Hugging Face's logo

unsloth
/
Llama-3.2-11B-Vision-Instruct-bnb-4bit

Image-to-Text
Transformers
Safetensors
PyTorch
English
mllama
llama-3
llama
meta
facebook
unsloth
multimodal
vision
text-generation-inference
4-bit precision
bitsandbytes
Model card Files
xet
Community
9
danielhanchen commited on
Commit
bde7cbc
·
verified ·
1 Parent(s): b9c6d51

Upload MllamaForConditionalGeneration

Browse files
Files changed (5) hide show
  1. config.json +218 -40
  2. generation_config.json +1 -1
  3. model-00001-of-00002.safetensors +3 -0
  4. model-00002-of-00002.safetensors +3 -0
  5. model.safetensors.index.json +0 -0
config.json CHANGED
@@ -1,34 +1,10 @@
1
  {
 
2
  "architectures": [
3
- "MllamaForCausalLM"
4
  ],
5
- "bos_token_id": 128000,
6
- "cross_attention_layers": [
7
- 3,
8
- 8,
9
- 13,
10
- 18,
11
- 23,
12
- 28,
13
- 33,
14
- 38
15
- ],
16
- "dropout": 0,
17
- "eos_token_id": [
18
- 128001,
19
- 128008,
20
- 128009
21
- ],
22
- "hidden_act": "silu",
23
- "hidden_size": 4096,
24
- "initializer_range": 0.02,
25
- "intermediate_size": 14336,
26
- "max_position_embeddings": 131072,
27
- "model_type": "mllama_text_model",
28
- "num_attention_heads": 32,
29
- "num_hidden_layers": 40,
30
- "num_key_value_heads": 8,
31
- "pad_token_id": 128004,
32
  "quantization_config": {
33
  "_load_in_4bit": true,
34
  "_load_in_8bit": false,
@@ -44,18 +20,220 @@
44
  "load_in_8bit": false,
45
  "quant_method": "bitsandbytes"
46
  },
47
- "rms_norm_eps": 1e-05,
48
- "rope_scaling": {
49
- "factor": 8.0,
50
- "high_freq_factor": 4.0,
51
- "low_freq_factor": 1.0,
52
- "original_max_position_embeddings": 8192,
53
- "rope_type": "llama3"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
54
  },
55
- "rope_theta": 500000.0,
56
- "tie_word_embeddings": false,
57
  "torch_dtype": "bfloat16",
58
- "transformers_version": "4.46.0.dev0",
59
- "use_cache": true,
60
- "vocab_size": 128256
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61
  }
 
1
  {
2
+ "_name_or_path": "meta-llama/Llama-3.2-11B-Vision-Instruct",
3
  "architectures": [
4
+ "MllamaForConditionalGeneration"
5
  ],
6
+ "image_token_index": 128256,
7
+ "model_type": "mllama",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
  "quantization_config": {
9
  "_load_in_4bit": true,
10
  "_load_in_8bit": false,
 
20
  "load_in_8bit": false,
21
  "quant_method": "bitsandbytes"
22
  },
23
+ "text_config": {
24
+ "_name_or_path": "",
25
+ "add_cross_attention": false,
26
+ "architectures": null,
27
+ "bad_words_ids": null,
28
+ "begin_suppress_tokens": null,
29
+ "bos_token_id": 128000,
30
+ "chunk_size_feed_forward": 0,
31
+ "cross_attention_hidden_size": null,
32
+ "cross_attention_layers": [
33
+ 3,
34
+ 8,
35
+ 13,
36
+ 18,
37
+ 23,
38
+ 28,
39
+ 33,
40
+ 38
41
+ ],
42
+ "decoder_start_token_id": null,
43
+ "diversity_penalty": 0.0,
44
+ "do_sample": false,
45
+ "dropout": 0,
46
+ "early_stopping": false,
47
+ "encoder_no_repeat_ngram_size": 0,
48
+ "eos_token_id": [
49
+ 128001,
50
+ 128008,
51
+ 128009
52
+ ],
53
+ "exponential_decay_length_penalty": null,
54
+ "finetuning_task": null,
55
+ "forced_bos_token_id": null,
56
+ "forced_eos_token_id": null,
57
+ "hidden_act": "silu",
58
+ "hidden_size": 4096,
59
+ "id2label": {
60
+ "0": "LABEL_0",
61
+ "1": "LABEL_1"
62
+ },
63
+ "initializer_range": 0.02,
64
+ "intermediate_size": 14336,
65
+ "is_decoder": false,
66
+ "is_encoder_decoder": false,
67
+ "label2id": {
68
+ "LABEL_0": 0,
69
+ "LABEL_1": 1
70
+ },
71
+ "length_penalty": 1.0,
72
+ "max_length": 20,
73
+ "max_position_embeddings": 131072,
74
+ "min_length": 0,
75
+ "model_type": "mllama_text_model",
76
+ "no_repeat_ngram_size": 0,
77
+ "num_attention_heads": 32,
78
+ "num_beam_groups": 1,
79
+ "num_beams": 1,
80
+ "num_hidden_layers": 40,
81
+ "num_key_value_heads": 8,
82
+ "num_return_sequences": 1,
83
+ "output_attentions": false,
84
+ "output_hidden_states": false,
85
+ "output_scores": false,
86
+ "pad_token_id": 128004,
87
+ "prefix": null,
88
+ "problem_type": null,
89
+ "pruned_heads": {},
90
+ "remove_invalid_values": false,
91
+ "repetition_penalty": 1.0,
92
+ "return_dict": true,
93
+ "return_dict_in_generate": false,
94
+ "rms_norm_eps": 1e-05,
95
+ "rope_scaling": {
96
+ "factor": 8.0,
97
+ "high_freq_factor": 4.0,
98
+ "low_freq_factor": 1.0,
99
+ "original_max_position_embeddings": 8192,
100
+ "rope_type": "llama3"
101
+ },
102
+ "rope_theta": 500000.0,
103
+ "sep_token_id": null,
104
+ "suppress_tokens": null,
105
+ "task_specific_params": null,
106
+ "temperature": 1.0,
107
+ "tf_legacy_loss": false,
108
+ "tie_encoder_decoder": false,
109
+ "tie_word_embeddings": false,
110
+ "tokenizer_class": null,
111
+ "top_k": 50,
112
+ "top_p": 1.0,
113
+ "torch_dtype": "bfloat16",
114
+ "torchscript": false,
115
+ "typical_p": 1.0,
116
+ "use_bfloat16": false,
117
+ "use_cache": true,
118
+ "vocab_size": 128256
119
  },
 
 
120
  "torch_dtype": "bfloat16",
121
+ "transformers_version": "4.45.0",
122
+ "vision_config": {
123
+ "_name_or_path": "",
124
+ "add_cross_attention": false,
125
+ "architectures": null,
126
+ "attention_heads": 16,
127
+ "bad_words_ids": null,
128
+ "begin_suppress_tokens": null,
129
+ "bos_token_id": null,
130
+ "chunk_size_feed_forward": 0,
131
+ "cross_attention_hidden_size": null,
132
+ "decoder_start_token_id": null,
133
+ "diversity_penalty": 0.0,
134
+ "do_sample": false,
135
+ "early_stopping": false,
136
+ "encoder_no_repeat_ngram_size": 0,
137
+ "eos_token_id": null,
138
+ "exponential_decay_length_penalty": null,
139
+ "finetuning_task": null,
140
+ "forced_bos_token_id": null,
141
+ "forced_eos_token_id": null,
142
+ "hidden_act": "gelu",
143
+ "hidden_size": 1280,
144
+ "id2label": {
145
+ "0": "LABEL_0",
146
+ "1": "LABEL_1"
147
+ },
148
+ "image_size": 560,
149
+ "initializer_range": 0.02,
150
+ "intermediate_layers_indices": [
151
+ 3,
152
+ 7,
153
+ 15,
154
+ 23,
155
+ 30
156
+ ],
157
+ "intermediate_size": 5120,
158
+ "is_decoder": false,
159
+ "is_encoder_decoder": false,
160
+ "label2id": {
161
+ "LABEL_0": 0,
162
+ "LABEL_1": 1
163
+ },
164
+ "length_penalty": 1.0,
165
+ "max_length": 20,
166
+ "max_num_tiles": 4,
167
+ "min_length": 0,
168
+ "model_type": "mllama_vision_model",
169
+ "no_repeat_ngram_size": 0,
170
+ "norm_eps": 1e-05,
171
+ "num_beam_groups": 1,
172
+ "num_beams": 1,
173
+ "num_channels": 3,
174
+ "num_global_layers": 8,
175
+ "num_hidden_layers": 32,
176
+ "num_return_sequences": 1,
177
+ "output_attentions": false,
178
+ "output_hidden_states": false,
179
+ "output_scores": false,
180
+ "pad_token_id": null,
181
+ "patch_size": 14,
182
+ "prefix": null,
183
+ "problem_type": null,
184
+ "pruned_heads": {},
185
+ "remove_invalid_values": false,
186
+ "repetition_penalty": 1.0,
187
+ "return_dict": true,
188
+ "return_dict_in_generate": false,
189
+ "sep_token_id": null,
190
+ "supported_aspect_ratios": [
191
+ [
192
+ 1,
193
+ 1
194
+ ],
195
+ [
196
+ 1,
197
+ 2
198
+ ],
199
+ [
200
+ 1,
201
+ 3
202
+ ],
203
+ [
204
+ 1,
205
+ 4
206
+ ],
207
+ [
208
+ 2,
209
+ 1
210
+ ],
211
+ [
212
+ 2,
213
+ 2
214
+ ],
215
+ [
216
+ 3,
217
+ 1
218
+ ],
219
+ [
220
+ 4,
221
+ 1
222
+ ]
223
+ ],
224
+ "suppress_tokens": null,
225
+ "task_specific_params": null,
226
+ "temperature": 1.0,
227
+ "tf_legacy_loss": false,
228
+ "tie_encoder_decoder": false,
229
+ "tie_word_embeddings": true,
230
+ "tokenizer_class": null,
231
+ "top_k": 50,
232
+ "top_p": 1.0,
233
+ "torch_dtype": "bfloat16",
234
+ "torchscript": false,
235
+ "typical_p": 1.0,
236
+ "use_bfloat16": false,
237
+ "vision_output_dim": 7680
238
+ }
239
  }
generation_config.json CHANGED
@@ -9,5 +9,5 @@
9
  "pad_token_id": 128004,
10
  "temperature": 0.6,
11
  "top_p": 0.9,
12
- "transformers_version": "4.46.0.dev0"
13
  }
 
9
  "pad_token_id": 128004,
10
  "temperature": 0.6,
11
  "top_p": 0.9,
12
+ "transformers_version": "4.45.0"
13
  }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:119b9d390428afb0e5cc62d835c5b0859e6b6f94dab5c387d014e485625b5adf
3
+ size 4998395628
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fde970a4c7fc1d8c2b09c0c7fed189cafa75d59713c8ddf0049ac81294600eb
3
+ size 2181540357
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff