Hugging Face's logo

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

Transformers
Safetensors
English
mllama
pretraining
llama-3
llama
meta
facebook
unsloth
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
danielhanchen commited on
Commit
3086d0a
•
1 Parent(s): 4b1e144

Upload MllamaForConditionalGeneration

Browse files
Files changed (5) hide show
  1. config.json +214 -36
  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
@@ -1,30 +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": 128001,
18
- "hidden_act": "silu",
19
- "hidden_size": 4096,
20
- "initializer_range": 0.02,
21
- "intermediate_size": 14336,
22
- "max_position_embeddings": 131072,
23
- "model_type": "mllama_text_model",
24
- "num_attention_heads": 32,
25
- "num_hidden_layers": 40,
26
- "num_key_value_heads": 8,
27
- "pad_token_id": 128004,
28
  "quantization_config": {
29
  "_load_in_4bit": true,
30
  "_load_in_8bit": false,
@@ -40,18 +20,216 @@
40
  "load_in_8bit": false,
41
  "quant_method": "bitsandbytes"
42
  },
43
- "rms_norm_eps": 1e-05,
44
- "rope_scaling": {
45
- "factor": 8.0,
46
- "high_freq_factor": 4.0,
47
- "low_freq_factor": 1.0,
48
- "original_max_position_embeddings": 8192,
49
- "rope_type": "llama3"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
50
  },
51
- "rope_theta": 500000.0,
52
- "tie_word_embeddings": false,
53
  "torch_dtype": "bfloat16",
54
- "transformers_version": "4.46.0.dev0",
55
- "use_cache": true,
56
- "vocab_size": 128256
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
57
  }
 
1
  {
2
+ "_name_or_path": "meta-llama/Llama-3.2-11B-Vision",
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": 128001,
49
+ "exponential_decay_length_penalty": null,
50
+ "finetuning_task": null,
51
+ "forced_bos_token_id": null,
52
+ "forced_eos_token_id": null,
53
+ "hidden_act": "silu",
54
+ "hidden_size": 4096,
55
+ "id2label": {
56
+ "0": "LABEL_0",
57
+ "1": "LABEL_1"
58
+ },
59
+ "initializer_range": 0.02,
60
+ "intermediate_size": 14336,
61
+ "is_decoder": false,
62
+ "is_encoder_decoder": false,
63
+ "label2id": {
64
+ "LABEL_0": 0,
65
+ "LABEL_1": 1
66
+ },
67
+ "length_penalty": 1.0,
68
+ "max_length": 20,
69
+ "max_position_embeddings": 131072,
70
+ "min_length": 0,
71
+ "model_type": "mllama_text_model",
72
+ "no_repeat_ngram_size": 0,
73
+ "num_attention_heads": 32,
74
+ "num_beam_groups": 1,
75
+ "num_beams": 1,
76
+ "num_hidden_layers": 40,
77
+ "num_key_value_heads": 8,
78
+ "num_return_sequences": 1,
79
+ "output_attentions": false,
80
+ "output_hidden_states": false,
81
+ "output_scores": false,
82
+ "pad_token_id": 128004,
83
+ "prefix": null,
84
+ "problem_type": null,
85
+ "pruned_heads": {},
86
+ "remove_invalid_values": false,
87
+ "repetition_penalty": 1.0,
88
+ "return_dict": true,
89
+ "return_dict_in_generate": false,
90
+ "rms_norm_eps": 1e-05,
91
+ "rope_scaling": {
92
+ "factor": 8.0,
93
+ "high_freq_factor": 4.0,
94
+ "low_freq_factor": 1.0,
95
+ "original_max_position_embeddings": 8192,
96
+ "rope_type": "llama3"
97
+ },
98
+ "rope_theta": 500000.0,
99
+ "sep_token_id": null,
100
+ "suppress_tokens": null,
101
+ "task_specific_params": null,
102
+ "temperature": 1.0,
103
+ "tf_legacy_loss": false,
104
+ "tie_encoder_decoder": false,
105
+ "tie_word_embeddings": false,
106
+ "tokenizer_class": null,
107
+ "top_k": 50,
108
+ "top_p": 1.0,
109
+ "torch_dtype": "bfloat16",
110
+ "torchscript": false,
111
+ "typical_p": 1.0,
112
+ "use_bfloat16": false,
113
+ "use_cache": true,
114
+ "vocab_size": 128256
115
  },
 
 
116
  "torch_dtype": "bfloat16",
117
+ "transformers_version": "4.45.0",
118
+ "vision_config": {
119
+ "_name_or_path": "",
120
+ "add_cross_attention": false,
121
+ "architectures": null,
122
+ "attention_heads": 16,
123
+ "bad_words_ids": null,
124
+ "begin_suppress_tokens": null,
125
+ "bos_token_id": null,
126
+ "chunk_size_feed_forward": 0,
127
+ "cross_attention_hidden_size": null,
128
+ "decoder_start_token_id": null,
129
+ "diversity_penalty": 0.0,
130
+ "do_sample": false,
131
+ "early_stopping": false,
132
+ "encoder_no_repeat_ngram_size": 0,
133
+ "eos_token_id": null,
134
+ "exponential_decay_length_penalty": null,
135
+ "finetuning_task": null,
136
+ "forced_bos_token_id": null,
137
+ "forced_eos_token_id": null,
138
+ "hidden_act": "gelu",
139
+ "hidden_size": 1280,
140
+ "id2label": {
141
+ "0": "LABEL_0",
142
+ "1": "LABEL_1"
143
+ },
144
+ "image_size": 448,
145
+ "initializer_range": 0.02,
146
+ "intermediate_layers_indices": [
147
+ 3,
148
+ 7,
149
+ 15,
150
+ 23,
151
+ 30
152
+ ],
153
+ "intermediate_size": 5120,
154
+ "is_decoder": false,
155
+ "is_encoder_decoder": false,
156
+ "label2id": {
157
+ "LABEL_0": 0,
158
+ "LABEL_1": 1
159
+ },
160
+ "length_penalty": 1.0,
161
+ "max_length": 20,
162
+ "max_num_tiles": 4,
163
+ "min_length": 0,
164
+ "model_type": "mllama_vision_model",
165
+ "no_repeat_ngram_size": 0,
166
+ "norm_eps": 1e-05,
167
+ "num_beam_groups": 1,
168
+ "num_beams": 1,
169
+ "num_channels": 3,
170
+ "num_global_layers": 8,
171
+ "num_hidden_layers": 32,
172
+ "num_return_sequences": 1,
173
+ "output_attentions": false,
174
+ "output_hidden_states": false,
175
+ "output_scores": false,
176
+ "pad_token_id": null,
177
+ "patch_size": 14,
178
+ "prefix": null,
179
+ "problem_type": null,
180
+ "pruned_heads": {},
181
+ "remove_invalid_values": false,
182
+ "repetition_penalty": 1.0,
183
+ "return_dict": true,
184
+ "return_dict_in_generate": false,
185
+ "sep_token_id": null,
186
+ "supported_aspect_ratios": [
187
+ [
188
+ 1,
189
+ 1
190
+ ],
191
+ [
192
+ 1,
193
+ 2
194
+ ],
195
+ [
196
+ 1,
197
+ 3
198
+ ],
199
+ [
200
+ 1,
201
+ 4
202
+ ],
203
+ [
204
+ 2,
205
+ 1
206
+ ],
207
+ [
208
+ 2,
209
+ 2
210
+ ],
211
+ [
212
+ 3,
213
+ 1
214
+ ],
215
+ [
216
+ 4,
217
+ 1
218
+ ]
219
+ ],
220
+ "suppress_tokens": null,
221
+ "task_specific_params": null,
222
+ "temperature": 1.0,
223
+ "tf_legacy_loss": false,
224
+ "tie_encoder_decoder": false,
225
+ "tie_word_embeddings": true,
226
+ "tokenizer_class": null,
227
+ "top_k": 50,
228
+ "top_p": 1.0,
229
+ "torch_dtype": "bfloat16",
230
+ "torchscript": false,
231
+ "typical_p": 1.0,
232
+ "use_bfloat16": false,
233
+ "vision_output_dim": 7680
234
+ }
235
  }
generation_config.json CHANGED
@@ -3,5 +3,5 @@
3
  "bos_token_id": 128000,
4
  "eos_token_id": 128001,
5
  "pad_token_id": 128004,
6
- "transformers_version": "4.46.0.dev0"
7
  }
 
3
  "bos_token_id": 128000,
4
  "eos_token_id": 128001,
5
  "pad_token_id": 128004,
6
+ "transformers_version": "4.45.0"
7
  }
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e69f8fcdfd5863e6bfa5b57101d06d00ff078b11e372ca71cbb4bde23882a42
3
+ size 4984953659
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48975aae8e6a2efcb65673e54d170db7958a59aedcecb4b03bb6b0e5e33342f8
3
+ size 2140423619
model.safetensors.index.json CHANGED
The diff for this file is too large to render. See raw diff