",
609
+ "model.llm.layers.40.mlp.gate.weight": "model-00016-of-00033.safetensors",
610
+ "model.llm.layers.9.attn.v_sconv.weight": "model-00016-of-00033.safetensors",
611
+ "model.llm.layers.9.mlp.gate.global_scale": "model-00016-of-00033.safetensors",
612
+ "model.llm.layers.9.mlp.gate.weight": "model-00016-of-00033.safetensors",
613
+ "model.llm.layers.1.attn.wk_dv.weight": "model-00017-of-00033.safetensors",
614
+ "model.llm.layers.1.mlp.w2_md.weight": "model-00017-of-00033.safetensors",
615
+ "model.llm.layers.10.attn.rel_logits_proj.proj": "model-00017-of-00033.safetensors",
616
+ "model.llm.layers.10.attn.wk_dv.weight": "model-00017-of-00033.safetensors",
617
+ "model.llm.layers.17.attn.wr_du.weight": "model-00017-of-00033.safetensors",
618
+ "model.llm.layers.22.attn.wk_dv.weight": "model-00017-of-00033.safetensors",
619
+ "model.llm.layers.24.attn_norm.weight": "model-00017-of-00033.safetensors",
620
+ "model.llm.layers.24.mlp.gate.global_scale": "model-00017-of-00033.safetensors",
621
+ "model.llm.layers.33.mlp.experts.w13_weight.weight_packed": "model-00017-of-00033.safetensors",
622
+ "model.llm.layers.33.mlp.experts.w13_weight.weight_scale": "model-00017-of-00033.safetensors",
623
+ "model.llm.layers.35.attn_norm.weight": "model-00017-of-00033.safetensors",
624
+ "model.llm.layers.35.mlp.experts.w2_weight.weight_packed": "model-00017-of-00033.safetensors",
625
+ "model.llm.layers.35.mlp.experts.w2_weight.weight_scale": "model-00017-of-00033.safetensors",
626
+ "model.llm.layers.36.mlp_norm.weight": "model-00017-of-00033.safetensors",
627
+ "model.llm.layers.40.mlp_norm.weight": "model-00017-of-00033.safetensors",
628
+ "model.llm.layers.41.attn.wr_du.weight": "model-00017-of-00033.safetensors",
629
+ "model.llm.layers.41.attn.wv_dv.weight": "model-00017-of-00033.safetensors",
630
+ "model.llm.layers.41.mlp.shared_experts.shared_w13_weight": "model-00017-of-00033.safetensors",
631
+ "model.llm.layers.8.attn.wk_dv.weight": "model-00017-of-00033.safetensors",
632
+ "model.visual.layers.linear_0.weight": "model-00017-of-00033.safetensors",
633
+ "model.llm.layers.24.attn_sconv.weight": "model-00017-of-00033.safetensors",
634
+ "model.llm.layers.26.mlp_sconv.weight": "model-00017-of-00033.safetensors",
635
+ "model.llm.layers.34.mlp_sconv.weight": "model-00017-of-00033.safetensors",
636
+ "model.llm.layers.38.mlp.gate.weight": "model-00017-of-00033.safetensors",
637
+ "model.llm.layers.39.mlp_sconv.weight": "model-00017-of-00033.safetensors",
638
+ "model.llm.layers.40.mlp_sconv.weight": "model-00017-of-00033.safetensors",
639
+ "model.llm.layers.1.mlp.global_scale": "model-00019-of-00033.safetensors",
640
+ "model.llm.layers.11.mlp.gate.global_scale": "model-00019-of-00033.safetensors",
641
+ "model.llm.layers.12.attn.q_norm.weight": "model-00019-of-00033.safetensors",
642
+ "model.llm.layers.12.mlp.experts.w2_weight.weight_packed": "model-00019-of-00033.safetensors",
643
+ "model.llm.layers.12.mlp.experts.w2_weight.weight_scale": "model-00019-of-00033.safetensors",
644
+ "model.llm.layers.12.mlp.shared_experts.shared_w13_weight": "model-00019-of-00033.safetensors",
645
+ "model.llm.layers.19.mlp.experts.w13_weight.weight_packed": "model-00019-of-00033.safetensors",
646
+ "model.llm.layers.19.mlp.experts.w13_weight.weight_scale": "model-00019-of-00033.safetensors",
647
+ "model.llm.layers.22.attn_norm.weight": "model-00019-of-00033.safetensors",
648
+ "model.llm.layers.29.attn.q_norm.weight": "model-00019-of-00033.safetensors",
649
+ "model.llm.layers.29.mlp.shared_experts.shared_w2_weight": "model-00019-of-00033.safetensors",
650
+ "model.llm.layers.31.attn.wk_dv.weight": "model-00019-of-00033.safetensors",
651
+ "model.llm.layers.31.attn.wv_dv.weight": "model-00019-of-00033.safetensors",
652
+ "model.llm.layers.32.attn.wk_dv.weight": "model-00019-of-00033.safetensors",
653
+ "model.llm.layers.32.mlp.shared_experts.shared_w13_weight": "model-00019-of-00033.safetensors",
654
+ "model.llm.layers.34.attn.q_norm.weight": "model-00019-of-00033.safetensors",
655
+ "model.llm.layers.37.mlp.gate.bias": "model-00019-of-00033.safetensors",
656
+ "model.llm.layers.7.attn_sconv.weight": "model-00019-of-00033.safetensors",
657
+ "model.llm.layers.1.attn.wq_du.weight": "model-00018-of-00033.safetensors",
658
+ "model.llm.layers.10.mlp.experts.w2_weight.weight_packed": "model-00018-of-00033.safetensors",
659
+ "model.llm.layers.10.mlp.experts.w2_weight.weight_scale": "model-00018-of-00033.safetensors",
660
+ "model.llm.layers.13.attn_norm.weight": "model-00018-of-00033.safetensors",
661
+ "model.llm.layers.13.mlp.experts.w13_weight.weight_packed": "model-00018-of-00033.safetensors",
662
+ "model.llm.layers.13.mlp.experts.w13_weight.weight_scale": "model-00018-of-00033.safetensors",
663
+ "model.llm.layers.13.mlp.experts.w2_weight.weight_packed": "model-00018-of-00033.safetensors",
664
+ "model.llm.layers.13.mlp.experts.w2_weight.weight_scale": "model-00018-of-00033.safetensors",
665
+ "model.llm.layers.14.mlp.gate.bias": "model-00018-of-00033.safetensors",
666
+ "model.llm.layers.17.mlp.gate.bias": "model-00018-of-00033.safetensors",
667
+ "model.llm.layers.18.mlp.gate.global_scale": "model-00018-of-00033.safetensors",
668
+ "model.llm.layers.18.mlp.shared_experts.shared_w13_weight": "model-00018-of-00033.safetensors",
669
+ "model.llm.layers.24.attn.wo_ud.weight": "model-00018-of-00033.safetensors",
670
+ "model.llm.layers.33.attn.q_norm.weight": "model-00018-of-00033.safetensors",
671
+ "model.llm.layers.11.mlp_sconv.weight": "model-00018-of-00033.safetensors",
672
+ "model.llm.layers.15.attn_sconv.weight": "model-00018-of-00033.safetensors",
673
+ "model.llm.layers.20.mlp.gate.weight": "model-00018-of-00033.safetensors",
674
+ "model.llm.layers.21.attn.k_sconv.weight": "model-00018-of-00033.safetensors",
675
+ "model.llm.layers.21.mlp_sconv.weight": "model-00018-of-00033.safetensors",
676
+ "model.llm.layers.25.mlp.gate.weight": "model-00018-of-00033.safetensors",
677
+ "model.llm.layers.30.attn.k_sconv.weight": "model-00018-of-00033.safetensors",
678
+ "model.llm.layers.11.attn.rel_logits_proj.proj": "model-00021-of-00033.safetensors",
679
+ "model.llm.layers.13.mlp.shared_experts.shared_w2_weight": "model-00021-of-00033.safetensors",
680
+ "model.llm.layers.16.attn.q_norm.weight": "model-00021-of-00033.safetensors",
681
+ "model.llm.layers.17.attn.wk_dv.weight": "model-00021-of-00033.safetensors",
682
+ "model.llm.layers.17.attn.wv_dv.weight": "model-00021-of-00033.safetensors",
683
+ "model.llm.layers.22.attn.wr_du.weight": "model-00021-of-00033.safetensors",
684
+ "model.llm.layers.3.attn.q_norm.weight": "model-00021-of-00033.safetensors",
685
+ "model.llm.layers.33.attn_norm.weight": "model-00021-of-00033.safetensors",
686
+ "model.llm.layers.36.attn.q_norm.weight": "model-00021-of-00033.safetensors",
687
+ "model.llm.layers.39.mlp.experts.w13_weight.weight_packed": "model-00021-of-00033.safetensors",
688
+ "model.llm.layers.39.mlp.experts.w13_weight.weight_scale": "model-00021-of-00033.safetensors",
689
+ "model.llm.layers.4.mlp.experts.w2_weight.weight_packed": "model-00021-of-00033.safetensors",
690
+ "model.llm.layers.4.mlp.experts.w2_weight.weight_scale": "model-00021-of-00033.safetensors",
691
+ "model.llm.layers.40.mlp.gate.global_scale": "model-00021-of-00033.safetensors",
692
+ "model.llm.layers.41.attn.k_norm.weight": "model-00021-of-00033.safetensors",
693
+ "model.llm.layers.10.attn.k_sconv.weight": "model-00021-of-00033.safetensors",
694
+ "model.llm.layers.17.attn.v_sconv.weight": "model-00021-of-00033.safetensors",
695
+ "model.llm.layers.1.mlp.w13_dn.weight": "model-00020-of-00033.safetensors",
696
+ "model.llm.layers.12.attn.k_norm.weight": "model-00020-of-00033.safetensors",
697
+ "model.llm.layers.12.attn.wr_du.weight": "model-00020-of-00033.safetensors",
698
+ "model.llm.layers.15.mlp.experts.w2_weight.weight_packed": "model-00020-of-00033.safetensors",
699
+ "model.llm.layers.15.mlp.experts.w2_weight.weight_scale": "model-00020-of-00033.safetensors",
700
+ "model.llm.layers.24.attn.k_norm.weight": "model-00020-of-00033.safetensors",
701
+ "model.llm.layers.25.attn.wr_du.weight": "model-00020-of-00033.safetensors",
702
+ "model.llm.layers.26.mlp.experts.w2_weight.weight_packed": "model-00020-of-00033.safetensors",
703
+ "model.llm.layers.26.mlp.experts.w2_weight.weight_scale": "model-00020-of-00033.safetensors",
704
+ "model.llm.layers.27.attn.rel_logits_proj.proj": "model-00020-of-00033.safetensors",
705
+ "model.llm.layers.28.mlp_norm.weight": "model-00020-of-00033.safetensors",
706
+ "model.llm.layers.35.attn.wv_dv.weight": "model-00020-of-00033.safetensors",
707
+ "model.llm.layers.37.mlp_norm.weight": "model-00020-of-00033.safetensors",
708
+ "model.llm.layers.39.mlp.gate.bias": "model-00020-of-00033.safetensors",
709
+ "model.llm.layers.4.attn.k_norm.weight": "model-00020-of-00033.safetensors",
710
+ "model.llm.layers.40.attn_norm.weight": "model-00020-of-00033.safetensors",
711
+ "model.llm.layers.6.attn_norm.weight": "model-00020-of-00033.safetensors",
712
+ "model.llm.layers.7.mlp.experts.w13_weight.weight_packed": "model-00020-of-00033.safetensors",
713
+ "model.llm.layers.7.mlp.experts.w13_weight.weight_scale": "model-00020-of-00033.safetensors",
714
+ "model.llm.layers.10.attn_sconv.weight": "model-00020-of-00033.safetensors",
715
+ "model.llm.layers.21.attn.v_sconv.weight": "model-00020-of-00033.safetensors",
716
+ "model.llm.layers.31.attn_sconv.weight": "model-00020-of-00033.safetensors",
717
+ "model.llm.layers.36.mlp.gate.weight": "model-00020-of-00033.safetensors",
718
+ "model.llm.layers.6.attn.k_sconv.weight": "model-00020-of-00033.safetensors",
719
+ "model.llm.layers.10.attn.wo_ud.weight": "model-00022-of-00033.safetensors",
720
+ "model.llm.layers.10.mlp.shared_experts.shared_w2_weight": "model-00022-of-00033.safetensors",
721
+ "model.llm.layers.11.attn.k_norm.weight": "model-00022-of-00033.safetensors",
722
+ "model.llm.layers.11.attn_norm.weight": "model-00022-of-00033.safetensors",
723
+ "model.llm.layers.14.attn.wr_du.weight": "model-00022-of-00033.safetensors",
724
+ "model.llm.layers.14.mlp_norm.weight": "model-00022-of-00033.safetensors",
725
+ "model.llm.layers.15.attn.k_norm.weight": "model-00022-of-00033.safetensors",
726
+ "model.llm.layers.15.attn.wv_dv.weight": "model-00022-of-00033.safetensors",
727
+ "model.llm.layers.18.mlp.experts.w13_weight.weight_packed": "model-00022-of-00033.safetensors",
728
+ "model.llm.layers.18.mlp.experts.w13_weight.weight_scale": "model-00022-of-00033.safetensors",
729
+ "model.llm.layers.19.attn.rel_logits_proj.proj": "model-00022-of-00033.safetensors",
730
+ "model.llm.layers.19.mlp.gate.global_scale": "model-00022-of-00033.safetensors",
731
+ "model.llm.layers.2.attn.wk_dv.weight": "model-00022-of-00033.safetensors",
732
+ "model.llm.layers.20.attn_norm.weight": "model-00022-of-00033.safetensors",
733
+ "model.llm.layers.23.attn.rel_logits_proj.proj": "model-00022-of-00033.safetensors",
734
+ "model.llm.layers.24.attn.wk_dv.weight": "model-00022-of-00033.safetensors",
735
+ "model.llm.layers.24.mlp.experts.w2_weight.weight_packed": "model-00022-of-00033.safetensors",
736
+ "model.llm.layers.24.mlp.experts.w2_weight.weight_scale": "model-00022-of-00033.safetensors",
737
+ "model.llm.layers.25.attn.wv_dv.weight": "model-00022-of-00033.safetensors",
738
+ "model.llm.layers.25.mlp.gate.bias": "model-00022-of-00033.safetensors",
739
+ "model.llm.layers.26.attn.wr_du.weight": "model-00022-of-00033.safetensors",
740
+ "model.llm.layers.26.mlp.gate.bias": "model-00022-of-00033.safetensors",
741
+ "model.llm.layers.28.attn_norm.weight": "model-00022-of-00033.safetensors",
742
+ "model.llm.layers.28.mlp.shared_experts.shared_w2_weight": "model-00022-of-00033.safetensors",
743
+ "model.llm.layers.29.attn.wk_dv.weight": "model-00022-of-00033.safetensors",
744
+ "model.llm.layers.29.mlp_norm.weight": "model-00022-of-00033.safetensors",
745
+ "model.llm.layers.31.mlp.gate.bias": "model-00022-of-00033.safetensors",
746
+ "model.llm.layers.31.mlp.gate.global_scale": "model-00022-of-00033.safetensors",
747
+ "model.llm.layers.32.attn.wr_du.weight": "model-00022-of-00033.safetensors",
748
+ "model.llm.layers.32.attn.wv_dv.weight": "model-00022-of-00033.safetensors",
749
+ "model.llm.layers.33.mlp.shared_experts.shared_w13_weight": "model-00022-of-00033.safetensors",
750
+ "model.llm.layers.34.mlp.gate.bias": "model-00022-of-00033.safetensors",
751
+ "model.llm.layers.35.attn.k_norm.weight": "model-00022-of-00033.safetensors",
752
+ "model.llm.layers.35.attn.wo_ud.weight": "model-00022-of-00033.safetensors",
753
+ "model.llm.layers.35.mlp.gate.global_scale": "model-00022-of-00033.safetensors",
754
+ "model.llm.layers.35.mlp.shared_experts.shared_w13_weight": "model-00022-of-00033.safetensors",
755
+ "model.llm.layers.40.attn.rel_logits_proj.proj": "model-00022-of-00033.safetensors",
756
+ "model.llm.layers.41.mlp.shared_experts.shared_w2_weight": "model-00022-of-00033.safetensors",
757
+ "model.llm.layers.5.attn.wk_dv.weight": "model-00022-of-00033.safetensors",
758
+ "model.llm.layers.5.mlp.experts.w2_weight.weight_packed": "model-00022-of-00033.safetensors",
759
+ "model.llm.layers.5.mlp.experts.w2_weight.weight_scale": "model-00022-of-00033.safetensors",
760
+ "model.llm.layers.6.attn.wk_dv.weight": "model-00022-of-00033.safetensors",
761
+ "model.llm.layers.6.attn.wq_du.weight": "model-00022-of-00033.safetensors",
762
+ "model.llm.layers.6.attn.wv_dv.weight": "model-00022-of-00033.safetensors",
763
+ "model.llm.layers.7.mlp.gate.bias": "model-00022-of-00033.safetensors",
764
+ "model.llm.layers.8.mlp.gate.bias": "model-00022-of-00033.safetensors",
765
+ "model.llm.layers.9.attn.wq_du.weight": "model-00022-of-00033.safetensors",
766
+ "model.llm.layers.16.attn_sconv.weight": "model-00022-of-00033.safetensors",
767
+ "model.llm.layers.24.mlp_sconv.weight": "model-00022-of-00033.safetensors",
768
+ "model.llm.layers.29.attn.k_sconv.weight": "model-00022-of-00033.safetensors",
769
+ "model.llm.layers.38.attn.k_sconv.weight": "model-00022-of-00033.safetensors",
770
+ "model.llm.layers.9.attn.k_sconv.weight": "model-00022-of-00033.safetensors",
771
+ "model.llm.layers.10.attn.wv_dv.weight": "model-00023-of-00033.safetensors",
772
+ "model.llm.layers.10.mlp.gate.global_scale": "model-00023-of-00033.safetensors",
773
+ "model.llm.layers.10.mlp.shared_experts.shared_w13_weight": "model-00023-of-00033.safetensors",
774
+ "model.llm.layers.11.attn.wr_du.weight": "model-00023-of-00033.safetensors",
775
+ "model.llm.layers.12.attn.wv_dv.weight": "model-00023-of-00033.safetensors",
776
+ "model.llm.layers.20.mlp.experts.w13_weight.weight_packed": "model-00023-of-00033.safetensors",
777
+ "model.llm.layers.20.mlp.experts.w13_weight.weight_scale": "model-00023-of-00033.safetensors",
778
+ "model.llm.layers.21.attn.wr_du.weight": "model-00023-of-00033.safetensors",
779
+ "model.llm.layers.23.mlp.shared_experts.shared_w2_weight": "model-00023-of-00033.safetensors",
780
+ "model.llm.layers.25.attn.q_norm.weight": "model-00023-of-00033.safetensors",
781
+ "model.llm.layers.26.mlp.shared_experts.shared_w2_weight": "model-00023-of-00033.safetensors",
782
+ "model.llm.layers.26.mlp_norm.weight": "model-00023-of-00033.safetensors",
783
+ "model.llm.layers.3.attn.wv_dv.weight": "model-00023-of-00033.safetensors",
784
+ "model.llm.layers.32.attn.k_norm.weight": "model-00023-of-00033.safetensors",
785
+ "model.llm.layers.37.attn.q_norm.weight": "model-00023-of-00033.safetensors",
786
+ "model.llm.layers.39.mlp.gate.global_scale": "model-00023-of-00033.safetensors",
787
+ "model.llm.layers.41.attn.q_norm.weight": "model-00023-of-00033.safetensors",
788
+ "model.llm.layers.41.mlp.gate.bias": "model-00023-of-00033.safetensors",
789
+ "model.llm.layers.7.attn.wk_dv.weight": "model-00023-of-00033.safetensors",
790
+ "model.llm.layers.7.mlp.shared_experts.shared_w2_weight": "model-00023-of-00033.safetensors",
791
+ "model.llm.layers.8.mlp.shared_experts.shared_w13_weight": "model-00023-of-00033.safetensors",
792
+ "model.llm.layers.9.mlp.experts.w13_weight.weight_packed": "model-00023-of-00033.safetensors",
793
+ "model.llm.layers.9.mlp.experts.w13_weight.weight_scale": "model-00023-of-00033.safetensors",
794
+ "model.llm.layers.22.mlp.gate.weight": "model-00023-of-00033.safetensors",
795
+ "model.llm.layers.23.attn.v_sconv.weight": "model-00023-of-00033.safetensors",
796
+ "model.llm.layers.32.attn_sconv.weight": "model-00023-of-00033.safetensors",
797
+ "model.llm.layers.33.attn.k_sconv.weight": "model-00023-of-00033.safetensors",
798
+ "model.llm.layers.5.mlp_sconv.weight": "model-00023-of-00033.safetensors",
799
+ "model.llm.layers.6.mlp.gate.weight": "model-00023-of-00033.safetensors",
800
+ "model.llm.layers.7.mlp_sconv.weight": "model-00023-of-00033.safetensors",
801
+ "model.llm.layers.10.attn_norm.weight": "model-00024-of-00033.safetensors",
802
+ "model.llm.layers.11.attn.wv_dv.weight": "model-00024-of-00033.safetensors",
803
+ "model.llm.layers.11.mlp.gate.bias": "model-00024-of-00033.safetensors",
804
+ "model.llm.layers.12.attn_norm.weight": "model-00024-of-00033.safetensors",
805
+ "model.llm.layers.13.mlp.shared_experts.shared_w13_weight": "model-00024-of-00033.safetensors",
806
+ "model.llm.layers.14.mlp.gate.global_scale": "model-00024-of-00033.safetensors",
807
+ "model.llm.layers.16.mlp.shared_experts.shared_w13_weight": "model-00024-of-00033.safetensors",
808
+ "model.llm.layers.17.mlp.experts.w13_weight.weight_packed": "model-00024-of-00033.safetensors",
809
+ "model.llm.layers.17.mlp.experts.w13_weight.weight_scale": "model-00024-of-00033.safetensors",
810
+ "model.llm.layers.19.attn.wk_dv.weight": "model-00024-of-00033.safetensors",
811
+ "model.llm.layers.19.attn.wr_du.weight": "model-00024-of-00033.safetensors",
812
+ "model.llm.layers.2.attn_norm.weight": "model-00024-of-00033.safetensors",
813
+ "model.llm.layers.28.attn.wv_dv.weight": "model-00024-of-00033.safetensors",
814
+ "model.llm.layers.3.mlp.gate.bias": "model-00024-of-00033.safetensors",
815
+ "model.llm.layers.31.attn_norm.weight": "model-00024-of-00033.safetensors",
816
+ "model.llm.layers.32.mlp.experts.w13_weight.weight_packed": "model-00024-of-00033.safetensors",
817
+ "model.llm.layers.32.mlp.experts.w13_weight.weight_scale": "model-00024-of-00033.safetensors",
818
+ "model.llm.layers.32.mlp.shared_experts.shared_w2_weight": "model-00024-of-00033.safetensors",
819
+ "model.llm.layers.33.attn.wk_dv.weight": "model-00024-of-00033.safetensors",
820
+ "model.llm.layers.33.attn.wo_ud.weight": "model-00024-of-00033.safetensors",
821
+ "model.llm.layers.34.attn.rel_logits_proj.proj": "model-00024-of-00033.safetensors",
822
+ "model.llm.layers.35.mlp.gate.bias": "model-00024-of-00033.safetensors",
823
+ "model.llm.layers.35.mlp.shared_experts.shared_w2_weight": "model-00024-of-00033.safetensors",
824
+ "model.llm.layers.37.attn.rel_logits_proj.proj": "model-00024-of-00033.safetensors",
825
+ "model.llm.layers.39.mlp_norm.weight": "model-00024-of-00033.safetensors",
826
+ "model.llm.layers.40.attn.wv_dv.weight": "model-00024-of-00033.safetensors",
827
+ "model.llm.layers.40.mlp.gate.bias": "model-00024-of-00033.safetensors",
828
+ "model.llm.layers.6.mlp.shared_experts.shared_w13_weight": "model-00024-of-00033.safetensors",
829
+ "model.llm.layers.8.mlp_norm.weight": "model-00024-of-00033.safetensors",
830
+ "model.llm.layers.9.attn.q_norm.weight": "model-00024-of-00033.safetensors",
831
+ "model.llm.layers.12.mlp.gate.global_scale": "model-00024-of-00033.safetensors",
832
+ "model.llm.layers.12.mlp.gate.weight": "model-00024-of-00033.safetensors",
833
+ "model.llm.layers.14.mlp_sconv.weight": "model-00024-of-00033.safetensors",
834
+ "model.llm.layers.16.attn.k_sconv.weight": "model-00024-of-00033.safetensors",
835
+ "model.llm.layers.17.mlp.gate.weight": "model-00024-of-00033.safetensors",
836
+ "model.llm.layers.2.attn.k_sconv.weight": "model-00024-of-00033.safetensors",
837
+ "model.llm.layers.23.attn.k_sconv.weight": "model-00024-of-00033.safetensors",
838
+ "model.llm.layers.40.attn_sconv.weight": "model-00024-of-00033.safetensors",
839
+ "model.llm.layers.5.attn.v_sconv.weight": "model-00024-of-00033.safetensors",
840
+ "model.llm.layers.10.mlp.experts.w13_weight.weight_packed": "model-00025-of-00033.safetensors",
841
+ "model.llm.layers.10.mlp.experts.w13_weight.weight_scale": "model-00025-of-00033.safetensors",
842
+ "model.llm.layers.15.attn.wo_ud.weight": "model-00025-of-00033.safetensors",
843
+ "model.llm.layers.23.mlp.experts.w13_weight.weight_packed": "model-00025-of-00033.safetensors",
844
+ "model.llm.layers.23.mlp.experts.w13_weight.weight_scale": "model-00025-of-00033.safetensors",
845
+ "model.llm.layers.3.mlp.gate.global_scale": "model-00025-of-00033.safetensors",
846
+ "model.llm.layers.31.attn.wo_ud.weight": "model-00025-of-00033.safetensors",
847
+ "model.llm.layers.30.mlp.gate.weight": "model-00025-of-00033.safetensors",
848
+ "model.llm.layers.36.attn_sconv.weight": "model-00025-of-00033.safetensors",
849
+ "model.llm.layers.5.attn_sconv.weight": "model-00025-of-00033.safetensors",
850
+ "model.mtp.layers.0.hidden_norm.weight": "model-00033-of-00033.safetensors",
851
+ "model.mtp.layers.0.input_proj.weight": "model-00033-of-00033.safetensors",
852
+ "model.mtp.layers.0.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
853
+ "model.mtp.layers.0.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
854
+ "model.mtp.layers.0.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
855
+ "model.mtp.layers.0.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
856
+ "model.mtp.layers.0.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
857
+ "model.mtp.layers.0.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
858
+ "model.mtp.layers.0.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
859
+ "model.mtp.layers.0.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
860
+ "model.mtp.layers.0.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
861
+ "model.mtp.layers.0.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
862
+ "model.mtp.layers.0.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
863
+ "model.mtp.layers.0.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
864
+ "model.mtp.layers.0.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
865
+ "model.mtp.layers.1.hidden_norm.weight": "model-00033-of-00033.safetensors",
866
+ "model.mtp.layers.1.input_proj.weight": "model-00033-of-00033.safetensors",
867
+ "model.mtp.layers.1.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
868
+ "model.mtp.layers.1.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
869
+ "model.mtp.layers.1.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
870
+ "model.mtp.layers.1.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
871
+ "model.mtp.layers.1.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
872
+ "model.mtp.layers.1.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
873
+ "model.mtp.layers.1.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
874
+ "model.mtp.layers.1.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
875
+ "model.mtp.layers.1.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
876
+ "model.mtp.layers.1.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
877
+ "model.mtp.layers.1.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
878
+ "model.mtp.layers.1.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
879
+ "model.mtp.layers.1.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
880
+ "model.mtp.layers.2.hidden_norm.weight": "model-00033-of-00033.safetensors",
881
+ "model.mtp.layers.2.input_proj.weight": "model-00033-of-00033.safetensors",
882
+ "model.mtp.layers.2.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
883
+ "model.mtp.layers.2.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
884
+ "model.mtp.layers.2.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
885
+ "model.mtp.layers.2.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
886
+ "model.mtp.layers.2.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
887
+ "model.mtp.layers.2.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
888
+ "model.mtp.layers.2.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
889
+ "model.mtp.layers.2.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
890
+ "model.mtp.layers.2.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
891
+ "model.mtp.layers.2.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
892
+ "model.mtp.layers.2.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
893
+ "model.mtp.layers.2.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
894
+ "model.mtp.layers.2.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
895
+ "model.mtp.layers.3.hidden_norm.weight": "model-00033-of-00033.safetensors",
896
+ "model.mtp.layers.3.input_proj.weight": "model-00033-of-00033.safetensors",
897
+ "model.mtp.layers.3.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
898
+ "model.mtp.layers.3.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
899
+ "model.mtp.layers.3.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
900
+ "model.mtp.layers.3.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
901
+ "model.mtp.layers.3.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
902
+ "model.mtp.layers.3.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
903
+ "model.mtp.layers.3.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
904
+ "model.mtp.layers.3.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
905
+ "model.mtp.layers.3.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
906
+ "model.mtp.layers.3.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
907
+ "model.mtp.layers.3.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
908
+ "model.mtp.layers.3.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
909
+ "model.mtp.layers.3.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
910
+ "model.mtp.layers.4.hidden_norm.weight": "model-00033-of-00033.safetensors",
911
+ "model.mtp.layers.4.input_proj.weight": "model-00033-of-00033.safetensors",
912
+ "model.mtp.layers.4.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
913
+ "model.mtp.layers.4.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
914
+ "model.mtp.layers.4.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
915
+ "model.mtp.layers.4.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
916
+ "model.mtp.layers.4.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
917
+ "model.mtp.layers.4.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
918
+ "model.mtp.layers.4.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
919
+ "model.mtp.layers.4.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
920
+ "model.mtp.layers.4.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
921
+ "model.mtp.layers.4.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
922
+ "model.mtp.layers.4.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
923
+ "model.mtp.layers.4.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
924
+ "model.mtp.layers.4.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
925
+ "model.mtp.layers.5.hidden_norm.weight": "model-00033-of-00033.safetensors",
926
+ "model.mtp.layers.5.input_proj.weight": "model-00033-of-00033.safetensors",
927
+ "model.mtp.layers.5.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
928
+ "model.mtp.layers.5.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
929
+ "model.mtp.layers.5.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
930
+ "model.mtp.layers.5.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
931
+ "model.mtp.layers.5.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
932
+ "model.mtp.layers.5.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
933
+ "model.mtp.layers.5.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
934
+ "model.mtp.layers.5.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
935
+ "model.mtp.layers.5.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
936
+ "model.mtp.layers.5.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
937
+ "model.mtp.layers.5.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
938
+ "model.mtp.layers.5.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
939
+ "model.mtp.layers.5.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
940
+ "model.mtp.layers.6.hidden_norm.weight": "model-00033-of-00033.safetensors",
941
+ "model.mtp.layers.6.input_proj.weight": "model-00033-of-00033.safetensors",
942
+ "model.mtp.layers.6.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
943
+ "model.mtp.layers.6.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
944
+ "model.mtp.layers.6.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
945
+ "model.mtp.layers.6.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
946
+ "model.mtp.layers.6.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
947
+ "model.mtp.layers.6.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
948
+ "model.mtp.layers.6.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
949
+ "model.mtp.layers.6.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
950
+ "model.mtp.layers.6.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
951
+ "model.mtp.layers.6.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
952
+ "model.mtp.layers.6.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
953
+ "model.mtp.layers.6.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
954
+ "model.mtp.layers.6.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
955
+ "model.mtp.layers.7.hidden_norm.weight": "model-00033-of-00033.safetensors",
956
+ "model.mtp.layers.7.input_proj.weight": "model-00033-of-00033.safetensors",
957
+ "model.mtp.layers.7.transformer_block.attn.k_norm.weight": "model-00033-of-00033.safetensors",
958
+ "model.mtp.layers.7.transformer_block.attn.q_norm.weight": "model-00033-of-00033.safetensors",
959
+ "model.mtp.layers.7.transformer_block.attn.rel_logits_proj.proj": "model-00033-of-00033.safetensors",
960
+ "model.mtp.layers.7.transformer_block.attn.wk_dv.weight": "model-00033-of-00033.safetensors",
961
+ "model.mtp.layers.7.transformer_block.attn.wo_ud.weight": "model-00033-of-00033.safetensors",
962
+ "model.mtp.layers.7.transformer_block.attn.wq_du.weight": "model-00033-of-00033.safetensors",
963
+ "model.mtp.layers.7.transformer_block.attn.wr_du.weight": "model-00033-of-00033.safetensors",
964
+ "model.mtp.layers.7.transformer_block.attn.wv_dv.weight": "model-00033-of-00033.safetensors",
965
+ "model.mtp.layers.7.transformer_block.attn_norm.weight": "model-00033-of-00033.safetensors",
966
+ "model.mtp.layers.7.transformer_block.mlp.global_scale": "model-00033-of-00033.safetensors",
967
+ "model.mtp.layers.7.transformer_block.mlp.w13_dn.weight": "model-00033-of-00033.safetensors",
968
+ "model.mtp.layers.7.transformer_block.mlp.w2_md.weight": "model-00033-of-00033.safetensors",
969
+ "model.mtp.layers.7.transformer_block.mlp_norm.weight": "model-00033-of-00033.safetensors",
970
+ "model.mtp.layers.0.embed_norm.weight": "model-00033-of-00033.safetensors",
971
+ "model.mtp.layers.0.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
972
+ "model.mtp.layers.0.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
973
+ "model.mtp.layers.0.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
974
+ "model.mtp.layers.0.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
975
+ "model.mtp.layers.1.embed_norm.weight": "model-00033-of-00033.safetensors",
976
+ "model.mtp.layers.1.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
977
+ "model.mtp.layers.1.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
978
+ "model.mtp.layers.1.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
979
+ "model.mtp.layers.1.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
980
+ "model.mtp.layers.2.embed_norm.weight": "model-00033-of-00033.safetensors",
981
+ "model.mtp.layers.2.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
982
+ "model.mtp.layers.2.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
983
+ "model.mtp.layers.2.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
984
+ "model.mtp.layers.2.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
985
+ "model.mtp.layers.3.embed_norm.weight": "model-00033-of-00033.safetensors",
986
+ "model.mtp.layers.3.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
987
+ "model.mtp.layers.3.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
988
+ "model.mtp.layers.3.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
989
+ "model.mtp.layers.3.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
990
+ "model.mtp.layers.4.embed_norm.weight": "model-00033-of-00033.safetensors",
991
+ "model.mtp.layers.4.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
992
+ "model.mtp.layers.4.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
993
+ "model.mtp.layers.4.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
994
+ "model.mtp.layers.4.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
995
+ "model.mtp.layers.5.embed_norm.weight": "model-00033-of-00033.safetensors",
996
+ "model.mtp.layers.5.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
997
+ "model.mtp.layers.5.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
998
+ "model.mtp.layers.5.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
999
+ "model.mtp.layers.5.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
1000
+ "model.mtp.layers.6.embed_norm.weight": "model-00033-of-00033.safetensors",
1001
+ "model.mtp.layers.6.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
1002
+ "model.mtp.layers.6.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
1003
+ "model.mtp.layers.6.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
1004
+ "model.mtp.layers.6.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
1005
+ "model.mtp.layers.7.embed_norm.weight": "model-00033-of-00033.safetensors",
1006
+ "model.mtp.layers.7.transformer_block.attn.k_sconv.weight": "model-00033-of-00033.safetensors",
1007
+ "model.mtp.layers.7.transformer_block.attn.v_sconv.weight": "model-00033-of-00033.safetensors",
1008
+ "model.mtp.layers.7.transformer_block.attn_sconv.weight": "model-00033-of-00033.safetensors",
1009
+ "model.mtp.layers.7.transformer_block.mlp_sconv.weight": "model-00033-of-00033.safetensors",
1010
+ "model.llm.layers.10.mlp_norm.weight": "model-00026-of-00033.safetensors",
1011
+ "model.llm.layers.14.attn.wk_dv.weight": "model-00026-of-00033.safetensors",
1012
+ "model.llm.layers.17.mlp.gate.global_scale": "model-00026-of-00033.safetensors",
1013
+ "model.llm.layers.20.mlp.experts.w2_weight.weight_packed": "model-00026-of-00033.safetensors",
1014
+ "model.llm.layers.20.mlp.experts.w2_weight.weight_scale": "model-00026-of-00033.safetensors",
1015
+ "model.llm.layers.34.mlp.experts.w2_weight.weight_packed": "model-00026-of-00033.safetensors",
1016
+ "model.llm.layers.34.mlp.experts.w2_weight.weight_scale": "model-00026-of-00033.safetensors",
1017
+ "model.llm.layers.38.mlp.gate.bias": "model-00026-of-00033.safetensors",
1018
+ "model.llm.layers.40.mlp.experts.w13_weight.weight_packed": "model-00026-of-00033.safetensors",
1019
+ "model.llm.layers.40.mlp.experts.w13_weight.weight_scale": "model-00026-of-00033.safetensors",
1020
+ "model.llm.layers.4.attn.k_sconv.weight": "model-00026-of-00033.safetensors",
1021
+ "model.llm.layers.2.attn.wr_du.weight": "model-00028-of-00033.safetensors",
1022
+ "model.llm.layers.21.mlp.experts.w13_weight.weight_packed": "model-00028-of-00033.safetensors",
1023
+ "model.llm.layers.21.mlp.experts.w13_weight.weight_scale": "model-00028-of-00033.safetensors",
1024
+ "model.llm.layers.21.mlp.shared_experts.shared_w13_weight": "model-00028-of-00033.safetensors",
1025
+ "model.llm.layers.22.attn.rel_logits_proj.proj": "model-00028-of-00033.safetensors",
1026
+ "model.llm.layers.22.mlp.experts.w13_weight.weight_packed": "model-00028-of-00033.safetensors",
1027
+ "model.llm.layers.22.mlp.experts.w13_weight.weight_scale": "model-00028-of-00033.safetensors",
1028
+ "model.llm.layers.25.attn.k_norm.weight": "model-00028-of-00033.safetensors",
1029
+ "model.llm.layers.29.attn.rel_logits_proj.proj": "model-00028-of-00033.safetensors",
1030
+ "model.llm.layers.29.mlp.shared_experts.shared_w13_weight": "model-00028-of-00033.safetensors",
1031
+ "model.llm.layers.30.attn.wv_dv.weight": "model-00028-of-00033.safetensors",
1032
+ "model.llm.layers.34.attn.k_norm.weight": "model-00028-of-00033.safetensors",
1033
+ "model.llm.layers.36.attn.k_norm.weight": "model-00028-of-00033.safetensors",
1034
+ "model.llm.layers.37.attn.k_norm.weight": "model-00028-of-00033.safetensors",
1035
+ "model.llm.layers.38.mlp_norm.weight": "model-00028-of-00033.safetensors",
1036
+ "model.llm.layers.39.attn.wv_dv.weight": "model-00028-of-00033.safetensors",
1037
+ "model.visual.final_norm.weight": "model-00028-of-00033.safetensors",
1038
+ "model.llm.layers.11.attn.v_sconv.weight": "model-00028-of-00033.safetensors",
1039
+ "model.llm.layers.15.attn.v_sconv.weight": "model-00028-of-00033.safetensors",
1040
+ "model.llm.layers.2.attn_sconv.weight": "model-00028-of-00033.safetensors",
1041
+ "model.llm.layers.20.attn_sconv.weight": "model-00028-of-00033.safetensors",
1042
+ "model.llm.layers.3.mlp_sconv.weight": "model-00028-of-00033.safetensors",
1043
+ "model.llm.layers.38.attn_sconv.weight": "model-00028-of-00033.safetensors",
1044
+ "model.llm.layers.39.attn.k_sconv.weight": "model-00028-of-00033.safetensors",
1045
+ "model.llm.layers.39.mlp.gate.weight": "model-00028-of-00033.safetensors",
1046
+ "model.llm.layers.7.mlp.gate.weight": "model-00028-of-00033.safetensors",
1047
+ "model.llm.layers.8.mlp_sconv.weight": "model-00028-of-00033.safetensors",
1048
+ "model.llm.layers.11.attn.q_norm.weight": "model-00027-of-00033.safetensors",
1049
+ "model.llm.layers.15.attn.q_norm.weight": "model-00027-of-00033.safetensors",
1050
+ "model.llm.layers.17.mlp_norm.weight": "model-00027-of-00033.safetensors",
1051
+ "model.llm.layers.18.mlp.experts.w2_weight.weight_packed": "model-00027-of-00033.safetensors",
1052
+ "model.llm.layers.18.mlp.experts.w2_weight.weight_scale": "model-00027-of-00033.safetensors",
1053
+ "model.llm.layers.28.mlp.gate.bias": "model-00027-of-00033.safetensors",
1054
+ "model.llm.layers.29.mlp.gate.bias": "model-00027-of-00033.safetensors",
1055
+ "model.llm.layers.3.mlp.experts.w13_weight.weight_packed": "model-00027-of-00033.safetensors",
1056
+ "model.llm.layers.3.mlp.experts.w13_weight.weight_scale": "model-00027-of-00033.safetensors",
1057
+ "model.llm.layers.30.attn.rel_logits_proj.proj": "model-00027-of-00033.safetensors",
1058
+ "model.llm.layers.30.attn.wk_dv.weight": "model-00027-of-00033.safetensors",
1059
+ "model.llm.layers.31.mlp.shared_experts.shared_w2_weight": "model-00027-of-00033.safetensors",
1060
+ "model.llm.layers.34.attn.wq_du.weight": "model-00027-of-00033.safetensors",
1061
+ "model.llm.layers.37.mlp.experts.w2_weight.weight_packed": "model-00027-of-00033.safetensors",
1062
+ "model.llm.layers.37.mlp.experts.w2_weight.weight_scale": "model-00027-of-00033.safetensors",
1063
+ "model.llm.layers.40.mlp.shared_experts.shared_w2_weight": "model-00027-of-00033.safetensors",
1064
+ "model.llm.layers.41.attn.wq_du.weight": "model-00027-of-00033.safetensors",
1065
+ "model.llm.layers.5.attn.wq_du.weight": "model-00027-of-00033.safetensors",
1066
+ "model.llm.layers.7.attn.wq_du.weight": "model-00027-of-00033.safetensors",
1067
+ "model.llm.layers.7.attn_norm.weight": "model-00027-of-00033.safetensors",
1068
+ "model.llm.layers.13.mlp.gate.weight": "model-00027-of-00033.safetensors",
1069
+ "model.llm.layers.25.attn_sconv.weight": "model-00027-of-00033.safetensors",
1070
+ "model.llm.layers.41.mlp_sconv.weight": "model-00027-of-00033.safetensors",
1071
+ "model.llm.layers.8.attn.v_sconv.weight": "model-00027-of-00033.safetensors",
1072
+ "model.llm.layers.11.attn.wq_du.weight": "model-00029-of-00033.safetensors",
1073
+ "model.llm.layers.15.mlp.gate.global_scale": "model-00029-of-00033.safetensors",
1074
+ "model.llm.layers.29.mlp.experts.w13_weight.weight_packed": "model-00029-of-00033.safetensors",
1075
+ "model.llm.layers.29.mlp.experts.w13_weight.weight_scale": "model-00029-of-00033.safetensors",
1076
+ "model.llm.layers.33.mlp.shared_experts.shared_w2_weight": "model-00029-of-00033.safetensors",
1077
+ "model.llm.layers.36.mlp.experts.w13_weight.weight_packed": "model-00029-of-00033.safetensors",
1078
+ "model.llm.layers.36.mlp.experts.w13_weight.weight_scale": "model-00029-of-00033.safetensors",
1079
+ "model.llm.layers.37.attn.wv_dv.weight": "model-00029-of-00033.safetensors",
1080
+ "model.llm.layers.4.mlp_norm.weight": "model-00029-of-00033.safetensors",
1081
+ "model.llm.layers.41.attn.wk_dv.weight": "model-00029-of-00033.safetensors",
1082
+ "model.llm.layers.5.attn.k_norm.weight": "model-00029-of-00033.safetensors",
1083
+ "model.llm.layers.7.mlp.shared_experts.shared_w13_weight": "model-00029-of-00033.safetensors",
1084
+ "model.llm.layers.7.mlp_norm.weight": "model-00029-of-00033.safetensors",
1085
+ "model.llm.layers.8.attn_norm.weight": "model-00029-of-00033.safetensors",
1086
+ "model.llm.layers.13.mlp_sconv.weight": "model-00029-of-00033.safetensors",
1087
+ "model.llm.layers.17.attn.k_sconv.weight": "model-00029-of-00033.safetensors",
1088
+ "model.llm.layers.22.attn.v_sconv.weight": "model-00029-of-00033.safetensors",
1089
+ "model.llm.layers.15.attn_norm.weight": "model-00032-of-00033.safetensors",
1090
+ "model.llm.layers.16.attn_norm.weight": "model-00032-of-00033.safetensors",
1091
+ "model.llm.layers.22.attn.q_norm.weight": "model-00032-of-00033.safetensors",
1092
+ "model.llm.layers.22.mlp.experts.w2_weight.weight_packed": "model-00032-of-00033.safetensors",
1093
+ "model.llm.layers.22.mlp.experts.w2_weight.weight_scale": "model-00032-of-00033.safetensors",
1094
+ "model.llm.layers.22.mlp.gate.bias": "model-00032-of-00033.safetensors",
1095
+ "model.llm.layers.26.attn.wq_du.weight": "model-00032-of-00033.safetensors",
1096
+ "model.llm.layers.26.mlp.gate.global_scale": "model-00032-of-00033.safetensors",
1097
+ "model.llm.layers.27.mlp.experts.w13_weight.weight_packed": "model-00032-of-00033.safetensors",
1098
+ "model.llm.layers.27.mlp.experts.w13_weight.weight_scale": "model-00032-of-00033.safetensors",
1099
+ "model.llm.layers.29.mlp.gate.global_scale": "model-00032-of-00033.safetensors",
1100
+ "model.llm.layers.30.mlp.shared_experts.shared_w2_weight": "model-00032-of-00033.safetensors",
1101
+ "model.llm.layers.5.attn.wo_ud.weight": "model-00032-of-00033.safetensors",
1102
+ "model.llm.layers.16.attn.v_sconv.weight": "model-00032-of-00033.safetensors",
1103
+ "model.llm.layers.17.attn_sconv.weight": "model-00032-of-00033.safetensors",
1104
+ "model.llm.layers.22.attn_sconv.weight": "model-00032-of-00033.safetensors",
1105
+ "model.llm.layers.27.mlp_sconv.weight": "model-00032-of-00033.safetensors",
1106
+ "model.llm.layers.30.attn.v_sconv.weight": "model-00032-of-00033.safetensors",
1107
+ "model.llm.layers.31.attn.k_sconv.weight": "model-00032-of-00033.safetensors",
1108
+ "model.llm.layers.32.mlp.gate.weight": "model-00032-of-00033.safetensors",
1109
+ "model.llm.layers.35.attn.k_sconv.weight": "model-00032-of-00033.safetensors",
1110
+ "model.llm.layers.37.attn.k_sconv.weight": "model-00032-of-00033.safetensors",
1111
+ "model.llm.layers.37.attn_sconv.weight": "model-00032-of-00033.safetensors",
1112
+ "model.llm.layers.18.mlp_norm.weight": "model-00030-of-00033.safetensors",
1113
+ "model.llm.layers.2.attn.wq_du.weight": "model-00030-of-00033.safetensors",
1114
+ "model.llm.layers.2.mlp.experts.w13_weight.weight_packed": "model-00030-of-00033.safetensors",
1115
+ "model.llm.layers.2.mlp.experts.w13_weight.weight_scale": "model-00030-of-00033.safetensors",
1116
+ "model.llm.layers.2.mlp.gate.bias": "model-00030-of-00033.safetensors",
1117
+ "model.llm.layers.2.mlp.shared_experts.shared_w13_weight": "model-00030-of-00033.safetensors",
1118
+ "model.llm.layers.21.attn_norm.weight": "model-00030-of-00033.safetensors",
1119
+ "model.llm.layers.22.mlp.shared_experts.shared_w2_weight": "model-00030-of-00033.safetensors",
1120
+ "model.llm.layers.28.mlp.experts.w13_weight.weight_packed": "model-00030-of-00033.safetensors",
1121
+ "model.llm.layers.28.mlp.experts.w13_weight.weight_scale": "model-00030-of-00033.safetensors",
1122
+ "model.llm.layers.32.attn.q_norm.weight": "model-00030-of-00033.safetensors",
1123
+ "model.llm.layers.7.mlp.gate.global_scale": "model-00030-of-00033.safetensors",
1124
+ "model.llm.layers.11.attn_sconv.weight": "model-00030-of-00033.safetensors",
1125
+ "model.llm.layers.13.attn.k_sconv.weight": "model-00030-of-00033.safetensors",
1126
+ "model.llm.layers.26.attn.v_sconv.weight": "model-00030-of-00033.safetensors",
1127
+ "model.llm.layers.34.mlp.gate.weight": "model-00030-of-00033.safetensors",
1128
+ "model.llm.layers.11.mlp.shared_experts.shared_w2_weight": "model-00031-of-00033.safetensors",
1129
+ "model.llm.layers.13.mlp.gate.global_scale": "model-00031-of-00033.safetensors",
1130
+ "model.llm.layers.6.mlp.experts.w13_weight.weight_packed": "model-00031-of-00033.safetensors",
1131
+ "model.llm.layers.6.mlp.experts.w13_weight.weight_scale": "model-00031-of-00033.safetensors",
1132
+ "model.llm.layers.8.mlp.experts.w13_weight.weight_packed": "model-00031-of-00033.safetensors",
1133
+ "model.llm.layers.8.mlp.experts.w13_weight.weight_scale": "model-00031-of-00033.safetensors"
1134
+ }
1135
+ }
processor_config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "audio_token": "<|unused_200053|>",
3
+ "audio_bos_token": "<|content_audio_input|>",
4
+ "dmel_max_value": 2.0,
5
+ "dmel_min_value": -7.0,
6
+ "feature_extractor": {
7
+ "audio_token_duration_s": 0.05,
8
+ "feature_extractor_type": "InklingFeatureExtractor",
9
+ "feature_size": 80,
10
+ "hop_length": 800,
11
+ "n_fft": 1600,
12
+ "padding_side": "right",
13
+ "padding_value": 0.0,
14
+ "return_attention_mask": true,
15
+ "sampling_rate": 16000,
16
+ "window_size": 1600,
17
+ "window_size_multiplier": 2.0
18
+ },
19
+ "image_processor": {
20
+ "do_convert_rgb": true,
21
+ "do_normalize": true,
22
+ "do_rescale": true,
23
+ "do_resize": true,
24
+ "image_mean": [
25
+ 0.48145466,
26
+ 0.4578275,
27
+ 0.40821073
28
+ ],
29
+ "image_processor_type": "InklingImageProcessor",
30
+ "image_std": [
31
+ 0.26862954,
32
+ 0.26130258,
33
+ 0.27577711
34
+ ],
35
+ "resample": 3,
36
+ "rescale_factor": 0.00392156862745098,
37
+ "size": {
38
+ "height": 40,
39
+ "width": 40
40
+ }
41
+ },
42
+ "image_token": "<|unused_200054|>",
43
+ "image_bos_token": "<|content_image|>",
44
+ "num_dmel_bins": 16,
45
+ "processor_class": "InklingProcessor"
46
+ }
quantization_config.json ADDED
@@ -0,0 +1,1963 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "quant_method": "auto-round",
3
+ "packing_format": "auto_round:llm_compressor",
4
+ "bits": 4,
5
+ "group_size": 32,
6
+ "sym": true,
7
+ "data_type": "mx_fp",
8
+ "iters": 0,
9
+ "model_free": true,
10
+ "autoround_version": "0.15.0",
11
+ "enable_quanted_input": false,
12
+ "block_name_to_quantize": "model.llm.layers",
13
+ "act_bits": 16,
14
+ "act_data_type": "fp",
15
+ "act_dynamic": true,
16
+ "act_group_size": 128,
17
+ "act_sym": true,
18
+ "extra_config": {
19
+ "mlp.experts": {
20
+ "act_bits": 4,
21
+ "act_group_size": 32,
22
+ "act_data_type": "mx_fp"
23
+ },
24
+ "model.audio.encoder": {
25
+ "bits": 16,
26
+ "data_type": "float",
27
+ "act_bits": 16,
28
+ "act_data_type": "float"
29
+ },
30
+ "model.llm.layers.13.attn.wr_du": {
31
+ "bits": 16,
32
+ "data_type": "float",
33
+ "act_bits": 16,
34
+ "act_data_type": "float"
35
+ },
36
+ "model.llm.layers.3.mlp.gate": {
37
+ "bits": 16,
38
+ "data_type": "float",
39
+ "act_bits": 16,
40
+ "act_data_type": "float"
41
+ },
42
+ "model.llm.layers.34.attn.wk_dv": {
43
+ "bits": 16,
44
+ "data_type": "float",
45
+ "act_bits": 16,
46
+ "act_data_type": "float"
47
+ },
48
+ "model.llm.layers.37.attn.wk_dv": {
49
+ "bits": 16,
50
+ "data_type": "float",
51
+ "act_bits": 16,
52
+ "act_data_type": "float"
53
+ },
54
+ "model.llm.layers.37.attn.wo_ud": {
55
+ "bits": 16,
56
+ "data_type": "float",
57
+ "act_bits": 16,
58
+ "act_data_type": "float"
59
+ },
60
+ "model.llm.layers.11.mlp.gate": {
61
+ "bits": 16,
62
+ "data_type": "float",
63
+ "act_bits": 16,
64
+ "act_data_type": "float"
65
+ },
66
+ "model.llm.layers.13.attn.wk_dv": {
67
+ "bits": 16,
68
+ "data_type": "float",
69
+ "act_bits": 16,
70
+ "act_data_type": "float"
71
+ },
72
+ "model.llm.layers.18.attn.wq_du": {
73
+ "bits": 16,
74
+ "data_type": "float",
75
+ "act_bits": 16,
76
+ "act_data_type": "float"
77
+ },
78
+ "model.llm.layers.19.attn.wo_ud": {
79
+ "bits": 16,
80
+ "data_type": "float",
81
+ "act_bits": 16,
82
+ "act_data_type": "float"
83
+ },
84
+ "model.llm.layers.2.attn.wv_dv": {
85
+ "bits": 16,
86
+ "data_type": "float",
87
+ "act_bits": 16,
88
+ "act_data_type": "float"
89
+ },
90
+ "model.llm.layers.2.mlp.gate": {
91
+ "bits": 16,
92
+ "data_type": "float",
93
+ "act_bits": 16,
94
+ "act_data_type": "float"
95
+ },
96
+ "model.llm.layers.20.attn.wk_dv": {
97
+ "bits": 16,
98
+ "data_type": "float",
99
+ "act_bits": 16,
100
+ "act_data_type": "float"
101
+ },
102
+ "model.llm.layers.21.attn.wk_dv": {
103
+ "bits": 16,
104
+ "data_type": "float",
105
+ "act_bits": 16,
106
+ "act_data_type": "float"
107
+ },
108
+ "model.llm.layers.22.attn.wq_du": {
109
+ "bits": 16,
110
+ "data_type": "float",
111
+ "act_bits": 16,
112
+ "act_data_type": "float"
113
+ },
114
+ "model.llm.layers.23.attn.wr_du": {
115
+ "bits": 16,
116
+ "data_type": "float",
117
+ "act_bits": 16,
118
+ "act_data_type": "float"
119
+ },
120
+ "model.llm.layers.25.attn.wo_ud": {
121
+ "bits": 16,
122
+ "data_type": "float",
123
+ "act_bits": 16,
124
+ "act_data_type": "float"
125
+ },
126
+ "model.llm.layers.27.attn.wk_dv": {
127
+ "bits": 16,
128
+ "data_type": "float",
129
+ "act_bits": 16,
130
+ "act_data_type": "float"
131
+ },
132
+ "model.llm.layers.29.attn.wv_dv": {
133
+ "bits": 16,
134
+ "data_type": "float",
135
+ "act_bits": 16,
136
+ "act_data_type": "float"
137
+ },
138
+ "model.llm.layers.3.attn.wo_ud": {
139
+ "bits": 16,
140
+ "data_type": "float",
141
+ "act_bits": 16,
142
+ "act_data_type": "float"
143
+ },
144
+ "model.llm.layers.3.attn.wr_du": {
145
+ "bits": 16,
146
+ "data_type": "float",
147
+ "act_bits": 16,
148
+ "act_data_type": "float"
149
+ },
150
+ "model.llm.layers.32.attn.wo_ud": {
151
+ "bits": 16,
152
+ "data_type": "float",
153
+ "act_bits": 16,
154
+ "act_data_type": "float"
155
+ },
156
+ "model.llm.layers.32.attn.wq_du": {
157
+ "bits": 16,
158
+ "data_type": "float",
159
+ "act_bits": 16,
160
+ "act_data_type": "float"
161
+ },
162
+ "model.llm.layers.37.attn.wr_du": {
163
+ "bits": 16,
164
+ "data_type": "float",
165
+ "act_bits": 16,
166
+ "act_data_type": "float"
167
+ },
168
+ "model.llm.layers.4.attn.wk_dv": {
169
+ "bits": 16,
170
+ "data_type": "float",
171
+ "act_bits": 16,
172
+ "act_data_type": "float"
173
+ },
174
+ "model.llm.layers.4.attn.wq_du": {
175
+ "bits": 16,
176
+ "data_type": "float",
177
+ "act_bits": 16,
178
+ "act_data_type": "float"
179
+ },
180
+ "model.llm.layers.40.attn.wr_du": {
181
+ "bits": 16,
182
+ "data_type": "float",
183
+ "act_bits": 16,
184
+ "act_data_type": "float"
185
+ },
186
+ "model.llm.layers.15.mlp.gate": {
187
+ "bits": 16,
188
+ "data_type": "float",
189
+ "act_bits": 16,
190
+ "act_data_type": "float"
191
+ },
192
+ "model.llm.layers.23.attn.wk_dv": {
193
+ "bits": 16,
194
+ "data_type": "float",
195
+ "act_bits": 16,
196
+ "act_data_type": "float"
197
+ },
198
+ "model.llm.layers.30.attn.wr_du": {
199
+ "bits": 16,
200
+ "data_type": "float",
201
+ "act_bits": 16,
202
+ "act_data_type": "float"
203
+ },
204
+ "model.llm.layers.34.attn.wv_dv": {
205
+ "bits": 16,
206
+ "data_type": "float",
207
+ "act_bits": 16,
208
+ "act_data_type": "float"
209
+ },
210
+ "model.llm.layers.36.attn.wr_du": {
211
+ "bits": 16,
212
+ "data_type": "float",
213
+ "act_bits": 16,
214
+ "act_data_type": "float"
215
+ },
216
+ "model.llm.layers.4.mlp.gate": {
217
+ "bits": 16,
218
+ "data_type": "float",
219
+ "act_bits": 16,
220
+ "act_data_type": "float"
221
+ },
222
+ "model.llm.layers.6.attn.wr_du": {
223
+ "bits": 16,
224
+ "data_type": "float",
225
+ "act_bits": 16,
226
+ "act_data_type": "float"
227
+ },
228
+ "model.llm.layers.8.attn.wr_du": {
229
+ "bits": 16,
230
+ "data_type": "float",
231
+ "act_bits": 16,
232
+ "act_data_type": "float"
233
+ },
234
+ "model.llm.layers.9.attn.wr_du": {
235
+ "bits": 16,
236
+ "data_type": "float",
237
+ "act_bits": 16,
238
+ "act_data_type": "float"
239
+ },
240
+ "model.llm.layers.18.mlp.gate": {
241
+ "bits": 16,
242
+ "data_type": "float",
243
+ "act_bits": 16,
244
+ "act_data_type": "float"
245
+ },
246
+ "model.llm.layers.27.mlp.gate": {
247
+ "bits": 16,
248
+ "data_type": "float",
249
+ "act_bits": 16,
250
+ "act_data_type": "float"
251
+ },
252
+ "model.llm.layers.30.attn.wq_du": {
253
+ "bits": 16,
254
+ "data_type": "float",
255
+ "act_bits": 16,
256
+ "act_data_type": "float"
257
+ },
258
+ "model.llm.layers.0.attn.wq_du": {
259
+ "bits": 16,
260
+ "data_type": "float",
261
+ "act_bits": 16,
262
+ "act_data_type": "float"
263
+ },
264
+ "model.llm.layers.0.mlp.w13_dn": {
265
+ "bits": 16,
266
+ "data_type": "float",
267
+ "act_bits": 16,
268
+ "act_data_type": "float"
269
+ },
270
+ "model.llm.layers.14.mlp.gate": {
271
+ "bits": 16,
272
+ "data_type": "float",
273
+ "act_bits": 16,
274
+ "act_data_type": "float"
275
+ },
276
+ "model.llm.layers.20.attn.wv_dv": {
277
+ "bits": 16,
278
+ "data_type": "float",
279
+ "act_bits": 16,
280
+ "act_data_type": "float"
281
+ },
282
+ "model.llm.layers.23.attn.wo_ud": {
283
+ "bits": 16,
284
+ "data_type": "float",
285
+ "act_bits": 16,
286
+ "act_data_type": "float"
287
+ },
288
+ "model.llm.layers.23.attn.wq_du": {
289
+ "bits": 16,
290
+ "data_type": "float",
291
+ "act_bits": 16,
292
+ "act_data_type": "float"
293
+ },
294
+ "model.llm.layers.23.attn.wv_dv": {
295
+ "bits": 16,
296
+ "data_type": "float",
297
+ "act_bits": 16,
298
+ "act_data_type": "float"
299
+ },
300
+ "model.llm.layers.24.attn.wq_du": {
301
+ "bits": 16,
302
+ "data_type": "float",
303
+ "act_bits": 16,
304
+ "act_data_type": "float"
305
+ },
306
+ "model.llm.layers.26.mlp.gate": {
307
+ "bits": 16,
308
+ "data_type": "float",
309
+ "act_bits": 16,
310
+ "act_data_type": "float"
311
+ },
312
+ "model.llm.layers.27.attn.wq_du": {
313
+ "bits": 16,
314
+ "data_type": "float",
315
+ "act_bits": 16,
316
+ "act_data_type": "float"
317
+ },
318
+ "model.llm.layers.35.attn.wk_dv": {
319
+ "bits": 16,
320
+ "data_type": "float",
321
+ "act_bits": 16,
322
+ "act_data_type": "float"
323
+ },
324
+ "model.llm.layers.36.attn.wo_ud": {
325
+ "bits": 16,
326
+ "data_type": "float",
327
+ "act_bits": 16,
328
+ "act_data_type": "float"
329
+ },
330
+ "model.llm.layers.37.mlp.gate": {
331
+ "bits": 16,
332
+ "data_type": "float",
333
+ "act_bits": 16,
334
+ "act_data_type": "float"
335
+ },
336
+ "model.llm.layers.41.mlp.gate": {
337
+ "bits": 16,
338
+ "data_type": "float",
339
+ "act_bits": 16,
340
+ "act_data_type": "float"
341
+ },
342
+ "model.llm.layers.5.mlp.gate": {
343
+ "bits": 16,
344
+ "data_type": "float",
345
+ "act_bits": 16,
346
+ "act_data_type": "float"
347
+ },
348
+ "model.llm.layers.7.attn.wr_du": {
349
+ "bits": 16,
350
+ "data_type": "float",
351
+ "act_bits": 16,
352
+ "act_data_type": "float"
353
+ },
354
+ "model.visual.layers.linear_2": {
355
+ "bits": 16,
356
+ "data_type": "float",
357
+ "act_bits": 16,
358
+ "act_data_type": "float"
359
+ },
360
+ "model.llm.layers.12.attn.wo_ud": {
361
+ "bits": 16,
362
+ "data_type": "float",
363
+ "act_bits": 16,
364
+ "act_data_type": "float"
365
+ },
366
+ "model.llm.layers.14.attn.wq_du": {
367
+ "bits": 16,
368
+ "data_type": "float",
369
+ "act_bits": 16,
370
+ "act_data_type": "float"
371
+ },
372
+ "model.llm.layers.28.attn.wq_du": {
373
+ "bits": 16,
374
+ "data_type": "float",
375
+ "act_bits": 16,
376
+ "act_data_type": "float"
377
+ },
378
+ "model.llm.layers.31.attn.wr_du": {
379
+ "bits": 16,
380
+ "data_type": "float",
381
+ "act_bits": 16,
382
+ "act_data_type": "float"
383
+ },
384
+ "model.llm.layers.35.attn.wr_du": {
385
+ "bits": 16,
386
+ "data_type": "float",
387
+ "act_bits": 16,
388
+ "act_data_type": "float"
389
+ },
390
+ "model.llm.layers.38.attn.wq_du": {
391
+ "bits": 16,
392
+ "data_type": "float",
393
+ "act_bits": 16,
394
+ "act_data_type": "float"
395
+ },
396
+ "model.llm.layers.8.attn.wo_ud": {
397
+ "bits": 16,
398
+ "data_type": "float",
399
+ "act_bits": 16,
400
+ "act_data_type": "float"
401
+ },
402
+ "model.llm.layers.9.attn.wk_dv": {
403
+ "bits": 16,
404
+ "data_type": "float",
405
+ "act_bits": 16,
406
+ "act_data_type": "float"
407
+ },
408
+ "model.llm.layers.11.attn.wk_dv": {
409
+ "bits": 16,
410
+ "data_type": "float",
411
+ "act_bits": 16,
412
+ "act_data_type": "float"
413
+ },
414
+ "model.llm.layers.12.attn.wq_du": {
415
+ "bits": 16,
416
+ "data_type": "float",
417
+ "act_bits": 16,
418
+ "act_data_type": "float"
419
+ },
420
+ "model.llm.layers.14.attn.wo_ud": {
421
+ "bits": 16,
422
+ "data_type": "float",
423
+ "act_bits": 16,
424
+ "act_data_type": "float"
425
+ },
426
+ "model.llm.layers.15.attn.wk_dv": {
427
+ "bits": 16,
428
+ "data_type": "float",
429
+ "act_bits": 16,
430
+ "act_data_type": "float"
431
+ },
432
+ "model.llm.layers.19.mlp.gate": {
433
+ "bits": 16,
434
+ "data_type": "float",
435
+ "act_bits": 16,
436
+ "act_data_type": "float"
437
+ },
438
+ "model.llm.layers.2.attn.wo_ud": {
439
+ "bits": 16,
440
+ "data_type": "float",
441
+ "act_bits": 16,
442
+ "act_data_type": "float"
443
+ },
444
+ "model.llm.layers.24.mlp.gate": {
445
+ "bits": 16,
446
+ "data_type": "float",
447
+ "act_bits": 16,
448
+ "act_data_type": "float"
449
+ },
450
+ "model.llm.layers.28.attn.wo_ud": {
451
+ "bits": 16,
452
+ "data_type": "float",
453
+ "act_bits": 16,
454
+ "act_data_type": "float"
455
+ },
456
+ "model.llm.layers.3.attn.wk_dv": {
457
+ "bits": 16,
458
+ "data_type": "float",
459
+ "act_bits": 16,
460
+ "act_data_type": "float"
461
+ },
462
+ "model.llm.layers.33.attn.wr_du": {
463
+ "bits": 16,
464
+ "data_type": "float",
465
+ "act_bits": 16,
466
+ "act_data_type": "float"
467
+ },
468
+ "model.llm.layers.34.attn.wr_du": {
469
+ "bits": 16,
470
+ "data_type": "float",
471
+ "act_bits": 16,
472
+ "act_data_type": "float"
473
+ },
474
+ "model.llm.layers.35.mlp.gate": {
475
+ "bits": 16,
476
+ "data_type": "float",
477
+ "act_bits": 16,
478
+ "act_data_type": "float"
479
+ },
480
+ "model.llm.layers.6.attn.wo_ud": {
481
+ "bits": 16,
482
+ "data_type": "float",
483
+ "act_bits": 16,
484
+ "act_data_type": "float"
485
+ },
486
+ "model.llm.layers.0.attn.wk_dv": {
487
+ "bits": 16,
488
+ "data_type": "float",
489
+ "act_bits": 16,
490
+ "act_data_type": "float"
491
+ },
492
+ "model.llm.layers.10.attn.wr_du": {
493
+ "bits": 16,
494
+ "data_type": "float",
495
+ "act_bits": 16,
496
+ "act_data_type": "float"
497
+ },
498
+ "model.llm.layers.13.attn.wv_dv": {
499
+ "bits": 16,
500
+ "data_type": "float",
501
+ "act_bits": 16,
502
+ "act_data_type": "float"
503
+ },
504
+ "model.llm.layers.15.attn.wr_du": {
505
+ "bits": 16,
506
+ "data_type": "float",
507
+ "act_bits": 16,
508
+ "act_data_type": "float"
509
+ },
510
+ "model.llm.layers.17.attn.wo_ud": {
511
+ "bits": 16,
512
+ "data_type": "float",
513
+ "act_bits": 16,
514
+ "act_data_type": "float"
515
+ },
516
+ "model.llm.layers.20.attn.wr_du": {
517
+ "bits": 16,
518
+ "data_type": "float",
519
+ "act_bits": 16,
520
+ "act_data_type": "float"
521
+ },
522
+ "model.llm.layers.36.attn.wv_dv": {
523
+ "bits": 16,
524
+ "data_type": "float",
525
+ "act_bits": 16,
526
+ "act_data_type": "float"
527
+ },
528
+ "model.llm.layers.38.attn.wk_dv": {
529
+ "bits": 16,
530
+ "data_type": "float",
531
+ "act_bits": 16,
532
+ "act_data_type": "float"
533
+ },
534
+ "model.llm.layers.41.attn.wo_ud": {
535
+ "bits": 16,
536
+ "data_type": "float",
537
+ "act_bits": 16,
538
+ "act_data_type": "float"
539
+ },
540
+ "model.llm.layers.9.attn.wo_ud": {
541
+ "bits": 16,
542
+ "data_type": "float",
543
+ "act_bits": 16,
544
+ "act_data_type": "float"
545
+ },
546
+ "model.llm.layers.9.attn.wv_dv": {
547
+ "bits": 16,
548
+ "data_type": "float",
549
+ "act_bits": 16,
550
+ "act_data_type": "float"
551
+ },
552
+ "model.llm.layers.0.attn.wo_ud": {
553
+ "bits": 16,
554
+ "data_type": "float",
555
+ "act_bits": 16,
556
+ "act_data_type": "float"
557
+ },
558
+ "model.llm.layers.10.attn.wq_du": {
559
+ "bits": 16,
560
+ "data_type": "float",
561
+ "act_bits": 16,
562
+ "act_data_type": "float"
563
+ },
564
+ "model.llm.layers.10.mlp.gate": {
565
+ "bits": 16,
566
+ "data_type": "float",
567
+ "act_bits": 16,
568
+ "act_data_type": "float"
569
+ },
570
+ "model.llm.layers.19.attn.wv_dv": {
571
+ "bits": 16,
572
+ "data_type": "float",
573
+ "act_bits": 16,
574
+ "act_data_type": "float"
575
+ },
576
+ "model.llm.layers.21.attn.wo_ud": {
577
+ "bits": 16,
578
+ "data_type": "float",
579
+ "act_bits": 16,
580
+ "act_data_type": "float"
581
+ },
582
+ "model.llm.layers.28.attn.wr_du": {
583
+ "bits": 16,
584
+ "data_type": "float",
585
+ "act_bits": 16,
586
+ "act_data_type": "float"
587
+ },
588
+ "model.llm.layers.30.attn.wo_ud": {
589
+ "bits": 16,
590
+ "data_type": "float",
591
+ "act_bits": 16,
592
+ "act_data_type": "float"
593
+ },
594
+ "model.llm.layers.35.attn.wq_du": {
595
+ "bits": 16,
596
+ "data_type": "float",
597
+ "act_bits": 16,
598
+ "act_data_type": "float"
599
+ },
600
+ "model.llm.layers.40.attn.wo_ud": {
601
+ "bits": 16,
602
+ "data_type": "float",
603
+ "act_bits": 16,
604
+ "act_data_type": "float"
605
+ },
606
+ "model.llm.layers.8.attn.wq_du": {
607
+ "bits": 16,
608
+ "data_type": "float",
609
+ "act_bits": 16,
610
+ "act_data_type": "float"
611
+ },
612
+ "model.llm.layers.0.attn.wr_du": {
613
+ "bits": 16,
614
+ "data_type": "float",
615
+ "act_bits": 16,
616
+ "act_data_type": "float"
617
+ },
618
+ "model.llm.layers.0.attn.wv_dv": {
619
+ "bits": 16,
620
+ "data_type": "float",
621
+ "act_bits": 16,
622
+ "act_data_type": "float"
623
+ },
624
+ "model.llm.layers.1.attn.wo_ud": {
625
+ "bits": 16,
626
+ "data_type": "float",
627
+ "act_bits": 16,
628
+ "act_data_type": "float"
629
+ },
630
+ "model.llm.layers.1.attn.wv_dv": {
631
+ "bits": 16,
632
+ "data_type": "float",
633
+ "act_bits": 16,
634
+ "act_data_type": "float"
635
+ },
636
+ "model.llm.layers.16.attn.wr_du": {
637
+ "bits": 16,
638
+ "data_type": "float",
639
+ "act_bits": 16,
640
+ "act_data_type": "float"
641
+ },
642
+ "model.llm.layers.21.attn.wq_du": {
643
+ "bits": 16,
644
+ "data_type": "float",
645
+ "act_bits": 16,
646
+ "act_data_type": "float"
647
+ },
648
+ "model.llm.layers.21.attn.wv_dv": {
649
+ "bits": 16,
650
+ "data_type": "float",
651
+ "act_bits": 16,
652
+ "act_data_type": "float"
653
+ },
654
+ "model.llm.layers.23.mlp.gate": {
655
+ "bits": 16,
656
+ "data_type": "float",
657
+ "act_bits": 16,
658
+ "act_data_type": "float"
659
+ },
660
+ "model.llm.layers.24.attn.wv_dv": {
661
+ "bits": 16,
662
+ "data_type": "float",
663
+ "act_bits": 16,
664
+ "act_data_type": "float"
665
+ },
666
+ "model.llm.layers.25.attn.wq_du": {
667
+ "bits": 16,
668
+ "data_type": "float",
669
+ "act_bits": 16,
670
+ "act_data_type": "float"
671
+ },
672
+ "model.llm.layers.26.attn.wk_dv": {
673
+ "bits": 16,
674
+ "data_type": "float",
675
+ "act_bits": 16,
676
+ "act_data_type": "float"
677
+ },
678
+ "model.llm.layers.27.attn.wo_ud": {
679
+ "bits": 16,
680
+ "data_type": "float",
681
+ "act_bits": 16,
682
+ "act_data_type": "float"
683
+ },
684
+ "model.llm.layers.29.attn.wq_du": {
685
+ "bits": 16,
686
+ "data_type": "float",
687
+ "act_bits": 16,
688
+ "act_data_type": "float"
689
+ },
690
+ "model.llm.layers.33.attn.wv_dv": {
691
+ "bits": 16,
692
+ "data_type": "float",
693
+ "act_bits": 16,
694
+ "act_data_type": "float"
695
+ },
696
+ "model.llm.layers.36.attn.wk_dv": {
697
+ "bits": 16,
698
+ "data_type": "float",
699
+ "act_bits": 16,
700
+ "act_data_type": "float"
701
+ },
702
+ "model.llm.layers.39.attn.wr_du": {
703
+ "bits": 16,
704
+ "data_type": "float",
705
+ "act_bits": 16,
706
+ "act_data_type": "float"
707
+ },
708
+ "model.llm.layers.13.attn.wq_du": {
709
+ "bits": 16,
710
+ "data_type": "float",
711
+ "act_bits": 16,
712
+ "act_data_type": "float"
713
+ },
714
+ "model.llm.layers.15.attn.wq_du": {
715
+ "bits": 16,
716
+ "data_type": "float",
717
+ "act_bits": 16,
718
+ "act_data_type": "float"
719
+ },
720
+ "model.llm.layers.16.attn.wq_du": {
721
+ "bits": 16,
722
+ "data_type": "float",
723
+ "act_bits": 16,
724
+ "act_data_type": "float"
725
+ },
726
+ "model.llm.layers.18.attn.wk_dv": {
727
+ "bits": 16,
728
+ "data_type": "float",
729
+ "act_bits": 16,
730
+ "act_data_type": "float"
731
+ },
732
+ "model.llm.layers.20.attn.wo_ud": {
733
+ "bits": 16,
734
+ "data_type": "float",
735
+ "act_bits": 16,
736
+ "act_data_type": "float"
737
+ },
738
+ "model.llm.layers.22.attn.wo_ud": {
739
+ "bits": 16,
740
+ "data_type": "float",
741
+ "act_bits": 16,
742
+ "act_data_type": "float"
743
+ },
744
+ "model.llm.layers.27.attn.wr_du": {
745
+ "bits": 16,
746
+ "data_type": "float",
747
+ "act_bits": 16,
748
+ "act_data_type": "float"
749
+ },
750
+ "model.llm.layers.29.mlp.gate": {
751
+ "bits": 16,
752
+ "data_type": "float",
753
+ "act_bits": 16,
754
+ "act_data_type": "float"
755
+ },
756
+ "model.llm.layers.16.attn.wv_dv": {
757
+ "bits": 16,
758
+ "data_type": "float",
759
+ "act_bits": 16,
760
+ "act_data_type": "float"
761
+ },
762
+ "model.llm.layers.17.attn.wq_du": {
763
+ "bits": 16,
764
+ "data_type": "float",
765
+ "act_bits": 16,
766
+ "act_data_type": "float"
767
+ },
768
+ "model.llm.layers.19.attn.wq_du": {
769
+ "bits": 16,
770
+ "data_type": "float",
771
+ "act_bits": 16,
772
+ "act_data_type": "float"
773
+ },
774
+ "model.llm.layers.27.attn.wv_dv": {
775
+ "bits": 16,
776
+ "data_type": "float",
777
+ "act_bits": 16,
778
+ "act_data_type": "float"
779
+ },
780
+ "model.llm.layers.29.attn.wo_ud": {
781
+ "bits": 16,
782
+ "data_type": "float",
783
+ "act_bits": 16,
784
+ "act_data_type": "float"
785
+ },
786
+ "model.llm.layers.38.attn.wv_dv": {
787
+ "bits": 16,
788
+ "data_type": "float",
789
+ "act_bits": 16,
790
+ "act_data_type": "float"
791
+ },
792
+ "model.llm.layers.39.attn.wk_dv": {
793
+ "bits": 16,
794
+ "data_type": "float",
795
+ "act_bits": 16,
796
+ "act_data_type": "float"
797
+ },
798
+ "model.llm.layers.4.attn.wo_ud": {
799
+ "bits": 16,
800
+ "data_type": "float",
801
+ "act_bits": 16,
802
+ "act_data_type": "float"
803
+ },
804
+ "model.llm.layers.14.attn.wv_dv": {
805
+ "bits": 16,
806
+ "data_type": "float",
807
+ "act_bits": 16,
808
+ "act_data_type": "float"
809
+ },
810
+ "model.llm.layers.16.attn.wk_dv": {
811
+ "bits": 16,
812
+ "data_type": "float",
813
+ "act_bits": 16,
814
+ "act_data_type": "float"
815
+ },
816
+ "model.llm.layers.18.attn.wo_ud": {
817
+ "bits": 16,
818
+ "data_type": "float",
819
+ "act_bits": 16,
820
+ "act_data_type": "float"
821
+ },
822
+ "model.llm.layers.22.attn.wv_dv": {
823
+ "bits": 16,
824
+ "data_type": "float",
825
+ "act_bits": 16,
826
+ "act_data_type": "float"
827
+ },
828
+ "model.llm.layers.28.mlp.gate": {
829
+ "bits": 16,
830
+ "data_type": "float",
831
+ "act_bits": 16,
832
+ "act_data_type": "float"
833
+ },
834
+ "model.llm.layers.31.mlp.gate": {
835
+ "bits": 16,
836
+ "data_type": "float",
837
+ "act_bits": 16,
838
+ "act_data_type": "float"
839
+ },
840
+ "model.llm.layers.40.attn.wk_dv": {
841
+ "bits": 16,
842
+ "data_type": "float",
843
+ "act_bits": 16,
844
+ "act_data_type": "float"
845
+ },
846
+ "model.llm.layers.5.attn.wv_dv": {
847
+ "bits": 16,
848
+ "data_type": "float",
849
+ "act_bits": 16,
850
+ "act_data_type": "float"
851
+ },
852
+ "model.llm.layers.7.attn.wo_ud": {
853
+ "bits": 16,
854
+ "data_type": "float",
855
+ "act_bits": 16,
856
+ "act_data_type": "float"
857
+ },
858
+ "model.llm.layers.8.mlp.gate": {
859
+ "bits": 16,
860
+ "data_type": "float",
861
+ "act_bits": 16,
862
+ "act_data_type": "float"
863
+ },
864
+ "model.visual.layers.linear_1": {
865
+ "bits": 16,
866
+ "data_type": "float",
867
+ "act_bits": 16,
868
+ "act_data_type": "float"
869
+ },
870
+ "model.visual.layers.linear_3": {
871
+ "bits": 16,
872
+ "data_type": "float",
873
+ "act_bits": 16,
874
+ "act_data_type": "float"
875
+ },
876
+ "model.llm.layers.0.mlp.w2_md": {
877
+ "bits": 16,
878
+ "data_type": "float",
879
+ "act_bits": 16,
880
+ "act_data_type": "float"
881
+ },
882
+ "model.llm.layers.18.attn.wv_dv": {
883
+ "bits": 16,
884
+ "data_type": "float",
885
+ "act_bits": 16,
886
+ "act_data_type": "float"
887
+ },
888
+ "model.llm.layers.20.attn.wq_du": {
889
+ "bits": 16,
890
+ "data_type": "float",
891
+ "act_bits": 16,
892
+ "act_data_type": "float"
893
+ },
894
+ "model.llm.layers.24.attn.wr_du": {
895
+ "bits": 16,
896
+ "data_type": "float",
897
+ "act_bits": 16,
898
+ "act_data_type": "float"
899
+ },
900
+ "model.llm.layers.25.attn.wk_dv": {
901
+ "bits": 16,
902
+ "data_type": "float",
903
+ "act_bits": 16,
904
+ "act_data_type": "float"
905
+ },
906
+ "model.llm.layers.3.attn.wq_du": {
907
+ "bits": 16,
908
+ "data_type": "float",
909
+ "act_bits": 16,
910
+ "act_data_type": "float"
911
+ },
912
+ "model.llm.layers.33.attn.wq_du": {
913
+ "bits": 16,
914
+ "data_type": "float",
915
+ "act_bits": 16,
916
+ "act_data_type": "float"
917
+ },
918
+ "model.llm.layers.33.mlp.gate": {
919
+ "bits": 16,
920
+ "data_type": "float",
921
+ "act_bits": 16,
922
+ "act_data_type": "float"
923
+ },
924
+ "model.llm.layers.39.attn.wo_ud": {
925
+ "bits": 16,
926
+ "data_type": "float",
927
+ "act_bits": 16,
928
+ "act_data_type": "float"
929
+ },
930
+ "model.llm.layers.12.attn.wk_dv": {
931
+ "bits": 16,
932
+ "data_type": "float",
933
+ "act_bits": 16,
934
+ "act_data_type": "float"
935
+ },
936
+ "model.llm.layers.13.attn.wo_ud": {
937
+ "bits": 16,
938
+ "data_type": "float",
939
+ "act_bits": 16,
940
+ "act_data_type": "float"
941
+ },
942
+ "model.llm.layers.16.attn.wo_ud": {
943
+ "bits": 16,
944
+ "data_type": "float",
945
+ "act_bits": 16,
946
+ "act_data_type": "float"
947
+ },
948
+ "model.llm.layers.21.mlp.gate": {
949
+ "bits": 16,
950
+ "data_type": "float",
951
+ "act_bits": 16,
952
+ "act_data_type": "float"
953
+ },
954
+ "model.llm.layers.28.attn.wk_dv": {
955
+ "bits": 16,
956
+ "data_type": "float",
957
+ "act_bits": 16,
958
+ "act_data_type": "float"
959
+ },
960
+ "model.llm.layers.29.attn.wr_du": {
961
+ "bits": 16,
962
+ "data_type": "float",
963
+ "act_bits": 16,
964
+ "act_data_type": "float"
965
+ },
966
+ "model.llm.layers.34.attn.wo_ud": {
967
+ "bits": 16,
968
+ "data_type": "float",
969
+ "act_bits": 16,
970
+ "act_data_type": "float"
971
+ },
972
+ "model.llm.layers.38.attn.wo_ud": {
973
+ "bits": 16,
974
+ "data_type": "float",
975
+ "act_bits": 16,
976
+ "act_data_type": "float"
977
+ },
978
+ "model.llm.layers.38.attn.wr_du": {
979
+ "bits": 16,
980
+ "data_type": "float",
981
+ "act_bits": 16,
982
+ "act_data_type": "float"
983
+ },
984
+ "model.llm.layers.4.attn.wr_du": {
985
+ "bits": 16,
986
+ "data_type": "float",
987
+ "act_bits": 16,
988
+ "act_data_type": "float"
989
+ },
990
+ "model.llm.layers.4.attn.wv_dv": {
991
+ "bits": 16,
992
+ "data_type": "float",
993
+ "act_bits": 16,
994
+ "act_data_type": "float"
995
+ },
996
+ "model.llm.layers.7.attn.wv_dv": {
997
+ "bits": 16,
998
+ "data_type": "float",
999
+ "act_bits": 16,
1000
+ "act_data_type": "float"
1001
+ },
1002
+ "model.llm.layers.1.attn.wr_du": {
1003
+ "bits": 16,
1004
+ "data_type": "float",
1005
+ "act_bits": 16,
1006
+ "act_data_type": "float"
1007
+ },
1008
+ "model.llm.layers.11.attn.wo_ud": {
1009
+ "bits": 16,
1010
+ "data_type": "float",
1011
+ "act_bits": 16,
1012
+ "act_data_type": "float"
1013
+ },
1014
+ "model.llm.layers.16.mlp.gate": {
1015
+ "bits": 16,
1016
+ "data_type": "float",
1017
+ "act_bits": 16,
1018
+ "act_data_type": "float"
1019
+ },
1020
+ "model.llm.layers.18.attn.wr_du": {
1021
+ "bits": 16,
1022
+ "data_type": "float",
1023
+ "act_bits": 16,
1024
+ "act_data_type": "float"
1025
+ },
1026
+ "model.llm.layers.26.attn.wo_ud": {
1027
+ "bits": 16,
1028
+ "data_type": "float",
1029
+ "act_bits": 16,
1030
+ "act_data_type": "float"
1031
+ },
1032
+ "model.llm.layers.26.attn.wv_dv": {
1033
+ "bits": 16,
1034
+ "data_type": "float",
1035
+ "act_bits": 16,
1036
+ "act_data_type": "float"
1037
+ },
1038
+ "model.llm.layers.31.attn.wq_du": {
1039
+ "bits": 16,
1040
+ "data_type": "float",
1041
+ "act_bits": 16,
1042
+ "act_data_type": "float"
1043
+ },
1044
+ "model.llm.layers.36.attn.wq_du": {
1045
+ "bits": 16,
1046
+ "data_type": "float",
1047
+ "act_bits": 16,
1048
+ "act_data_type": "float"
1049
+ },
1050
+ "model.llm.layers.37.attn.wq_du": {
1051
+ "bits": 16,
1052
+ "data_type": "float",
1053
+ "act_bits": 16,
1054
+ "act_data_type": "float"
1055
+ },
1056
+ "model.llm.layers.39.attn.wq_du": {
1057
+ "bits": 16,
1058
+ "data_type": "float",
1059
+ "act_bits": 16,
1060
+ "act_data_type": "float"
1061
+ },
1062
+ "model.llm.layers.40.attn.wq_du": {
1063
+ "bits": 16,
1064
+ "data_type": "float",
1065
+ "act_bits": 16,
1066
+ "act_data_type": "float"
1067
+ },
1068
+ "model.llm.layers.40.mlp.gate": {
1069
+ "bits": 16,
1070
+ "data_type": "float",
1071
+ "act_bits": 16,
1072
+ "act_data_type": "float"
1073
+ },
1074
+ "model.llm.layers.5.attn.wr_du": {
1075
+ "bits": 16,
1076
+ "data_type": "float",
1077
+ "act_bits": 16,
1078
+ "act_data_type": "float"
1079
+ },
1080
+ "model.llm.layers.8.attn.wv_dv": {
1081
+ "bits": 16,
1082
+ "data_type": "float",
1083
+ "act_bits": 16,
1084
+ "act_data_type": "float"
1085
+ },
1086
+ "model.llm.layers.9.mlp.gate": {
1087
+ "bits": 16,
1088
+ "data_type": "float",
1089
+ "act_bits": 16,
1090
+ "act_data_type": "float"
1091
+ },
1092
+ "model.llm.layers.1.attn.wk_dv": {
1093
+ "bits": 16,
1094
+ "data_type": "float",
1095
+ "act_bits": 16,
1096
+ "act_data_type": "float"
1097
+ },
1098
+ "model.llm.layers.1.mlp.w2_md": {
1099
+ "bits": 16,
1100
+ "data_type": "float",
1101
+ "act_bits": 16,
1102
+ "act_data_type": "float"
1103
+ },
1104
+ "model.llm.layers.10.attn.wk_dv": {
1105
+ "bits": 16,
1106
+ "data_type": "float",
1107
+ "act_bits": 16,
1108
+ "act_data_type": "float"
1109
+ },
1110
+ "model.llm.layers.17.attn.wr_du": {
1111
+ "bits": 16,
1112
+ "data_type": "float",
1113
+ "act_bits": 16,
1114
+ "act_data_type": "float"
1115
+ },
1116
+ "model.llm.layers.22.attn.wk_dv": {
1117
+ "bits": 16,
1118
+ "data_type": "float",
1119
+ "act_bits": 16,
1120
+ "act_data_type": "float"
1121
+ },
1122
+ "model.llm.layers.38.mlp.gate": {
1123
+ "bits": 16,
1124
+ "data_type": "float",
1125
+ "act_bits": 16,
1126
+ "act_data_type": "float"
1127
+ },
1128
+ "model.llm.layers.41.attn.wr_du": {
1129
+ "bits": 16,
1130
+ "data_type": "float",
1131
+ "act_bits": 16,
1132
+ "act_data_type": "float"
1133
+ },
1134
+ "model.llm.layers.41.attn.wv_dv": {
1135
+ "bits": 16,
1136
+ "data_type": "float",
1137
+ "act_bits": 16,
1138
+ "act_data_type": "float"
1139
+ },
1140
+ "model.llm.layers.8.attn.wk_dv": {
1141
+ "bits": 16,
1142
+ "data_type": "float",
1143
+ "act_bits": 16,
1144
+ "act_data_type": "float"
1145
+ },
1146
+ "model.visual.layers.linear_0": {
1147
+ "bits": 16,
1148
+ "data_type": "float",
1149
+ "act_bits": 16,
1150
+ "act_data_type": "float"
1151
+ },
1152
+ "model.llm.layers.31.attn.wk_dv": {
1153
+ "bits": 16,
1154
+ "data_type": "float",
1155
+ "act_bits": 16,
1156
+ "act_data_type": "float"
1157
+ },
1158
+ "model.llm.layers.31.attn.wv_dv": {
1159
+ "bits": 16,
1160
+ "data_type": "float",
1161
+ "act_bits": 16,
1162
+ "act_data_type": "float"
1163
+ },
1164
+ "model.llm.layers.32.attn.wk_dv": {
1165
+ "bits": 16,
1166
+ "data_type": "float",
1167
+ "act_bits": 16,
1168
+ "act_data_type": "float"
1169
+ },
1170
+ "model.llm.layers.1.attn.wq_du": {
1171
+ "bits": 16,
1172
+ "data_type": "float",
1173
+ "act_bits": 16,
1174
+ "act_data_type": "float"
1175
+ },
1176
+ "model.llm.layers.20.mlp.gate": {
1177
+ "bits": 16,
1178
+ "data_type": "float",
1179
+ "act_bits": 16,
1180
+ "act_data_type": "float"
1181
+ },
1182
+ "model.llm.layers.24.attn.wo_ud": {
1183
+ "bits": 16,
1184
+ "data_type": "float",
1185
+ "act_bits": 16,
1186
+ "act_data_type": "float"
1187
+ },
1188
+ "model.llm.layers.25.mlp.gate": {
1189
+ "bits": 16,
1190
+ "data_type": "float",
1191
+ "act_bits": 16,
1192
+ "act_data_type": "float"
1193
+ },
1194
+ "model.llm.layers.17.attn.wk_dv": {
1195
+ "bits": 16,
1196
+ "data_type": "float",
1197
+ "act_bits": 16,
1198
+ "act_data_type": "float"
1199
+ },
1200
+ "model.llm.layers.17.attn.wv_dv": {
1201
+ "bits": 16,
1202
+ "data_type": "float",
1203
+ "act_bits": 16,
1204
+ "act_data_type": "float"
1205
+ },
1206
+ "model.llm.layers.22.attn.wr_du": {
1207
+ "bits": 16,
1208
+ "data_type": "float",
1209
+ "act_bits": 16,
1210
+ "act_data_type": "float"
1211
+ },
1212
+ "model.llm.layers.1.mlp.w13_dn": {
1213
+ "bits": 16,
1214
+ "data_type": "float",
1215
+ "act_bits": 16,
1216
+ "act_data_type": "float"
1217
+ },
1218
+ "model.llm.layers.12.attn.wr_du": {
1219
+ "bits": 16,
1220
+ "data_type": "float",
1221
+ "act_bits": 16,
1222
+ "act_data_type": "float"
1223
+ },
1224
+ "model.llm.layers.25.attn.wr_du": {
1225
+ "bits": 16,
1226
+ "data_type": "float",
1227
+ "act_bits": 16,
1228
+ "act_data_type": "float"
1229
+ },
1230
+ "model.llm.layers.35.attn.wv_dv": {
1231
+ "bits": 16,
1232
+ "data_type": "float",
1233
+ "act_bits": 16,
1234
+ "act_data_type": "float"
1235
+ },
1236
+ "model.llm.layers.36.mlp.gate": {
1237
+ "bits": 16,
1238
+ "data_type": "float",
1239
+ "act_bits": 16,
1240
+ "act_data_type": "float"
1241
+ },
1242
+ "model.llm.layers.10.attn.wo_ud": {
1243
+ "bits": 16,
1244
+ "data_type": "float",
1245
+ "act_bits": 16,
1246
+ "act_data_type": "float"
1247
+ },
1248
+ "model.llm.layers.14.attn.wr_du": {
1249
+ "bits": 16,
1250
+ "data_type": "float",
1251
+ "act_bits": 16,
1252
+ "act_data_type": "float"
1253
+ },
1254
+ "model.llm.layers.15.attn.wv_dv": {
1255
+ "bits": 16,
1256
+ "data_type": "float",
1257
+ "act_bits": 16,
1258
+ "act_data_type": "float"
1259
+ },
1260
+ "model.llm.layers.2.attn.wk_dv": {
1261
+ "bits": 16,
1262
+ "data_type": "float",
1263
+ "act_bits": 16,
1264
+ "act_data_type": "float"
1265
+ },
1266
+ "model.llm.layers.24.attn.wk_dv": {
1267
+ "bits": 16,
1268
+ "data_type": "float",
1269
+ "act_bits": 16,
1270
+ "act_data_type": "float"
1271
+ },
1272
+ "model.llm.layers.25.attn.wv_dv": {
1273
+ "bits": 16,
1274
+ "data_type": "float",
1275
+ "act_bits": 16,
1276
+ "act_data_type": "float"
1277
+ },
1278
+ "model.llm.layers.26.attn.wr_du": {
1279
+ "bits": 16,
1280
+ "data_type": "float",
1281
+ "act_bits": 16,
1282
+ "act_data_type": "float"
1283
+ },
1284
+ "model.llm.layers.29.attn.wk_dv": {
1285
+ "bits": 16,
1286
+ "data_type": "float",
1287
+ "act_bits": 16,
1288
+ "act_data_type": "float"
1289
+ },
1290
+ "model.llm.layers.32.attn.wr_du": {
1291
+ "bits": 16,
1292
+ "data_type": "float",
1293
+ "act_bits": 16,
1294
+ "act_data_type": "float"
1295
+ },
1296
+ "model.llm.layers.32.attn.wv_dv": {
1297
+ "bits": 16,
1298
+ "data_type": "float",
1299
+ "act_bits": 16,
1300
+ "act_data_type": "float"
1301
+ },
1302
+ "model.llm.layers.35.attn.wo_ud": {
1303
+ "bits": 16,
1304
+ "data_type": "float",
1305
+ "act_bits": 16,
1306
+ "act_data_type": "float"
1307
+ },
1308
+ "model.llm.layers.5.attn.wk_dv": {
1309
+ "bits": 16,
1310
+ "data_type": "float",
1311
+ "act_bits": 16,
1312
+ "act_data_type": "float"
1313
+ },
1314
+ "model.llm.layers.6.attn.wk_dv": {
1315
+ "bits": 16,
1316
+ "data_type": "float",
1317
+ "act_bits": 16,
1318
+ "act_data_type": "float"
1319
+ },
1320
+ "model.llm.layers.6.attn.wq_du": {
1321
+ "bits": 16,
1322
+ "data_type": "float",
1323
+ "act_bits": 16,
1324
+ "act_data_type": "float"
1325
+ },
1326
+ "model.llm.layers.6.attn.wv_dv": {
1327
+ "bits": 16,
1328
+ "data_type": "float",
1329
+ "act_bits": 16,
1330
+ "act_data_type": "float"
1331
+ },
1332
+ "model.llm.layers.9.attn.wq_du": {
1333
+ "bits": 16,
1334
+ "data_type": "float",
1335
+ "act_bits": 16,
1336
+ "act_data_type": "float"
1337
+ },
1338
+ "model.llm.layers.10.attn.wv_dv": {
1339
+ "bits": 16,
1340
+ "data_type": "float",
1341
+ "act_bits": 16,
1342
+ "act_data_type": "float"
1343
+ },
1344
+ "model.llm.layers.11.attn.wr_du": {
1345
+ "bits": 16,
1346
+ "data_type": "float",
1347
+ "act_bits": 16,
1348
+ "act_data_type": "float"
1349
+ },
1350
+ "model.llm.layers.12.attn.wv_dv": {
1351
+ "bits": 16,
1352
+ "data_type": "float",
1353
+ "act_bits": 16,
1354
+ "act_data_type": "float"
1355
+ },
1356
+ "model.llm.layers.21.attn.wr_du": {
1357
+ "bits": 16,
1358
+ "data_type": "float",
1359
+ "act_bits": 16,
1360
+ "act_data_type": "float"
1361
+ },
1362
+ "model.llm.layers.22.mlp.gate": {
1363
+ "bits": 16,
1364
+ "data_type": "float",
1365
+ "act_bits": 16,
1366
+ "act_data_type": "float"
1367
+ },
1368
+ "model.llm.layers.3.attn.wv_dv": {
1369
+ "bits": 16,
1370
+ "data_type": "float",
1371
+ "act_bits": 16,
1372
+ "act_data_type": "float"
1373
+ },
1374
+ "model.llm.layers.6.mlp.gate": {
1375
+ "bits": 16,
1376
+ "data_type": "float",
1377
+ "act_bits": 16,
1378
+ "act_data_type": "float"
1379
+ },
1380
+ "model.llm.layers.7.attn.wk_dv": {
1381
+ "bits": 16,
1382
+ "data_type": "float",
1383
+ "act_bits": 16,
1384
+ "act_data_type": "float"
1385
+ },
1386
+ "model.llm.layers.11.attn.wv_dv": {
1387
+ "bits": 16,
1388
+ "data_type": "float",
1389
+ "act_bits": 16,
1390
+ "act_data_type": "float"
1391
+ },
1392
+ "model.llm.layers.12.mlp.gate": {
1393
+ "bits": 16,
1394
+ "data_type": "float",
1395
+ "act_bits": 16,
1396
+ "act_data_type": "float"
1397
+ },
1398
+ "model.llm.layers.17.mlp.gate": {
1399
+ "bits": 16,
1400
+ "data_type": "float",
1401
+ "act_bits": 16,
1402
+ "act_data_type": "float"
1403
+ },
1404
+ "model.llm.layers.19.attn.wk_dv": {
1405
+ "bits": 16,
1406
+ "data_type": "float",
1407
+ "act_bits": 16,
1408
+ "act_data_type": "float"
1409
+ },
1410
+ "model.llm.layers.19.attn.wr_du": {
1411
+ "bits": 16,
1412
+ "data_type": "float",
1413
+ "act_bits": 16,
1414
+ "act_data_type": "float"
1415
+ },
1416
+ "model.llm.layers.28.attn.wv_dv": {
1417
+ "bits": 16,
1418
+ "data_type": "float",
1419
+ "act_bits": 16,
1420
+ "act_data_type": "float"
1421
+ },
1422
+ "model.llm.layers.33.attn.wk_dv": {
1423
+ "bits": 16,
1424
+ "data_type": "float",
1425
+ "act_bits": 16,
1426
+ "act_data_type": "float"
1427
+ },
1428
+ "model.llm.layers.33.attn.wo_ud": {
1429
+ "bits": 16,
1430
+ "data_type": "float",
1431
+ "act_bits": 16,
1432
+ "act_data_type": "float"
1433
+ },
1434
+ "model.llm.layers.40.attn.wv_dv": {
1435
+ "bits": 16,
1436
+ "data_type": "float",
1437
+ "act_bits": 16,
1438
+ "act_data_type": "float"
1439
+ },
1440
+ "model.llm.layers.15.attn.wo_ud": {
1441
+ "bits": 16,
1442
+ "data_type": "float",
1443
+ "act_bits": 16,
1444
+ "act_data_type": "float"
1445
+ },
1446
+ "model.llm.layers.30.mlp.gate": {
1447
+ "bits": 16,
1448
+ "data_type": "float",
1449
+ "act_bits": 16,
1450
+ "act_data_type": "float"
1451
+ },
1452
+ "model.llm.layers.31.attn.wo_ud": {
1453
+ "bits": 16,
1454
+ "data_type": "float",
1455
+ "act_bits": 16,
1456
+ "act_data_type": "float"
1457
+ },
1458
+ "model.mtp.layers.0.input_proj": {
1459
+ "bits": 16,
1460
+ "data_type": "float",
1461
+ "act_bits": 16,
1462
+ "act_data_type": "float"
1463
+ },
1464
+ "model.mtp.layers.0.transformer_block.attn.wk_dv": {
1465
+ "bits": 16,
1466
+ "data_type": "float",
1467
+ "act_bits": 16,
1468
+ "act_data_type": "float"
1469
+ },
1470
+ "model.mtp.layers.0.transformer_block.attn.wo_ud": {
1471
+ "bits": 16,
1472
+ "data_type": "float",
1473
+ "act_bits": 16,
1474
+ "act_data_type": "float"
1475
+ },
1476
+ "model.mtp.layers.0.transformer_block.attn.wq_du": {
1477
+ "bits": 16,
1478
+ "data_type": "float",
1479
+ "act_bits": 16,
1480
+ "act_data_type": "float"
1481
+ },
1482
+ "model.mtp.layers.0.transformer_block.attn.wr_du": {
1483
+ "bits": 16,
1484
+ "data_type": "float",
1485
+ "act_bits": 16,
1486
+ "act_data_type": "float"
1487
+ },
1488
+ "model.mtp.layers.0.transformer_block.attn.wv_dv": {
1489
+ "bits": 16,
1490
+ "data_type": "float",
1491
+ "act_bits": 16,
1492
+ "act_data_type": "float"
1493
+ },
1494
+ "model.mtp.layers.0.transformer_block.mlp.w13_dn": {
1495
+ "bits": 16,
1496
+ "data_type": "float",
1497
+ "act_bits": 16,
1498
+ "act_data_type": "float"
1499
+ },
1500
+ "model.mtp.layers.0.transformer_block.mlp.w2_md": {
1501
+ "bits": 16,
1502
+ "data_type": "float",
1503
+ "act_bits": 16,
1504
+ "act_data_type": "float"
1505
+ },
1506
+ "model.mtp.layers.1.input_proj": {
1507
+ "bits": 16,
1508
+ "data_type": "float",
1509
+ "act_bits": 16,
1510
+ "act_data_type": "float"
1511
+ },
1512
+ "model.mtp.layers.1.transformer_block.attn.wk_dv": {
1513
+ "bits": 16,
1514
+ "data_type": "float",
1515
+ "act_bits": 16,
1516
+ "act_data_type": "float"
1517
+ },
1518
+ "model.mtp.layers.1.transformer_block.attn.wo_ud": {
1519
+ "bits": 16,
1520
+ "data_type": "float",
1521
+ "act_bits": 16,
1522
+ "act_data_type": "float"
1523
+ },
1524
+ "model.mtp.layers.1.transformer_block.attn.wq_du": {
1525
+ "bits": 16,
1526
+ "data_type": "float",
1527
+ "act_bits": 16,
1528
+ "act_data_type": "float"
1529
+ },
1530
+ "model.mtp.layers.1.transformer_block.attn.wr_du": {
1531
+ "bits": 16,
1532
+ "data_type": "float",
1533
+ "act_bits": 16,
1534
+ "act_data_type": "float"
1535
+ },
1536
+ "model.mtp.layers.1.transformer_block.attn.wv_dv": {
1537
+ "bits": 16,
1538
+ "data_type": "float",
1539
+ "act_bits": 16,
1540
+ "act_data_type": "float"
1541
+ },
1542
+ "model.mtp.layers.1.transformer_block.mlp.w13_dn": {
1543
+ "bits": 16,
1544
+ "data_type": "float",
1545
+ "act_bits": 16,
1546
+ "act_data_type": "float"
1547
+ },
1548
+ "model.mtp.layers.1.transformer_block.mlp.w2_md": {
1549
+ "bits": 16,
1550
+ "data_type": "float",
1551
+ "act_bits": 16,
1552
+ "act_data_type": "float"
1553
+ },
1554
+ "model.mtp.layers.2.input_proj": {
1555
+ "bits": 16,
1556
+ "data_type": "float",
1557
+ "act_bits": 16,
1558
+ "act_data_type": "float"
1559
+ },
1560
+ "model.mtp.layers.2.transformer_block.attn.wk_dv": {
1561
+ "bits": 16,
1562
+ "data_type": "float",
1563
+ "act_bits": 16,
1564
+ "act_data_type": "float"
1565
+ },
1566
+ "model.mtp.layers.2.transformer_block.attn.wo_ud": {
1567
+ "bits": 16,
1568
+ "data_type": "float",
1569
+ "act_bits": 16,
1570
+ "act_data_type": "float"
1571
+ },
1572
+ "model.mtp.layers.2.transformer_block.attn.wq_du": {
1573
+ "bits": 16,
1574
+ "data_type": "float",
1575
+ "act_bits": 16,
1576
+ "act_data_type": "float"
1577
+ },
1578
+ "model.mtp.layers.2.transformer_block.attn.wr_du": {
1579
+ "bits": 16,
1580
+ "data_type": "float",
1581
+ "act_bits": 16,
1582
+ "act_data_type": "float"
1583
+ },
1584
+ "model.mtp.layers.2.transformer_block.attn.wv_dv": {
1585
+ "bits": 16,
1586
+ "data_type": "float",
1587
+ "act_bits": 16,
1588
+ "act_data_type": "float"
1589
+ },
1590
+ "model.mtp.layers.2.transformer_block.mlp.w13_dn": {
1591
+ "bits": 16,
1592
+ "data_type": "float",
1593
+ "act_bits": 16,
1594
+ "act_data_type": "float"
1595
+ },
1596
+ "model.mtp.layers.2.transformer_block.mlp.w2_md": {
1597
+ "bits": 16,
1598
+ "data_type": "float",
1599
+ "act_bits": 16,
1600
+ "act_data_type": "float"
1601
+ },
1602
+ "model.mtp.layers.3.input_proj": {
1603
+ "bits": 16,
1604
+ "data_type": "float",
1605
+ "act_bits": 16,
1606
+ "act_data_type": "float"
1607
+ },
1608
+ "model.mtp.layers.3.transformer_block.attn.wk_dv": {
1609
+ "bits": 16,
1610
+ "data_type": "float",
1611
+ "act_bits": 16,
1612
+ "act_data_type": "float"
1613
+ },
1614
+ "model.mtp.layers.3.transformer_block.attn.wo_ud": {
1615
+ "bits": 16,
1616
+ "data_type": "float",
1617
+ "act_bits": 16,
1618
+ "act_data_type": "float"
1619
+ },
1620
+ "model.mtp.layers.3.transformer_block.attn.wq_du": {
1621
+ "bits": 16,
1622
+ "data_type": "float",
1623
+ "act_bits": 16,
1624
+ "act_data_type": "float"
1625
+ },
1626
+ "model.mtp.layers.3.transformer_block.attn.wr_du": {
1627
+ "bits": 16,
1628
+ "data_type": "float",
1629
+ "act_bits": 16,
1630
+ "act_data_type": "float"
1631
+ },
1632
+ "model.mtp.layers.3.transformer_block.attn.wv_dv": {
1633
+ "bits": 16,
1634
+ "data_type": "float",
1635
+ "act_bits": 16,
1636
+ "act_data_type": "float"
1637
+ },
1638
+ "model.mtp.layers.3.transformer_block.mlp.w13_dn": {
1639
+ "bits": 16,
1640
+ "data_type": "float",
1641
+ "act_bits": 16,
1642
+ "act_data_type": "float"
1643
+ },
1644
+ "model.mtp.layers.3.transformer_block.mlp.w2_md": {
1645
+ "bits": 16,
1646
+ "data_type": "float",
1647
+ "act_bits": 16,
1648
+ "act_data_type": "float"
1649
+ },
1650
+ "model.mtp.layers.4.input_proj": {
1651
+ "bits": 16,
1652
+ "data_type": "float",
1653
+ "act_bits": 16,
1654
+ "act_data_type": "float"
1655
+ },
1656
+ "model.mtp.layers.4.transformer_block.attn.wk_dv": {
1657
+ "bits": 16,
1658
+ "data_type": "float",
1659
+ "act_bits": 16,
1660
+ "act_data_type": "float"
1661
+ },
1662
+ "model.mtp.layers.4.transformer_block.attn.wo_ud": {
1663
+ "bits": 16,
1664
+ "data_type": "float",
1665
+ "act_bits": 16,
1666
+ "act_data_type": "float"
1667
+ },
1668
+ "model.mtp.layers.4.transformer_block.attn.wq_du": {
1669
+ "bits": 16,
1670
+ "data_type": "float",
1671
+ "act_bits": 16,
1672
+ "act_data_type": "float"
1673
+ },
1674
+ "model.mtp.layers.4.transformer_block.attn.wr_du": {
1675
+ "bits": 16,
1676
+ "data_type": "float",
1677
+ "act_bits": 16,
1678
+ "act_data_type": "float"
1679
+ },
1680
+ "model.mtp.layers.4.transformer_block.attn.wv_dv": {
1681
+ "bits": 16,
1682
+ "data_type": "float",
1683
+ "act_bits": 16,
1684
+ "act_data_type": "float"
1685
+ },
1686
+ "model.mtp.layers.4.transformer_block.mlp.w13_dn": {
1687
+ "bits": 16,
1688
+ "data_type": "float",
1689
+ "act_bits": 16,
1690
+ "act_data_type": "float"
1691
+ },
1692
+ "model.mtp.layers.4.transformer_block.mlp.w2_md": {
1693
+ "bits": 16,
1694
+ "data_type": "float",
1695
+ "act_bits": 16,
1696
+ "act_data_type": "float"
1697
+ },
1698
+ "model.mtp.layers.5.input_proj": {
1699
+ "bits": 16,
1700
+ "data_type": "float",
1701
+ "act_bits": 16,
1702
+ "act_data_type": "float"
1703
+ },
1704
+ "model.mtp.layers.5.transformer_block.attn.wk_dv": {
1705
+ "bits": 16,
1706
+ "data_type": "float",
1707
+ "act_bits": 16,
1708
+ "act_data_type": "float"
1709
+ },
1710
+ "model.mtp.layers.5.transformer_block.attn.wo_ud": {
1711
+ "bits": 16,
1712
+ "data_type": "float",
1713
+ "act_bits": 16,
1714
+ "act_data_type": "float"
1715
+ },
1716
+ "model.mtp.layers.5.transformer_block.attn.wq_du": {
1717
+ "bits": 16,
1718
+ "data_type": "float",
1719
+ "act_bits": 16,
1720
+ "act_data_type": "float"
1721
+ },
1722
+ "model.mtp.layers.5.transformer_block.attn.wr_du": {
1723
+ "bits": 16,
1724
+ "data_type": "float",
1725
+ "act_bits": 16,
1726
+ "act_data_type": "float"
1727
+ },
1728
+ "model.mtp.layers.5.transformer_block.attn.wv_dv": {
1729
+ "bits": 16,
1730
+ "data_type": "float",
1731
+ "act_bits": 16,
1732
+ "act_data_type": "float"
1733
+ },
1734
+ "model.mtp.layers.5.transformer_block.mlp.w13_dn": {
1735
+ "bits": 16,
1736
+ "data_type": "float",
1737
+ "act_bits": 16,
1738
+ "act_data_type": "float"
1739
+ },
1740
+ "model.mtp.layers.5.transformer_block.mlp.w2_md": {
1741
+ "bits": 16,
1742
+ "data_type": "float",
1743
+ "act_bits": 16,
1744
+ "act_data_type": "float"
1745
+ },
1746
+ "model.mtp.layers.6.input_proj": {
1747
+ "bits": 16,
1748
+ "data_type": "float",
1749
+ "act_bits": 16,
1750
+ "act_data_type": "float"
1751
+ },
1752
+ "model.mtp.layers.6.transformer_block.attn.wk_dv": {
1753
+ "bits": 16,
1754
+ "data_type": "float",
1755
+ "act_bits": 16,
1756
+ "act_data_type": "float"
1757
+ },
1758
+ "model.mtp.layers.6.transformer_block.attn.wo_ud": {
1759
+ "bits": 16,
1760
+ "data_type": "float",
1761
+ "act_bits": 16,
1762
+ "act_data_type": "float"
1763
+ },
1764
+ "model.mtp.layers.6.transformer_block.attn.wq_du": {
1765
+ "bits": 16,
1766
+ "data_type": "float",
1767
+ "act_bits": 16,
1768
+ "act_data_type": "float"
1769
+ },
1770
+ "model.mtp.layers.6.transformer_block.attn.wr_du": {
1771
+ "bits": 16,
1772
+ "data_type": "float",
1773
+ "act_bits": 16,
1774
+ "act_data_type": "float"
1775
+ },
1776
+ "model.mtp.layers.6.transformer_block.attn.wv_dv": {
1777
+ "bits": 16,
1778
+ "data_type": "float",
1779
+ "act_bits": 16,
1780
+ "act_data_type": "float"
1781
+ },
1782
+ "model.mtp.layers.6.transformer_block.mlp.w13_dn": {
1783
+ "bits": 16,
1784
+ "data_type": "float",
1785
+ "act_bits": 16,
1786
+ "act_data_type": "float"
1787
+ },
1788
+ "model.mtp.layers.6.transformer_block.mlp.w2_md": {
1789
+ "bits": 16,
1790
+ "data_type": "float",
1791
+ "act_bits": 16,
1792
+ "act_data_type": "float"
1793
+ },
1794
+ "model.mtp.layers.7.input_proj": {
1795
+ "bits": 16,
1796
+ "data_type": "float",
1797
+ "act_bits": 16,
1798
+ "act_data_type": "float"
1799
+ },
1800
+ "model.mtp.layers.7.transformer_block.attn.wk_dv": {
1801
+ "bits": 16,
1802
+ "data_type": "float",
1803
+ "act_bits": 16,
1804
+ "act_data_type": "float"
1805
+ },
1806
+ "model.mtp.layers.7.transformer_block.attn.wo_ud": {
1807
+ "bits": 16,
1808
+ "data_type": "float",
1809
+ "act_bits": 16,
1810
+ "act_data_type": "float"
1811
+ },
1812
+ "model.mtp.layers.7.transformer_block.attn.wq_du": {
1813
+ "bits": 16,
1814
+ "data_type": "float",
1815
+ "act_bits": 16,
1816
+ "act_data_type": "float"
1817
+ },
1818
+ "model.mtp.layers.7.transformer_block.attn.wr_du": {
1819
+ "bits": 16,
1820
+ "data_type": "float",
1821
+ "act_bits": 16,
1822
+ "act_data_type": "float"
1823
+ },
1824
+ "model.mtp.layers.7.transformer_block.attn.wv_dv": {
1825
+ "bits": 16,
1826
+ "data_type": "float",
1827
+ "act_bits": 16,
1828
+ "act_data_type": "float"
1829
+ },
1830
+ "model.mtp.layers.7.transformer_block.mlp.w13_dn": {
1831
+ "bits": 16,
1832
+ "data_type": "float",
1833
+ "act_bits": 16,
1834
+ "act_data_type": "float"
1835
+ },
1836
+ "model.mtp.layers.7.transformer_block.mlp.w2_md": {
1837
+ "bits": 16,
1838
+ "data_type": "float",
1839
+ "act_bits": 16,
1840
+ "act_data_type": "float"
1841
+ },
1842
+ "model.llm.layers.14.attn.wk_dv": {
1843
+ "bits": 16,
1844
+ "data_type": "float",
1845
+ "act_bits": 16,
1846
+ "act_data_type": "float"
1847
+ },
1848
+ "model.llm.layers.2.attn.wr_du": {
1849
+ "bits": 16,
1850
+ "data_type": "float",
1851
+ "act_bits": 16,
1852
+ "act_data_type": "float"
1853
+ },
1854
+ "model.llm.layers.30.attn.wv_dv": {
1855
+ "bits": 16,
1856
+ "data_type": "float",
1857
+ "act_bits": 16,
1858
+ "act_data_type": "float"
1859
+ },
1860
+ "model.llm.layers.39.attn.wv_dv": {
1861
+ "bits": 16,
1862
+ "data_type": "float",
1863
+ "act_bits": 16,
1864
+ "act_data_type": "float"
1865
+ },
1866
+ "model.llm.layers.39.mlp.gate": {
1867
+ "bits": 16,
1868
+ "data_type": "float",
1869
+ "act_bits": 16,
1870
+ "act_data_type": "float"
1871
+ },
1872
+ "model.llm.layers.7.mlp.gate": {
1873
+ "bits": 16,
1874
+ "data_type": "float",
1875
+ "act_bits": 16,
1876
+ "act_data_type": "float"
1877
+ },
1878
+ "model.llm.layers.13.mlp.gate": {
1879
+ "bits": 16,
1880
+ "data_type": "float",
1881
+ "act_bits": 16,
1882
+ "act_data_type": "float"
1883
+ },
1884
+ "model.llm.layers.30.attn.wk_dv": {
1885
+ "bits": 16,
1886
+ "data_type": "float",
1887
+ "act_bits": 16,
1888
+ "act_data_type": "float"
1889
+ },
1890
+ "model.llm.layers.34.attn.wq_du": {
1891
+ "bits": 16,
1892
+ "data_type": "float",
1893
+ "act_bits": 16,
1894
+ "act_data_type": "float"
1895
+ },
1896
+ "model.llm.layers.41.attn.wq_du": {
1897
+ "bits": 16,
1898
+ "data_type": "float",
1899
+ "act_bits": 16,
1900
+ "act_data_type": "float"
1901
+ },
1902
+ "model.llm.layers.5.attn.wq_du": {
1903
+ "bits": 16,
1904
+ "data_type": "float",
1905
+ "act_bits": 16,
1906
+ "act_data_type": "float"
1907
+ },
1908
+ "model.llm.layers.7.attn.wq_du": {
1909
+ "bits": 16,
1910
+ "data_type": "float",
1911
+ "act_bits": 16,
1912
+ "act_data_type": "float"
1913
+ },
1914
+ "model.llm.layers.11.attn.wq_du": {
1915
+ "bits": 16,
1916
+ "data_type": "float",
1917
+ "act_bits": 16,
1918
+ "act_data_type": "float"
1919
+ },
1920
+ "model.llm.layers.37.attn.wv_dv": {
1921
+ "bits": 16,
1922
+ "data_type": "float",
1923
+ "act_bits": 16,
1924
+ "act_data_type": "float"
1925
+ },
1926
+ "model.llm.layers.41.attn.wk_dv": {
1927
+ "bits": 16,
1928
+ "data_type": "float",
1929
+ "act_bits": 16,
1930
+ "act_data_type": "float"
1931
+ },
1932
+ "model.llm.layers.26.attn.wq_du": {
1933
+ "bits": 16,
1934
+ "data_type": "float",
1935
+ "act_bits": 16,
1936
+ "act_data_type": "float"
1937
+ },
1938
+ "model.llm.layers.32.mlp.gate": {
1939
+ "bits": 16,
1940
+ "data_type": "float",
1941
+ "act_bits": 16,
1942
+ "act_data_type": "float"
1943
+ },
1944
+ "model.llm.layers.5.attn.wo_ud": {
1945
+ "bits": 16,
1946
+ "data_type": "float",
1947
+ "act_bits": 16,
1948
+ "act_data_type": "float"
1949
+ },
1950
+ "model.llm.layers.2.attn.wq_du": {
1951
+ "bits": 16,
1952
+ "data_type": "float",
1953
+ "act_bits": 16,
1954
+ "act_data_type": "float"
1955
+ },
1956
+ "model.llm.layers.34.mlp.gate": {
1957
+ "bits": 16,
1958
+ "data_type": "float",
1959
+ "act_bits": 16,
1960
+ "act_data_type": "float"
1961
+ }
1962
+ }
1963
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|endoftext|>",
4
+ "<|message_user|>",
5
+ "<|message_model|>",
6
+ "<|message_system|>",
7
+ "<|message_tool|>",
8
+ "<|content_text|>",
9
+ "<|content_image|>",
10
+ "<|content_model_end_sampling|>",
11
+ "<|content_thinking|>",
12
+ "<|end_message|>",
13
+ "<|content_audio_input|>",
14
+ "<|content_tool_error|>",
15
+ "<|audio|>",
16
+ "<|content_xml|>",
17
+ "<|begin_of_text|>",
18
+ "<|audio_end|>",
19
+ "<|content_invoke_tool_json|>",
20
+ "<|content_invoke_tool_text|>"
21
+ ]
22
+ }
tiktoken/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc253fd2b702f7a6da7105eaa8f3463b2f1247e83614f23e5323b921088bed2a
3
+ size 3615874
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fb6333a7db8fe5da90728e741e4a3ee4ac2ae12c5dd4958cc6f31688787d3c2
3
+ size 27875797