Hugging Face's logo

314e
/
llama3.2-quantised

Image-Text-to-Text
Transformers
Safetensors
mllama
conversational
text-generation-inference
Inference Endpoints
compressed-tensors
Model card Files Community
developer-314e commited on
Commit
5c1fa39
·
verified ·
1 Parent(s): 58539b4

Upload MllamaForConditionalGeneration

Browse files
Files changed (6) hide show
  1. README.md +199 -0
  2. config.json +835 -0
  3. generation_config.json +13 -0
  4. model-00001-of-00002.safetensors +3 -0
  5. model-00002-of-00002.safetensors +3 -0
  6. model.safetensors.index.json +0 -0
README.md ADDED
@@ -0,0 +1,199 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ tags: []
4
+ ---
5
+
6
+ # Model Card for Model ID
7
+
8
+ <!-- Provide a quick summary of what the model is/does. -->
9
+
10
+
11
+
12
+ ## Model Details
13
+
14
+ ### Model Description
15
+
16
+ <!-- Provide a longer summary of what this model is. -->
17
+
18
+ This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
19
+
20
+ - **Developed by:** [More Information Needed]
21
+ - **Funded by [optional]:** [More Information Needed]
22
+ - **Shared by [optional]:** [More Information Needed]
23
+ - **Model type:** [More Information Needed]
24
+ - **Language(s) (NLP):** [More Information Needed]
25
+ - **License:** [More Information Needed]
26
+ - **Finetuned from model [optional]:** [More Information Needed]
27
+
28
+ ### Model Sources [optional]
29
+
30
+ <!-- Provide the basic links for the model. -->
31
+
32
+ - **Repository:** [More Information Needed]
33
+ - **Paper [optional]:** [More Information Needed]
34
+ - **Demo [optional]:** [More Information Needed]
35
+
36
+ ## Uses
37
+
38
+ <!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
39
+
40
+ ### Direct Use
41
+
42
+ <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
43
+
44
+ [More Information Needed]
45
+
46
+ ### Downstream Use [optional]
47
+
48
+ <!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
49
+
50
+ [More Information Needed]
51
+
52
+ ### Out-of-Scope Use
53
+
54
+ <!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
55
+
56
+ [More Information Needed]
57
+
58
+ ## Bias, Risks, and Limitations
59
+
60
+ <!-- This section is meant to convey both technical and sociotechnical limitations. -->
61
+
62
+ [More Information Needed]
63
+
64
+ ### Recommendations
65
+
66
+ <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
67
+
68
+ Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
69
+
70
+ ## How to Get Started with the Model
71
+
72
+ Use the code below to get started with the model.
73
+
74
+ [More Information Needed]
75
+
76
+ ## Training Details
77
+
78
+ ### Training Data
79
+
80
+ <!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
81
+
82
+ [More Information Needed]
83
+
84
+ ### Training Procedure
85
+
86
+ <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
87
+
88
+ #### Preprocessing [optional]
89
+
90
+ [More Information Needed]
91
+
92
+
93
+ #### Training Hyperparameters
94
+
95
+ - **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
96
+
97
+ #### Speeds, Sizes, Times [optional]
98
+
99
+ <!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
100
+
101
+ [More Information Needed]
102
+
103
+ ## Evaluation
104
+
105
+ <!-- This section describes the evaluation protocols and provides the results. -->
106
+
107
+ ### Testing Data, Factors & Metrics
108
+
109
+ #### Testing Data
110
+
111
+ <!-- This should link to a Dataset Card if possible. -->
112
+
113
+ [More Information Needed]
114
+
115
+ #### Factors
116
+
117
+ <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
118
+
119
+ [More Information Needed]
120
+
121
+ #### Metrics
122
+
123
+ <!-- These are the evaluation metrics being used, ideally with a description of why. -->
124
+
125
+ [More Information Needed]
126
+
127
+ ### Results
128
+
129
+ [More Information Needed]
130
+
131
+ #### Summary
132
+
133
+
134
+
135
+ ## Model Examination [optional]
136
+
137
+ <!-- Relevant interpretability work for the model goes here -->
138
+
139
+ [More Information Needed]
140
+
141
+ ## Environmental Impact
142
+
143
+ <!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
144
+
145
+ Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
146
+
147
+ - **Hardware Type:** [More Information Needed]
148
+ - **Hours used:** [More Information Needed]
149
+ - **Cloud Provider:** [More Information Needed]
150
+ - **Compute Region:** [More Information Needed]
151
+ - **Carbon Emitted:** [More Information Needed]
152
+
153
+ ## Technical Specifications [optional]
154
+
155
+ ### Model Architecture and Objective
156
+
157
+ [More Information Needed]
158
+
159
+ ### Compute Infrastructure
160
+
161
+ [More Information Needed]
162
+
163
+ #### Hardware
164
+
165
+ [More Information Needed]
166
+
167
+ #### Software
168
+
169
+ [More Information Needed]
170
+
171
+ ## Citation [optional]
172
+
173
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
174
+
175
+ **BibTeX:**
176
+
177
+ [More Information Needed]
178
+
179
+ **APA:**
180
+
181
+ [More Information Needed]
182
+
183
+ ## Glossary [optional]
184
+
185
+ <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
186
+
187
+ [More Information Needed]
188
+
189
+ ## More Information [optional]
190
+
191
+ [More Information Needed]
192
+
193
+ ## Model Card Authors [optional]
194
+
195
+ [More Information Needed]
196
+
197
+ ## Model Card Contact
198
+
199
+ [More Information Needed]
config.json ADDED
@@ -0,0 +1,835 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "Llama-3.2-11B-Vision-Instruct-W4A16-G128",
3
+ "architectures": [
4
+ "MllamaForConditionalGeneration"
5
+ ],
6
+ "image_token_index": 128256,
7
+ "model_type": "mllama",
8
+ "quantization_config": {
9
+ "config_groups": {
10
+ "group_0": {
11
+ "input_activations": null,
12
+ "output_activations": null,
13
+ "targets": [
14
+ "Linear"
15
+ ],
16
+ "weights": {
17
+ "actorder": null,
18
+ "block_structure": null,
19
+ "dynamic": false,
20
+ "group_size": 128,
21
+ "num_bits": 4,
22
+ "observer": "minmax",
23
+ "observer_kwargs": {},
24
+ "strategy": "group",
25
+ "symmetric": true,
26
+ "type": "int"
27
+ }
28
+ }
29
+ },
30
+ "format": "pack-quantized",
31
+ "global_compression_ratio": 1.363428206546751,
32
+ "ignore": [
33
+ "vision_model.transformer.layers.0.self_attn.q_proj",
34
+ "vision_model.transformer.layers.0.self_attn.k_proj",
35
+ "vision_model.transformer.layers.0.self_attn.v_proj",
36
+ "vision_model.transformer.layers.0.self_attn.o_proj",
37
+ "vision_model.transformer.layers.0.mlp.fc1",
38
+ "vision_model.transformer.layers.0.mlp.fc2",
39
+ "vision_model.transformer.layers.1.self_attn.q_proj",
40
+ "vision_model.transformer.layers.1.self_attn.k_proj",
41
+ "vision_model.transformer.layers.1.self_attn.v_proj",
42
+ "vision_model.transformer.layers.1.self_attn.o_proj",
43
+ "vision_model.transformer.layers.1.mlp.fc1",
44
+ "vision_model.transformer.layers.1.mlp.fc2",
45
+ "vision_model.transformer.layers.2.self_attn.q_proj",
46
+ "vision_model.transformer.layers.2.self_attn.k_proj",
47
+ "vision_model.transformer.layers.2.self_attn.v_proj",
48
+ "vision_model.transformer.layers.2.self_attn.o_proj",
49
+ "vision_model.transformer.layers.2.mlp.fc1",
50
+ "vision_model.transformer.layers.2.mlp.fc2",
51
+ "vision_model.transformer.layers.3.self_attn.q_proj",
52
+ "vision_model.transformer.layers.3.self_attn.k_proj",
53
+ "vision_model.transformer.layers.3.self_attn.v_proj",
54
+ "vision_model.transformer.layers.3.self_attn.o_proj",
55
+ "vision_model.transformer.layers.3.mlp.fc1",
56
+ "vision_model.transformer.layers.3.mlp.fc2",
57
+ "vision_model.transformer.layers.4.self_attn.q_proj",
58
+ "vision_model.transformer.layers.4.self_attn.k_proj",
59
+ "vision_model.transformer.layers.4.self_attn.v_proj",
60
+ "vision_model.transformer.layers.4.self_attn.o_proj",
61
+ "vision_model.transformer.layers.4.mlp.fc1",
62
+ "vision_model.transformer.layers.4.mlp.fc2",
63
+ "vision_model.transformer.layers.5.self_attn.q_proj",
64
+ "vision_model.transformer.layers.5.self_attn.k_proj",
65
+ "vision_model.transformer.layers.5.self_attn.v_proj",
66
+ "vision_model.transformer.layers.5.self_attn.o_proj",
67
+ "vision_model.transformer.layers.5.mlp.fc1",
68
+ "vision_model.transformer.layers.5.mlp.fc2",
69
+ "vision_model.transformer.layers.6.self_attn.q_proj",
70
+ "vision_model.transformer.layers.6.self_attn.k_proj",
71
+ "vision_model.transformer.layers.6.self_attn.v_proj",
72
+ "vision_model.transformer.layers.6.self_attn.o_proj",
73
+ "vision_model.transformer.layers.6.mlp.fc1",
74
+ "vision_model.transformer.layers.6.mlp.fc2",
75
+ "vision_model.transformer.layers.7.self_attn.q_proj",
76
+ "vision_model.transformer.layers.7.self_attn.k_proj",
77
+ "vision_model.transformer.layers.7.self_attn.v_proj",
78
+ "vision_model.transformer.layers.7.self_attn.o_proj",
79
+ "vision_model.transformer.layers.7.mlp.fc1",
80
+ "vision_model.transformer.layers.7.mlp.fc2",
81
+ "vision_model.transformer.layers.8.self_attn.q_proj",
82
+ "vision_model.transformer.layers.8.self_attn.k_proj",
83
+ "vision_model.transformer.layers.8.self_attn.v_proj",
84
+ "vision_model.transformer.layers.8.self_attn.o_proj",
85
+ "vision_model.transformer.layers.8.mlp.fc1",
86
+ "vision_model.transformer.layers.8.mlp.fc2",
87
+ "vision_model.transformer.layers.9.self_attn.q_proj",
88
+ "vision_model.transformer.layers.9.self_attn.k_proj",
89
+ "vision_model.transformer.layers.9.self_attn.v_proj",
90
+ "vision_model.transformer.layers.9.self_attn.o_proj",
91
+ "vision_model.transformer.layers.9.mlp.fc1",
92
+ "vision_model.transformer.layers.9.mlp.fc2",
93
+ "vision_model.transformer.layers.10.self_attn.q_proj",
94
+ "vision_model.transformer.layers.10.self_attn.k_proj",
95
+ "vision_model.transformer.layers.10.self_attn.v_proj",
96
+ "vision_model.transformer.layers.10.self_attn.o_proj",
97
+ "vision_model.transformer.layers.10.mlp.fc1",
98
+ "vision_model.transformer.layers.10.mlp.fc2",
99
+ "vision_model.transformer.layers.11.self_attn.q_proj",
100
+ "vision_model.transformer.layers.11.self_attn.k_proj",
101
+ "vision_model.transformer.layers.11.self_attn.v_proj",
102
+ "vision_model.transformer.layers.11.self_attn.o_proj",
103
+ "vision_model.transformer.layers.11.mlp.fc1",
104
+ "vision_model.transformer.layers.11.mlp.fc2",
105
+ "vision_model.transformer.layers.12.self_attn.q_proj",
106
+ "vision_model.transformer.layers.12.self_attn.k_proj",
107
+ "vision_model.transformer.layers.12.self_attn.v_proj",
108
+ "vision_model.transformer.layers.12.self_attn.o_proj",
109
+ "vision_model.transformer.layers.12.mlp.fc1",
110
+ "vision_model.transformer.layers.12.mlp.fc2",
111
+ "vision_model.transformer.layers.13.self_attn.q_proj",
112
+ "vision_model.transformer.layers.13.self_attn.k_proj",
113
+ "vision_model.transformer.layers.13.self_attn.v_proj",
114
+ "vision_model.transformer.layers.13.self_attn.o_proj",
115
+ "vision_model.transformer.layers.13.mlp.fc1",
116
+ "vision_model.transformer.layers.13.mlp.fc2",
117
+ "vision_model.transformer.layers.14.self_attn.q_proj",
118
+ "vision_model.transformer.layers.14.self_attn.k_proj",
119
+ "vision_model.transformer.layers.14.self_attn.v_proj",
120
+ "vision_model.transformer.layers.14.self_attn.o_proj",
121
+ "vision_model.transformer.layers.14.mlp.fc1",
122
+ "vision_model.transformer.layers.14.mlp.fc2",
123
+ "vision_model.transformer.layers.15.self_attn.q_proj",
124
+ "vision_model.transformer.layers.15.self_attn.k_proj",
125
+ "vision_model.transformer.layers.15.self_attn.v_proj",
126
+ "vision_model.transformer.layers.15.self_attn.o_proj",
127
+ "vision_model.transformer.layers.15.mlp.fc1",
128
+ "vision_model.transformer.layers.15.mlp.fc2",
129
+ "vision_model.transformer.layers.16.self_attn.q_proj",
130
+ "vision_model.transformer.layers.16.self_attn.k_proj",
131
+ "vision_model.transformer.layers.16.self_attn.v_proj",
132
+ "vision_model.transformer.layers.16.self_attn.o_proj",
133
+ "vision_model.transformer.layers.16.mlp.fc1",
134
+ "vision_model.transformer.layers.16.mlp.fc2",
135
+ "vision_model.transformer.layers.17.self_attn.q_proj",
136
+ "vision_model.transformer.layers.17.self_attn.k_proj",
137
+ "vision_model.transformer.layers.17.self_attn.v_proj",
138
+ "vision_model.transformer.layers.17.self_attn.o_proj",
139
+ "vision_model.transformer.layers.17.mlp.fc1",
140
+ "vision_model.transformer.layers.17.mlp.fc2",
141
+ "vision_model.transformer.layers.18.self_attn.q_proj",
142
+ "vision_model.transformer.layers.18.self_attn.k_proj",
143
+ "vision_model.transformer.layers.18.self_attn.v_proj",
144
+ "vision_model.transformer.layers.18.self_attn.o_proj",
145
+ "vision_model.transformer.layers.18.mlp.fc1",
146
+ "vision_model.transformer.layers.18.mlp.fc2",
147
+ "vision_model.transformer.layers.19.self_attn.q_proj",
148
+ "vision_model.transformer.layers.19.self_attn.k_proj",
149
+ "vision_model.transformer.layers.19.self_attn.v_proj",
150
+ "vision_model.transformer.layers.19.self_attn.o_proj",
151
+ "vision_model.transformer.layers.19.mlp.fc1",
152
+ "vision_model.transformer.layers.19.mlp.fc2",
153
+ "vision_model.transformer.layers.20.self_attn.q_proj",
154
+ "vision_model.transformer.layers.20.self_attn.k_proj",
155
+ "vision_model.transformer.layers.20.self_attn.v_proj",
156
+ "vision_model.transformer.layers.20.self_attn.o_proj",
157
+ "vision_model.transformer.layers.20.mlp.fc1",
158
+ "vision_model.transformer.layers.20.mlp.fc2",
159
+ "vision_model.transformer.layers.21.self_attn.q_proj",
160
+ "vision_model.transformer.layers.21.self_attn.k_proj",
161
+ "vision_model.transformer.layers.21.self_attn.v_proj",
162
+ "vision_model.transformer.layers.21.self_attn.o_proj",
163
+ "vision_model.transformer.layers.21.mlp.fc1",
164
+ "vision_model.transformer.layers.21.mlp.fc2",
165
+ "vision_model.transformer.layers.22.self_attn.q_proj",
166
+ "vision_model.transformer.layers.22.self_attn.k_proj",
167
+ "vision_model.transformer.layers.22.self_attn.v_proj",
168
+ "vision_model.transformer.layers.22.self_attn.o_proj",
169
+ "vision_model.transformer.layers.22.mlp.fc1",
170
+ "vision_model.transformer.layers.22.mlp.fc2",
171
+ "vision_model.transformer.layers.23.self_attn.q_proj",
172
+ "vision_model.transformer.layers.23.self_attn.k_proj",
173
+ "vision_model.transformer.layers.23.self_attn.v_proj",
174
+ "vision_model.transformer.layers.23.self_attn.o_proj",
175
+ "vision_model.transformer.layers.23.mlp.fc1",
176
+ "vision_model.transformer.layers.23.mlp.fc2",
177
+ "vision_model.transformer.layers.24.self_attn.q_proj",
178
+ "vision_model.transformer.layers.24.self_attn.k_proj",
179
+ "vision_model.transformer.layers.24.self_attn.v_proj",
180
+ "vision_model.transformer.layers.24.self_attn.o_proj",
181
+ "vision_model.transformer.layers.24.mlp.fc1",
182
+ "vision_model.transformer.layers.24.mlp.fc2",
183
+ "vision_model.transformer.layers.25.self_attn.q_proj",
184
+ "vision_model.transformer.layers.25.self_attn.k_proj",
185
+ "vision_model.transformer.layers.25.self_attn.v_proj",
186
+ "vision_model.transformer.layers.25.self_attn.o_proj",
187
+ "vision_model.transformer.layers.25.mlp.fc1",
188
+ "vision_model.transformer.layers.25.mlp.fc2",
189
+ "vision_model.transformer.layers.26.self_attn.q_proj",
190
+ "vision_model.transformer.layers.26.self_attn.k_proj",
191
+ "vision_model.transformer.layers.26.self_attn.v_proj",
192
+ "vision_model.transformer.layers.26.self_attn.o_proj",
193
+ "vision_model.transformer.layers.26.mlp.fc1",
194
+ "vision_model.transformer.layers.26.mlp.fc2",
195
+ "vision_model.transformer.layers.27.self_attn.q_proj",
196
+ "vision_model.transformer.layers.27.self_attn.k_proj",
197
+ "vision_model.transformer.layers.27.self_attn.v_proj",
198
+ "vision_model.transformer.layers.27.self_attn.o_proj",
199
+ "vision_model.transformer.layers.27.mlp.fc1",
200
+ "vision_model.transformer.layers.27.mlp.fc2",
201
+ "vision_model.transformer.layers.28.self_attn.q_proj",
202
+ "vision_model.transformer.layers.28.self_attn.k_proj",
203
+ "vision_model.transformer.layers.28.self_attn.v_proj",
204
+ "vision_model.transformer.layers.28.self_attn.o_proj",
205
+ "vision_model.transformer.layers.28.mlp.fc1",
206
+ "vision_model.transformer.layers.28.mlp.fc2",
207
+ "vision_model.transformer.layers.29.self_attn.q_proj",
208
+ "vision_model.transformer.layers.29.self_attn.k_proj",
209
+ "vision_model.transformer.layers.29.self_attn.v_proj",
210
+ "vision_model.transformer.layers.29.self_attn.o_proj",
211
+ "vision_model.transformer.layers.29.mlp.fc1",
212
+ "vision_model.transformer.layers.29.mlp.fc2",
213
+ "vision_model.transformer.layers.30.self_attn.q_proj",
214
+ "vision_model.transformer.layers.30.self_attn.k_proj",
215
+ "vision_model.transformer.layers.30.self_attn.v_proj",
216
+ "vision_model.transformer.layers.30.self_attn.o_proj",
217
+ "vision_model.transformer.layers.30.mlp.fc1",
218
+ "vision_model.transformer.layers.30.mlp.fc2",
219
+ "vision_model.transformer.layers.31.self_attn.q_proj",
220
+ "vision_model.transformer.layers.31.self_attn.k_proj",
221
+ "vision_model.transformer.layers.31.self_attn.v_proj",
222
+ "vision_model.transformer.layers.31.self_attn.o_proj",
223
+ "vision_model.transformer.layers.31.mlp.fc1",
224
+ "vision_model.transformer.layers.31.mlp.fc2",
225
+ "vision_model.global_transformer.layers.0.self_attn.q_proj",
226
+ "vision_model.global_transformer.layers.0.self_attn.k_proj",
227
+ "vision_model.global_transformer.layers.0.self_attn.v_proj",
228
+ "vision_model.global_transformer.layers.0.self_attn.o_proj",
229
+ "vision_model.global_transformer.layers.0.mlp.fc1",
230
+ "vision_model.global_transformer.layers.0.mlp.fc2",
231
+ "vision_model.global_transformer.layers.1.self_attn.q_proj",
232
+ "vision_model.global_transformer.layers.1.self_attn.k_proj",
233
+ "vision_model.global_transformer.layers.1.self_attn.v_proj",
234
+ "vision_model.global_transformer.layers.1.self_attn.o_proj",
235
+ "vision_model.global_transformer.layers.1.mlp.fc1",
236
+ "vision_model.global_transformer.layers.1.mlp.fc2",
237
+ "vision_model.global_transformer.layers.2.self_attn.q_proj",
238
+ "vision_model.global_transformer.layers.2.self_attn.k_proj",
239
+ "vision_model.global_transformer.layers.2.self_attn.v_proj",
240
+ "vision_model.global_transformer.layers.2.self_attn.o_proj",
241
+ "vision_model.global_transformer.layers.2.mlp.fc1",
242
+ "vision_model.global_transformer.layers.2.mlp.fc2",
243
+ "vision_model.global_transformer.layers.3.self_attn.q_proj",
244
+ "vision_model.global_transformer.layers.3.self_attn.k_proj",
245
+ "vision_model.global_transformer.layers.3.self_attn.v_proj",
246
+ "vision_model.global_transformer.layers.3.self_attn.o_proj",
247
+ "vision_model.global_transformer.layers.3.mlp.fc1",
248
+ "vision_model.global_transformer.layers.3.mlp.fc2",
249
+ "vision_model.global_transformer.layers.4.self_attn.q_proj",
250
+ "vision_model.global_transformer.layers.4.self_attn.k_proj",
251
+ "vision_model.global_transformer.layers.4.self_attn.v_proj",
252
+ "vision_model.global_transformer.layers.4.self_attn.o_proj",
253
+ "vision_model.global_transformer.layers.4.mlp.fc1",
254
+ "vision_model.global_transformer.layers.4.mlp.fc2",
255
+ "vision_model.global_transformer.layers.5.self_attn.q_proj",
256
+ "vision_model.global_transformer.layers.5.self_attn.k_proj",
257
+ "vision_model.global_transformer.layers.5.self_attn.v_proj",
258
+ "vision_model.global_transformer.layers.5.self_attn.o_proj",
259
+ "vision_model.global_transformer.layers.5.mlp.fc1",
260
+ "vision_model.global_transformer.layers.5.mlp.fc2",
261
+ "vision_model.global_transformer.layers.6.self_attn.q_proj",
262
+ "vision_model.global_transformer.layers.6.self_attn.k_proj",
263
+ "vision_model.global_transformer.layers.6.self_attn.v_proj",
264
+ "vision_model.global_transformer.layers.6.self_attn.o_proj",
265
+ "vision_model.global_transformer.layers.6.mlp.fc1",
266
+ "vision_model.global_transformer.layers.6.mlp.fc2",
267
+ "vision_model.global_transformer.layers.7.self_attn.q_proj",
268
+ "vision_model.global_transformer.layers.7.self_attn.k_proj",
269
+ "vision_model.global_transformer.layers.7.self_attn.v_proj",
270
+ "vision_model.global_transformer.layers.7.self_attn.o_proj",
271
+ "vision_model.global_transformer.layers.7.mlp.fc1",
272
+ "vision_model.global_transformer.layers.7.mlp.fc2",
273
+ "language_model.lm_head",
274
+ "multi_modal_projector"
275
+ ],
276
+ "kv_cache_scheme": null,
277
+ "quant_method": "compressed-tensors",
278
+ "quantization_status": "compressed",
279
+ "sparsity_config": {
280
+ "format": "dense",
281
+ "global_sparsity": 0.13409465208989965,
282
+ "ignore": [
283
+ "vision_model.transformer.layers.0.self_attn.q_proj",
284
+ "vision_model.transformer.layers.0.self_attn.k_proj",
285
+ "vision_model.transformer.layers.0.self_attn.v_proj",
286
+ "vision_model.transformer.layers.0.self_attn.o_proj",
287
+ "vision_model.transformer.layers.0.mlp.fc1",
288
+ "vision_model.transformer.layers.0.mlp.fc2",
289
+ "vision_model.transformer.layers.1.self_attn.q_proj",
290
+ "vision_model.transformer.layers.1.self_attn.k_proj",
291
+ "vision_model.transformer.layers.1.self_attn.v_proj",
292
+ "vision_model.transformer.layers.1.self_attn.o_proj",
293
+ "vision_model.transformer.layers.1.mlp.fc1",
294
+ "vision_model.transformer.layers.1.mlp.fc2",
295
+ "vision_model.transformer.layers.2.self_attn.q_proj",
296
+ "vision_model.transformer.layers.2.self_attn.k_proj",
297
+ "vision_model.transformer.layers.2.self_attn.v_proj",
298
+ "vision_model.transformer.layers.2.self_attn.o_proj",
299
+ "vision_model.transformer.layers.2.mlp.fc1",
300
+ "vision_model.transformer.layers.2.mlp.fc2",
301
+ "vision_model.transformer.layers.3.self_attn.q_proj",
302
+ "vision_model.transformer.layers.3.self_attn.k_proj",
303
+ "vision_model.transformer.layers.3.self_attn.v_proj",
304
+ "vision_model.transformer.layers.3.self_attn.o_proj",
305
+ "vision_model.transformer.layers.3.mlp.fc1",
306
+ "vision_model.transformer.layers.3.mlp.fc2",
307
+ "vision_model.transformer.layers.4.self_attn.q_proj",
308
+ "vision_model.transformer.layers.4.self_attn.k_proj",
309
+ "vision_model.transformer.layers.4.self_attn.v_proj",
310
+ "vision_model.transformer.layers.4.self_attn.o_proj",
311
+ "vision_model.transformer.layers.4.mlp.fc1",
312
+ "vision_model.transformer.layers.4.mlp.fc2",
313
+ "vision_model.transformer.layers.5.self_attn.q_proj",
314
+ "vision_model.transformer.layers.5.self_attn.k_proj",
315
+ "vision_model.transformer.layers.5.self_attn.v_proj",
316
+ "vision_model.transformer.layers.5.self_attn.o_proj",
317
+ "vision_model.transformer.layers.5.mlp.fc1",
318
+ "vision_model.transformer.layers.5.mlp.fc2",
319
+ "vision_model.transformer.layers.6.self_attn.q_proj",
320
+ "vision_model.transformer.layers.6.self_attn.k_proj",
321
+ "vision_model.transformer.layers.6.self_attn.v_proj",
322
+ "vision_model.transformer.layers.6.self_attn.o_proj",
323
+ "vision_model.transformer.layers.6.mlp.fc1",
324
+ "vision_model.transformer.layers.6.mlp.fc2",
325
+ "vision_model.transformer.layers.7.self_attn.q_proj",
326
+ "vision_model.transformer.layers.7.self_attn.k_proj",
327
+ "vision_model.transformer.layers.7.self_attn.v_proj",
328
+ "vision_model.transformer.layers.7.self_attn.o_proj",
329
+ "vision_model.transformer.layers.7.mlp.fc1",
330
+ "vision_model.transformer.layers.7.mlp.fc2",
331
+ "vision_model.transformer.layers.8.self_attn.q_proj",
332
+ "vision_model.transformer.layers.8.self_attn.k_proj",
333
+ "vision_model.transformer.layers.8.self_attn.v_proj",
334
+ "vision_model.transformer.layers.8.self_attn.o_proj",
335
+ "vision_model.transformer.layers.8.mlp.fc1",
336
+ "vision_model.transformer.layers.8.mlp.fc2",
337
+ "vision_model.transformer.layers.9.self_attn.q_proj",
338
+ "vision_model.transformer.layers.9.self_attn.k_proj",
339
+ "vision_model.transformer.layers.9.self_attn.v_proj",
340
+ "vision_model.transformer.layers.9.self_attn.o_proj",
341
+ "vision_model.transformer.layers.9.mlp.fc1",
342
+ "vision_model.transformer.layers.9.mlp.fc2",
343
+ "vision_model.transformer.layers.10.self_attn.q_proj",
344
+ "vision_model.transformer.layers.10.self_attn.k_proj",
345
+ "vision_model.transformer.layers.10.self_attn.v_proj",
346
+ "vision_model.transformer.layers.10.self_attn.o_proj",
347
+ "vision_model.transformer.layers.10.mlp.fc1",
348
+ "vision_model.transformer.layers.10.mlp.fc2",
349
+ "vision_model.transformer.layers.11.self_attn.q_proj",
350
+ "vision_model.transformer.layers.11.self_attn.k_proj",
351
+ "vision_model.transformer.layers.11.self_attn.v_proj",
352
+ "vision_model.transformer.layers.11.self_attn.o_proj",
353
+ "vision_model.transformer.layers.11.mlp.fc1",
354
+ "vision_model.transformer.layers.11.mlp.fc2",
355
+ "vision_model.transformer.layers.12.self_attn.q_proj",
356
+ "vision_model.transformer.layers.12.self_attn.k_proj",
357
+ "vision_model.transformer.layers.12.self_attn.v_proj",
358
+ "vision_model.transformer.layers.12.self_attn.o_proj",
359
+ "vision_model.transformer.layers.12.mlp.fc1",
360
+ "vision_model.transformer.layers.12.mlp.fc2",
361
+ "vision_model.transformer.layers.13.self_attn.q_proj",
362
+ "vision_model.transformer.layers.13.self_attn.k_proj",
363
+ "vision_model.transformer.layers.13.self_attn.v_proj",
364
+ "vision_model.transformer.layers.13.self_attn.o_proj",
365
+ "vision_model.transformer.layers.13.mlp.fc1",
366
+ "vision_model.transformer.layers.13.mlp.fc2",
367
+ "vision_model.transformer.layers.14.self_attn.q_proj",
368
+ "vision_model.transformer.layers.14.self_attn.k_proj",
369
+ "vision_model.transformer.layers.14.self_attn.v_proj",
370
+ "vision_model.transformer.layers.14.self_attn.o_proj",
371
+ "vision_model.transformer.layers.14.mlp.fc1",
372
+ "vision_model.transformer.layers.14.mlp.fc2",
373
+ "vision_model.transformer.layers.15.self_attn.q_proj",
374
+ "vision_model.transformer.layers.15.self_attn.k_proj",
375
+ "vision_model.transformer.layers.15.self_attn.v_proj",
376
+ "vision_model.transformer.layers.15.self_attn.o_proj",
377
+ "vision_model.transformer.layers.15.mlp.fc1",
378
+ "vision_model.transformer.layers.15.mlp.fc2",
379
+ "vision_model.transformer.layers.16.self_attn.q_proj",
380
+ "vision_model.transformer.layers.16.self_attn.k_proj",
381
+ "vision_model.transformer.layers.16.self_attn.v_proj",
382
+ "vision_model.transformer.layers.16.self_attn.o_proj",
383
+ "vision_model.transformer.layers.16.mlp.fc1",
384
+ "vision_model.transformer.layers.16.mlp.fc2",
385
+ "vision_model.transformer.layers.17.self_attn.q_proj",
386
+ "vision_model.transformer.layers.17.self_attn.k_proj",
387
+ "vision_model.transformer.layers.17.self_attn.v_proj",
388
+ "vision_model.transformer.layers.17.self_attn.o_proj",
389
+ "vision_model.transformer.layers.17.mlp.fc1",
390
+ "vision_model.transformer.layers.17.mlp.fc2",
391
+ "vision_model.transformer.layers.18.self_attn.q_proj",
392
+ "vision_model.transformer.layers.18.self_attn.k_proj",
393
+ "vision_model.transformer.layers.18.self_attn.v_proj",
394
+ "vision_model.transformer.layers.18.self_attn.o_proj",
395
+ "vision_model.transformer.layers.18.mlp.fc1",
396
+ "vision_model.transformer.layers.18.mlp.fc2",
397
+ "vision_model.transformer.layers.19.self_attn.q_proj",
398
+ "vision_model.transformer.layers.19.self_attn.k_proj",
399
+ "vision_model.transformer.layers.19.self_attn.v_proj",
400
+ "vision_model.transformer.layers.19.self_attn.o_proj",
401
+ "vision_model.transformer.layers.19.mlp.fc1",
402
+ "vision_model.transformer.layers.19.mlp.fc2",
403
+ "vision_model.transformer.layers.20.self_attn.q_proj",
404
+ "vision_model.transformer.layers.20.self_attn.k_proj",
405
+ "vision_model.transformer.layers.20.self_attn.v_proj",
406
+ "vision_model.transformer.layers.20.self_attn.o_proj",
407
+ "vision_model.transformer.layers.20.mlp.fc1",
408
+ "vision_model.transformer.layers.20.mlp.fc2",
409
+ "vision_model.transformer.layers.21.self_attn.q_proj",
410
+ "vision_model.transformer.layers.21.self_attn.k_proj",
411
+ "vision_model.transformer.layers.21.self_attn.v_proj",
412
+ "vision_model.transformer.layers.21.self_attn.o_proj",
413
+ "vision_model.transformer.layers.21.mlp.fc1",
414
+ "vision_model.transformer.layers.21.mlp.fc2",
415
+ "vision_model.transformer.layers.22.self_attn.q_proj",
416
+ "vision_model.transformer.layers.22.self_attn.k_proj",
417
+ "vision_model.transformer.layers.22.self_attn.v_proj",
418
+ "vision_model.transformer.layers.22.self_attn.o_proj",
419
+ "vision_model.transformer.layers.22.mlp.fc1",
420
+ "vision_model.transformer.layers.22.mlp.fc2",
421
+ "vision_model.transformer.layers.23.self_attn.q_proj",
422
+ "vision_model.transformer.layers.23.self_attn.k_proj",
423
+ "vision_model.transformer.layers.23.self_attn.v_proj",
424
+ "vision_model.transformer.layers.23.self_attn.o_proj",
425
+ "vision_model.transformer.layers.23.mlp.fc1",
426
+ "vision_model.transformer.layers.23.mlp.fc2",
427
+ "vision_model.transformer.layers.24.self_attn.q_proj",
428
+ "vision_model.transformer.layers.24.self_attn.k_proj",
429
+ "vision_model.transformer.layers.24.self_attn.v_proj",
430
+ "vision_model.transformer.layers.24.self_attn.o_proj",
431
+ "vision_model.transformer.layers.24.mlp.fc1",
432
+ "vision_model.transformer.layers.24.mlp.fc2",
433
+ "vision_model.transformer.layers.25.self_attn.q_proj",
434
+ "vision_model.transformer.layers.25.self_attn.k_proj",
435
+ "vision_model.transformer.layers.25.self_attn.v_proj",
436
+ "vision_model.transformer.layers.25.self_attn.o_proj",
437
+ "vision_model.transformer.layers.25.mlp.fc1",
438
+ "vision_model.transformer.layers.25.mlp.fc2",
439
+ "vision_model.transformer.layers.26.self_attn.q_proj",
440
+ "vision_model.transformer.layers.26.self_attn.k_proj",
441
+ "vision_model.transformer.layers.26.self_attn.v_proj",
442
+ "vision_model.transformer.layers.26.self_attn.o_proj",
443
+ "vision_model.transformer.layers.26.mlp.fc1",
444
+ "vision_model.transformer.layers.26.mlp.fc2",
445
+ "vision_model.transformer.layers.27.self_attn.q_proj",
446
+ "vision_model.transformer.layers.27.self_attn.k_proj",
447
+ "vision_model.transformer.layers.27.self_attn.v_proj",
448
+ "vision_model.transformer.layers.27.self_attn.o_proj",
449
+ "vision_model.transformer.layers.27.mlp.fc1",
450
+ "vision_model.transformer.layers.27.mlp.fc2",
451
+ "vision_model.transformer.layers.28.self_attn.q_proj",
452
+ "vision_model.transformer.layers.28.self_attn.k_proj",
453
+ "vision_model.transformer.layers.28.self_attn.v_proj",
454
+ "vision_model.transformer.layers.28.self_attn.o_proj",
455
+ "vision_model.transformer.layers.28.mlp.fc1",
456
+ "vision_model.transformer.layers.28.mlp.fc2",
457
+ "vision_model.transformer.layers.29.self_attn.q_proj",
458
+ "vision_model.transformer.layers.29.self_attn.k_proj",
459
+ "vision_model.transformer.layers.29.self_attn.v_proj",
460
+ "vision_model.transformer.layers.29.self_attn.o_proj",
461
+ "vision_model.transformer.layers.29.mlp.fc1",
462
+ "vision_model.transformer.layers.29.mlp.fc2",
463
+ "vision_model.transformer.layers.30.self_attn.q_proj",
464
+ "vision_model.transformer.layers.30.self_attn.k_proj",
465
+ "vision_model.transformer.layers.30.self_attn.v_proj",
466
+ "vision_model.transformer.layers.30.self_attn.o_proj",
467
+ "vision_model.transformer.layers.30.mlp.fc1",
468
+ "vision_model.transformer.layers.30.mlp.fc2",
469
+ "vision_model.transformer.layers.31.self_attn.q_proj",
470
+ "vision_model.transformer.layers.31.self_attn.k_proj",
471
+ "vision_model.transformer.layers.31.self_attn.v_proj",
472
+ "vision_model.transformer.layers.31.self_attn.o_proj",
473
+ "vision_model.transformer.layers.31.mlp.fc1",
474
+ "vision_model.transformer.layers.31.mlp.fc2",
475
+ "vision_model.global_transformer.layers.0.self_attn.q_proj",
476
+ "vision_model.global_transformer.layers.0.self_attn.k_proj",
477
+ "vision_model.global_transformer.layers.0.self_attn.v_proj",
478
+ "vision_model.global_transformer.layers.0.self_attn.o_proj",
479
+ "vision_model.global_transformer.layers.0.mlp.fc1",
480
+ "vision_model.global_transformer.layers.0.mlp.fc2",
481
+ "vision_model.global_transformer.layers.1.self_attn.q_proj",
482
+ "vision_model.global_transformer.layers.1.self_attn.k_proj",
483
+ "vision_model.global_transformer.layers.1.self_attn.v_proj",
484
+ "vision_model.global_transformer.layers.1.self_attn.o_proj",
485
+ "vision_model.global_transformer.layers.1.mlp.fc1",
486
+ "vision_model.global_transformer.layers.1.mlp.fc2",
487
+ "vision_model.global_transformer.layers.2.self_attn.q_proj",
488
+ "vision_model.global_transformer.layers.2.self_attn.k_proj",
489
+ "vision_model.global_transformer.layers.2.self_attn.v_proj",
490
+ "vision_model.global_transformer.layers.2.self_attn.o_proj",
491
+ "vision_model.global_transformer.layers.2.mlp.fc1",
492
+ "vision_model.global_transformer.layers.2.mlp.fc2",
493
+ "vision_model.global_transformer.layers.3.self_attn.q_proj",
494
+ "vision_model.global_transformer.layers.3.self_attn.k_proj",
495
+ "vision_model.global_transformer.layers.3.self_attn.v_proj",
496
+ "vision_model.global_transformer.layers.3.self_attn.o_proj",
497
+ "vision_model.global_transformer.layers.3.mlp.fc1",
498
+ "vision_model.global_transformer.layers.3.mlp.fc2",
499
+ "vision_model.global_transformer.layers.4.self_attn.q_proj",
500
+ "vision_model.global_transformer.layers.4.self_attn.k_proj",
501
+ "vision_model.global_transformer.layers.4.self_attn.v_proj",
502
+ "vision_model.global_transformer.layers.4.self_attn.o_proj",
503
+ "vision_model.global_transformer.layers.4.mlp.fc1",
504
+ "vision_model.global_transformer.layers.4.mlp.fc2",
505
+ "vision_model.global_transformer.layers.5.self_attn.q_proj",
506
+ "vision_model.global_transformer.layers.5.self_attn.k_proj",
507
+ "vision_model.global_transformer.layers.5.self_attn.v_proj",
508
+ "vision_model.global_transformer.layers.5.self_attn.o_proj",
509
+ "vision_model.global_transformer.layers.5.mlp.fc1",
510
+ "vision_model.global_transformer.layers.5.mlp.fc2",
511
+ "vision_model.global_transformer.layers.6.self_attn.q_proj",
512
+ "vision_model.global_transformer.layers.6.self_attn.k_proj",
513
+ "vision_model.global_transformer.layers.6.self_attn.v_proj",
514
+ "vision_model.global_transformer.layers.6.self_attn.o_proj",
515
+ "vision_model.global_transformer.layers.6.mlp.fc1",
516
+ "vision_model.global_transformer.layers.6.mlp.fc2",
517
+ "vision_model.global_transformer.layers.7.self_attn.q_proj",
518
+ "vision_model.global_transformer.layers.7.self_attn.k_proj",
519
+ "vision_model.global_transformer.layers.7.self_attn.v_proj",
520
+ "vision_model.global_transformer.layers.7.self_attn.o_proj",
521
+ "vision_model.global_transformer.layers.7.mlp.fc1",
522
+ "vision_model.global_transformer.layers.7.mlp.fc2",
523
+ "language_model.model.layers.0.self_attn.v_proj",
524
+ "language_model.model.layers.0.self_attn.o_proj",
525
+ "language_model.model.layers.0.mlp.gate_proj",
526
+ "language_model.model.layers.0.mlp.up_proj",
527
+ "language_model.model.layers.0.mlp.down_proj",
528
+ "language_model.model.layers.1.self_attn.q_proj",
529
+ "language_model.model.layers.1.self_attn.k_proj",
530
+ "language_model.model.layers.1.self_attn.v_proj",
531
+ "language_model.model.layers.1.self_attn.o_proj",
532
+ "language_model.model.layers.1.mlp.gate_proj",
533
+ "language_model.model.layers.1.mlp.up_proj",
534
+ "language_model.model.layers.1.mlp.down_proj",
535
+ "language_model.model.layers.2.self_attn.q_proj",
536
+ "language_model.model.layers.2.self_attn.k_proj",
537
+ "language_model.model.layers.2.self_attn.v_proj",
538
+ "language_model.model.layers.2.self_attn.o_proj",
539
+ "language_model.model.layers.2.mlp.gate_proj",
540
+ "language_model.model.layers.2.mlp.up_proj",
541
+ "language_model.model.layers.2.mlp.down_proj",
542
+ "language_model.model.layers.3.cross_attn.q_proj",
543
+ "language_model.model.layers.3.cross_attn.k_proj",
544
+ "language_model.model.layers.3.cross_attn.v_proj",
545
+ "language_model.model.layers.3.cross_attn.o_proj",
546
+ "language_model.model.layers.3.mlp.gate_proj",
547
+ "language_model.model.layers.3.mlp.up_proj",
548
+ "language_model.model.layers.3.mlp.down_proj",
549
+ "language_model.model.layers.4.self_attn.q_proj",
550
+ "language_model.model.layers.4.self_attn.k_proj",
551
+ "language_model.model.layers.4.self_attn.v_proj",
552
+ "language_model.model.layers.4.self_attn.o_proj",
553
+ "language_model.model.layers.4.mlp.gate_proj",
554
+ "language_model.model.layers.4.mlp.up_proj",
555
+ "language_model.model.layers.4.mlp.down_proj",
556
+ "language_model.model.layers.5.self_attn.q_proj",
557
+ "language_model.model.layers.5.self_attn.k_proj",
558
+ "language_model.model.layers.5.self_attn.v_proj",
559
+ "language_model.model.layers.5.self_attn.o_proj",
560
+ "language_model.model.layers.5.mlp.gate_proj",
561
+ "language_model.model.layers.5.mlp.up_proj",
562
+ "language_model.model.layers.5.mlp.down_proj",
563
+ "language_model.model.layers.6.self_attn.q_proj",
564
+ "language_model.model.layers.6.self_attn.k_proj",
565
+ "language_model.model.layers.6.self_attn.v_proj",
566
+ "language_model.model.layers.6.self_attn.o_proj",
567
+ "language_model.model.layers.6.mlp.gate_proj",
568
+ "language_model.model.layers.6.mlp.up_proj",
569
+ "language_model.model.layers.6.mlp.down_proj",
570
+ "language_model.model.layers.7.self_attn.q_proj",
571
+ "language_model.model.layers.7.self_attn.k_proj",
572
+ "language_model.model.layers.7.self_attn.v_proj",
573
+ "language_model.model.layers.7.self_attn.o_proj",
574
+ "language_model.model.layers.7.mlp.gate_proj",
575
+ "language_model.model.layers.7.mlp.up_proj",
576
+ "language_model.model.layers.7.mlp.down_proj",
577
+ "language_model.model.layers.8.cross_attn.q_proj",
578
+ "language_model.model.layers.8.cross_attn.k_proj",
579
+ "language_model.model.layers.8.cross_attn.v_proj",
580
+ "language_model.model.layers.8.cross_attn.o_proj",
581
+ "language_model.model.layers.8.mlp.gate_proj",
582
+ "language_model.model.layers.8.mlp.up_proj",
583
+ "language_model.model.layers.8.mlp.down_proj",
584
+ "language_model.model.layers.9.self_attn.q_proj",
585
+ "language_model.model.layers.9.self_attn.k_proj",
586
+ "language_model.model.layers.9.self_attn.v_proj",
587
+ "language_model.model.layers.9.self_attn.o_proj",
588
+ "language_model.model.layers.9.mlp.gate_proj",
589
+ "language_model.model.layers.9.mlp.up_proj",
590
+ "language_model.model.layers.9.mlp.down_proj",
591
+ "language_model.model.layers.10.self_attn.q_proj",
592
+ "language_model.model.layers.10.self_attn.k_proj",
593
+ "language_model.model.layers.10.self_attn.v_proj",
594
+ "language_model.model.layers.10.self_attn.o_proj",
595
+ "language_model.model.layers.10.mlp.gate_proj",
596
+ "language_model.model.layers.10.mlp.up_proj",
597
+ "language_model.model.layers.10.mlp.down_proj",
598
+ "language_model.model.layers.11.self_attn.q_proj",
599
+ "language_model.model.layers.11.self_attn.k_proj",
600
+ "language_model.model.layers.11.self_attn.v_proj",
601
+ "language_model.model.layers.11.self_attn.o_proj",
602
+ "language_model.model.layers.11.mlp.gate_proj",
603
+ "language_model.model.layers.11.mlp.up_proj",
604
+ "language_model.model.layers.11.mlp.down_proj",
605
+ "language_model.model.layers.12.self_attn.q_proj",
606
+ "language_model.model.layers.12.self_attn.k_proj",
607
+ "language_model.model.layers.12.self_attn.v_proj",
608
+ "language_model.model.layers.12.self_attn.o_proj",
609
+ "language_model.model.layers.12.mlp.gate_proj",
610
+ "language_model.model.layers.12.mlp.up_proj",
611
+ "language_model.model.layers.12.mlp.down_proj",
612
+ "language_model.model.layers.13.cross_attn.q_proj",
613
+ "language_model.model.layers.13.cross_attn.k_proj",
614
+ "language_model.model.layers.13.cross_attn.v_proj",
615
+ "language_model.model.layers.13.cross_attn.o_proj",
616
+ "language_model.model.layers.13.mlp.gate_proj",
617
+ "language_model.model.layers.13.mlp.up_proj",
618
+ "language_model.model.layers.13.mlp.down_proj",
619
+ "language_model.model.layers.14.self_attn.q_proj",
620
+ "language_model.model.layers.14.self_attn.k_proj",
621
+ "language_model.model.layers.14.self_attn.v_proj",
622
+ "language_model.model.layers.14.self_attn.o_proj",
623
+ "language_model.model.layers.14.mlp.gate_proj",
624
+ "language_model.model.layers.14.mlp.up_proj",
625
+ "language_model.model.layers.14.mlp.down_proj",
626
+ "language_model.model.layers.15.self_attn.q_proj",
627
+ "language_model.model.layers.15.self_attn.k_proj",
628
+ "language_model.model.layers.15.self_attn.v_proj",
629
+ "language_model.model.layers.15.self_attn.o_proj",
630
+ "language_model.model.layers.15.mlp.gate_proj",
631
+ "language_model.model.layers.15.mlp.up_proj",
632
+ "language_model.model.layers.15.mlp.down_proj",
633
+ "language_model.model.layers.16.self_attn.q_proj",
634
+ "language_model.model.layers.16.self_attn.k_proj",
635
+ "language_model.model.layers.16.self_attn.v_proj",
636
+ "language_model.model.layers.16.self_attn.o_proj",
637
+ "language_model.model.layers.16.mlp.gate_proj",
638
+ "language_model.model.layers.16.mlp.up_proj",
639
+ "language_model.model.layers.16.mlp.down_proj",
640
+ "language_model.model.layers.17.self_attn.q_proj",
641
+ "language_model.model.layers.17.self_attn.k_proj",
642
+ "language_model.model.layers.17.self_attn.v_proj",
643
+ "language_model.model.layers.17.self_attn.o_proj",
644
+ "language_model.model.layers.17.mlp.gate_proj",
645
+ "language_model.model.layers.17.mlp.up_proj",
646
+ "language_model.model.layers.17.mlp.down_proj",
647
+ "language_model.model.layers.18.cross_attn.q_proj",
648
+ "language_model.model.layers.18.cross_attn.k_proj",
649
+ "language_model.model.layers.18.cross_attn.v_proj",
650
+ "language_model.model.layers.18.cross_attn.o_proj",
651
+ "language_model.model.layers.18.mlp.gate_proj",
652
+ "language_model.model.layers.18.mlp.up_proj",
653
+ "language_model.model.layers.18.mlp.down_proj",
654
+ "language_model.model.layers.19.self_attn.q_proj",
655
+ "language_model.model.layers.19.self_attn.k_proj",
656
+ "language_model.model.layers.19.self_attn.v_proj",
657
+ "language_model.model.layers.19.self_attn.o_proj",
658
+ "language_model.model.layers.19.mlp.gate_proj",
659
+ "language_model.model.layers.19.mlp.up_proj",
660
+ "language_model.model.layers.19.mlp.down_proj",
661
+ "language_model.model.layers.20.self_attn.q_proj",
662
+ "language_model.model.layers.20.self_attn.k_proj",
663
+ "language_model.model.layers.20.self_attn.v_proj",
664
+ "language_model.model.layers.20.self_attn.o_proj",
665
+ "language_model.model.layers.20.mlp.gate_proj",
666
+ "language_model.model.layers.20.mlp.up_proj",
667
+ "language_model.model.layers.20.mlp.down_proj",
668
+ "language_model.model.layers.21.self_attn.q_proj",
669
+ "language_model.model.layers.21.self_attn.k_proj",
670
+ "language_model.model.layers.21.self_attn.v_proj",
671
+ "language_model.model.layers.21.self_attn.o_proj",
672
+ "language_model.model.layers.21.mlp.gate_proj",
673
+ "language_model.model.layers.21.mlp.up_proj",
674
+ "language_model.model.layers.21.mlp.down_proj",
675
+ "language_model.model.layers.22.self_attn.q_proj",
676
+ "language_model.model.layers.22.self_attn.k_proj",
677
+ "language_model.model.layers.22.self_attn.v_proj",
678
+ "language_model.model.layers.22.self_attn.o_proj",
679
+ "language_model.model.layers.22.mlp.gate_proj",
680
+ "language_model.model.layers.22.mlp.up_proj",
681
+ "language_model.model.layers.22.mlp.down_proj",
682
+ "language_model.model.layers.23.cross_attn.q_proj",
683
+ "language_model.model.layers.23.cross_attn.k_proj",
684
+ "language_model.model.layers.23.cross_attn.v_proj",
685
+ "language_model.model.layers.23.cross_attn.o_proj",
686
+ "language_model.model.layers.23.mlp.gate_proj",
687
+ "language_model.model.layers.23.mlp.up_proj",
688
+ "language_model.model.layers.23.mlp.down_proj",
689
+ "language_model.model.layers.24.self_attn.q_proj",
690
+ "language_model.model.layers.24.self_attn.k_proj",
691
+ "language_model.model.layers.24.self_attn.v_proj",
692
+ "language_model.model.layers.24.self_attn.o_proj",
693
+ "language_model.model.layers.24.mlp.gate_proj",
694
+ "language_model.model.layers.24.mlp.up_proj",
695
+ "language_model.model.layers.24.mlp.down_proj",
696
+ "language_model.model.layers.25.self_attn.q_proj",
697
+ "language_model.model.layers.25.self_attn.k_proj",
698
+ "language_model.model.layers.25.self_attn.v_proj",
699
+ "language_model.model.layers.25.self_attn.o_proj",
700
+ "language_model.model.layers.25.mlp.gate_proj",
701
+ "language_model.model.layers.25.mlp.up_proj",
702
+ "language_model.model.layers.25.mlp.down_proj",
703
+ "language_model.model.layers.26.self_attn.q_proj",
704
+ "language_model.model.layers.26.self_attn.k_proj",
705
+ "language_model.model.layers.26.self_attn.v_proj",
706
+ "language_model.model.layers.26.self_attn.o_proj",
707
+ "language_model.model.layers.26.mlp.gate_proj",
708
+ "language_model.model.layers.26.mlp.up_proj",
709
+ "language_model.model.layers.26.mlp.down_proj",
710
+ "language_model.model.layers.27.self_attn.q_proj",
711
+ "language_model.model.layers.27.self_attn.k_proj",
712
+ "language_model.model.layers.27.self_attn.v_proj",
713
+ "language_model.model.layers.27.self_attn.o_proj",
714
+ "language_model.model.layers.27.mlp.gate_proj",
715
+ "language_model.model.layers.27.mlp.up_proj",
716
+ "language_model.model.layers.27.mlp.down_proj",
717
+ "language_model.model.layers.28.cross_attn.q_proj",
718
+ "language_model.model.layers.28.cross_attn.k_proj",
719
+ "language_model.model.layers.28.cross_attn.v_proj",
720
+ "language_model.model.layers.28.cross_attn.o_proj",
721
+ "language_model.model.layers.28.mlp.gate_proj",
722
+ "language_model.model.layers.28.mlp.up_proj",
723
+ "language_model.model.layers.28.mlp.down_proj",
724
+ "language_model.model.layers.29.self_attn.q_proj",
725
+ "language_model.model.layers.29.self_attn.k_proj",
726
+ "language_model.model.layers.29.self_attn.v_proj",
727
+ "language_model.model.layers.29.self_attn.o_proj",
728
+ "language_model.model.layers.29.mlp.gate_proj",
729
+ "language_model.model.layers.29.mlp.up_proj",
730
+ "language_model.model.layers.29.mlp.down_proj",
731
+ "language_model.model.layers.30.self_attn.q_proj",
732
+ "language_model.model.layers.30.self_attn.k_proj",
733
+ "language_model.model.layers.30.self_attn.v_proj",
734
+ "language_model.model.layers.30.self_attn.o_proj",
735
+ "language_model.model.layers.30.mlp.gate_proj",
736
+ "language_model.model.layers.30.mlp.up_proj",
737
+ "language_model.model.layers.30.mlp.down_proj",
738
+ "language_model.model.layers.31.self_attn.q_proj",
739
+ "language_model.model.layers.31.self_attn.k_proj",
740
+ "language_model.model.layers.31.self_attn.v_proj",
741
+ "language_model.model.layers.31.self_attn.o_proj",
742
+ "language_model.model.layers.31.mlp.gate_proj",
743
+ "language_model.model.layers.31.mlp.up_proj",
744
+ "language_model.model.layers.31.mlp.down_proj",
745
+ "language_model.model.layers.32.self_attn.q_proj",
746
+ "language_model.model.layers.32.self_attn.k_proj",
747
+ "language_model.model.layers.32.self_attn.v_proj",
748
+ "language_model.model.layers.32.self_attn.o_proj",
749
+ "language_model.model.layers.32.mlp.gate_proj",
750
+ "language_model.model.layers.32.mlp.up_proj",
751
+ "language_model.model.layers.32.mlp.down_proj",
752
+ "language_model.model.layers.33.cross_attn.q_proj",
753
+ "language_model.model.layers.33.cross_attn.k_proj",
754
+ "language_model.model.layers.33.cross_attn.v_proj",
755
+ "language_model.model.layers.33.cross_attn.o_proj",
756
+ "language_model.model.layers.33.mlp.gate_proj",
757
+ "language_model.model.layers.33.mlp.up_proj",
758
+ "language_model.model.layers.33.mlp.down_proj",
759
+ "language_model.model.layers.34.self_attn.q_proj",
760
+ "language_model.model.layers.34.self_attn.k_proj",
761
+ "language_model.model.layers.34.self_attn.v_proj",
762
+ "language_model.model.layers.34.self_attn.o_proj",
763
+ "language_model.model.layers.34.mlp.gate_proj",
764
+ "language_model.model.layers.34.mlp.up_proj",
765
+ "language_model.model.layers.34.mlp.down_proj",
766
+ "language_model.model.layers.35.self_attn.q_proj",
767
+ "language_model.model.layers.35.self_attn.k_proj",
768
+ "language_model.model.layers.35.self_attn.v_proj",
769
+ "language_model.model.layers.35.self_attn.o_proj",
770
+ "language_model.model.layers.35.mlp.gate_proj",
771
+ "language_model.model.layers.35.mlp.up_proj",
772
+ "language_model.model.layers.35.mlp.down_proj",
773
+ "language_model.model.layers.36.self_attn.q_proj",
774
+ "language_model.model.layers.36.self_attn.k_proj",
775
+ "language_model.model.layers.36.self_attn.v_proj",
776
+ "language_model.model.layers.36.self_attn.o_proj",
777
+ "language_model.model.layers.36.mlp.gate_proj",
778
+ "language_model.model.layers.36.mlp.up_proj",
779
+ "language_model.model.layers.36.mlp.down_proj",
780
+ "language_model.model.layers.37.self_attn.q_proj",
781
+ "language_model.model.layers.37.self_attn.k_proj",
782
+ "language_model.model.layers.37.self_attn.v_proj",
783
+ "language_model.model.layers.37.self_attn.o_proj",
784
+ "language_model.model.layers.37.mlp.gate_proj",
785
+ "language_model.model.layers.37.mlp.up_proj",
786
+ "language_model.model.layers.37.mlp.down_proj",
787
+ "language_model.model.layers.38.cross_attn.q_proj",
788
+ "language_model.model.layers.38.cross_attn.k_proj",
789
+ "language_model.model.layers.38.cross_attn.v_proj",
790
+ "language_model.model.layers.38.cross_attn.o_proj",
791
+ "language_model.model.layers.38.mlp.gate_proj",
792
+ "language_model.model.layers.38.mlp.up_proj",
793
+ "language_model.model.layers.38.mlp.down_proj",
794
+ "language_model.model.layers.39.self_attn.q_proj",
795
+ "language_model.model.layers.39.self_attn.k_proj",
796
+ "language_model.model.layers.39.self_attn.v_proj",
797
+ "language_model.model.layers.39.self_attn.o_proj",
798
+ "language_model.model.layers.39.mlp.gate_proj",
799
+ "language_model.model.layers.39.mlp.up_proj",
800
+ "language_model.model.layers.39.mlp.down_proj",
801
+ "language_model.lm_head",
802
+ "multi_modal_projector"
803
+ ],
804
+ "registry_requires_subclass": false,
805
+ "sparsity_structure": "unstructured",
806
+ "targets": [
807
+ "language_model.model.layers.0.self_attn.q_proj",
808
+ "language_model.model.layers.0.self_attn.k_proj"
809
+ ]
810
+ }
811
+ },
812
+ "text_config": {
813
+ "eos_token_id": [
814
+ 128001,
815
+ 128008,
816
+ 128009
817
+ ],
818
+ "model_type": "mllama_text_model",
819
+ "rope_scaling": {
820
+ "factor": 8.0,
821
+ "high_freq_factor": 4.0,
822
+ "low_freq_factor": 1.0,
823
+ "original_max_position_embeddings": 8192,
824
+ "rope_type": "llama3"
825
+ },
826
+ "torch_dtype": "bfloat16"
827
+ },
828
+ "torch_dtype": "float16",
829
+ "transformers_version": "4.47.1",
830
+ "vision_config": {
831
+ "image_size": 560,
832
+ "model_type": "mllama_vision_model",
833
+ "torch_dtype": "bfloat16"
834
+ }
835
+ }
generation_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token_id": 128000,
3
+ "do_sample": true,
4
+ "eos_token_id": [
5
+ 128001,
6
+ 128008,
7
+ 128009
8
+ ],
9
+ "pad_token_id": 128004,
10
+ "temperature": 0.6,
11
+ "top_p": 0.9,
12
+ "transformers_version": "4.47.1"
13
+ }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:05857ba7521250d366a2dddc9fd56553ea82c088d6db6a502cd221d9840f6d7e
3
+ size 4998152470
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:db00a3eb235415ed58755d915f3797be03f48ad269f44d3c0f4d6c26d7b94582
3
+ size 3392570984
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff