Visual Question Answering
Transformers
Safetensors
Chinese
English
QH_360VL
text-generation
custom_code
qingshan777 commited on
Commit
bdf4f88
1 Parent(s): 33a7b4e

Delete proj_config.json

Browse files
Files changed (1) hide show
  1. proj_config.json +0 -337
proj_config.json DELETED
@@ -1,337 +0,0 @@
1
- {
2
- "_commit_hash": null,
3
- "_name_or_path": "",
4
- "architectures": [
5
- "HoneybeeForConditionalGeneration"
6
- ],
7
- "hidden_act": "silu",
8
- "hidden_size": 5120,
9
- "initializer_factor": 1.0,
10
- "initializer_range": 0.02,
11
- "intermediate_size": 13824,
12
- "keys_to_ignore_at_inference": [
13
- "past_key_values"
14
- ],
15
- "lm_config": {
16
- "_name_or_path": "",
17
- "add_cross_attention": false,
18
- "architectures": null,
19
- "bad_words_ids": null,
20
- "begin_suppress_tokens": null,
21
- "bos_token_id": null,
22
- "chunk_size_feed_forward": 0,
23
- "cross_attention_hidden_size": null,
24
- "decoder_start_token_id": null,
25
- "delta_model_name_or_path": null,
26
- "diversity_penalty": 0.0,
27
- "do_sample": false,
28
- "early_stopping": false,
29
- "encoder_no_repeat_ngram_size": 0,
30
- "eos_token_id": null,
31
- "exponential_decay_length_penalty": null,
32
- "finetuning_task": null,
33
- "forced_bos_token_id": null,
34
- "forced_eos_token_id": null,
35
- "id2label": {
36
- "0": "LABEL_0",
37
- "1": "LABEL_1"
38
- },
39
- "is_decoder": false,
40
- "is_encoder_decoder": false,
41
- "label2id": {
42
- "LABEL_0": 0,
43
- "LABEL_1": 1
44
- },
45
- "length_penalty": 1.0,
46
- "max_length": 20,
47
- "min_length": 0,
48
- "model_type": "mllm_lm",
49
- "no_repeat_ngram_size": 0,
50
- "num_beam_groups": 1,
51
- "num_beams": 1,
52
- "num_return_sequences": 1,
53
- "output_attentions": false,
54
- "output_hidden_states": false,
55
- "output_scores": false,
56
- "pad_token_id": null,
57
- "prefix": null,
58
- "pretrained_lm_name_or_path": "/hbox2dir/vicuna-13b-v1.5",
59
- "pretrained_tokenizer_name_or_path": "/hbox2dir/vicuna-13b-v1.5",
60
- "problem_type": null,
61
- "pruned_heads": {},
62
- "remove_invalid_values": false,
63
- "repetition_penalty": 1.0,
64
- "return_dict": true,
65
- "return_dict_in_generate": false,
66
- "sep_token_id": null,
67
- "suppress_tokens": null,
68
- "task_specific_params": null,
69
- "temperature": 1.0,
70
- "tf_legacy_loss": false,
71
- "tie_encoder_decoder": false,
72
- "tie_word_embeddings": true,
73
- "tokenizer_class": null,
74
- "top_k": 50,
75
- "top_p": 1.0,
76
- "torch_dtype": null,
77
- "torchscript": false,
78
- "transformers_version": "4.30.2",
79
- "typical_p": 1.0,
80
- "use_bfloat16": false
81
- },
82
- "max_position_embeddings": 4096,
83
- "model_type": "mllm",
84
- "num_attention_heads": 40,
85
- "num_hidden_layers": 40,
86
- "num_key_value_heads": 40,
87
- "num_query_tokens": 576,
88
- "pretraining_tp": 1,
89
- "rms_norm_eps": 1e-05,
90
- "rope_scaling": null,
91
- "text_config": {
92
- "_name_or_path": "",
93
- "add_cross_attention": false,
94
- "architectures": [
95
- "LlamaForCausalLM"
96
- ],
97
- "bad_words_ids": null,
98
- "begin_suppress_tokens": null,
99
- "bos_token_id": 1,
100
- "chunk_size_feed_forward": 0,
101
- "cross_attention_hidden_size": null,
102
- "decoder_start_token_id": null,
103
- "diversity_penalty": 0.0,
104
- "do_sample": false,
105
- "early_stopping": false,
106
- "encoder_no_repeat_ngram_size": 0,
107
- "eos_token_id": 2,
108
- "exponential_decay_length_penalty": null,
109
- "finetuning_task": null,
110
- "forced_bos_token_id": null,
111
- "forced_eos_token_id": null,
112
- "hidden_act": "silu",
113
- "hidden_size": 5120,
114
- "id2label": {
115
- "0": "LABEL_0",
116
- "1": "LABEL_1"
117
- },
118
- "initializer_range": 0.02,
119
- "intermediate_size": 13824,
120
- "is_decoder": false,
121
- "is_encoder_decoder": false,
122
- "label2id": {
123
- "LABEL_0": 0,
124
- "LABEL_1": 1
125
- },
126
- "length_penalty": 1.0,
127
- "max_length": 4096,
128
- "max_position_embeddings": 4096,
129
- "min_length": 0,
130
- "model_type": "llama",
131
- "no_repeat_ngram_size": 0,
132
- "num_attention_heads": 40,
133
- "num_beam_groups": 1,
134
- "num_beams": 1,
135
- "num_hidden_layers": 40,
136
- "num_key_value_heads": 40,
137
- "num_return_sequences": 1,
138
- "output_attentions": false,
139
- "output_hidden_states": false,
140
- "output_scores": false,
141
- "pad_token_id": 0,
142
- "prefix": null,
143
- "pretraining_tp": 1,
144
- "problem_type": null,
145
- "pruned_heads": {},
146
- "remove_invalid_values": false,
147
- "repetition_penalty": 1.0,
148
- "return_dict": true,
149
- "return_dict_in_generate": false,
150
- "rms_norm_eps": 1e-05,
151
- "rope_scaling": null,
152
- "sep_token_id": null,
153
- "suppress_tokens": null,
154
- "task_specific_params": null,
155
- "temperature": 1.0,
156
- "tf_legacy_loss": false,
157
- "tie_encoder_decoder": false,
158
- "tie_word_embeddings": false,
159
- "tokenizer_class": null,
160
- "top_k": 50,
161
- "top_p": 1.0,
162
- "torch_dtype": "float16",
163
- "torchscript": false,
164
- "transformers_version": "4.30.2",
165
- "typical_p": 1.0,
166
- "use_bfloat16": false,
167
- "use_cache": true,
168
- "vocab_size": 32000
169
- },
170
- "tie_word_embeddings": false,
171
- "torch_dtype": "bfloat16",
172
- "transformers_version": null,
173
- "use_cache": true,
174
- "use_decoder_only_language_model": true,
175
- "vision_config": {
176
- "_name_or_path": "",
177
- "add_cross_attention": false,
178
- "architectures": null,
179
- "attention_dropout": 0.0,
180
- "bad_words_ids": null,
181
- "begin_suppress_tokens": null,
182
- "bos_token_id": null,
183
- "chunk_size_feed_forward": 0,
184
- "cross_attention_hidden_size": null,
185
- "decoder_start_token_id": null,
186
- "diversity_penalty": 0.0,
187
- "do_sample": false,
188
- "dropout": 0.0,
189
- "early_stopping": false,
190
- "encoder_no_repeat_ngram_size": 0,
191
- "encoder_type": "openai.clip",
192
- "eos_token_id": null,
193
- "exponential_decay_length_penalty": null,
194
- "finetuning_task": null,
195
- "forced_bos_token_id": null,
196
- "forced_eos_token_id": null,
197
- "hidden_act": "quick_gelu",
198
- "hidden_size": 1024,
199
- "id2label": {
200
- "0": "LABEL_0",
201
- "1": "LABEL_1"
202
- },
203
- "image_size": 336,
204
- "initializer_factor": 1.0,
205
- "initializer_range": 0.02,
206
- "intermediate_size": 4096,
207
- "is_decoder": false,
208
- "is_encoder_decoder": false,
209
- "label2id": {
210
- "LABEL_0": 0,
211
- "LABEL_1": 1
212
- },
213
- "layer_norm_eps": 1e-05,
214
- "length_penalty": 1.0,
215
- "max_length": 20,
216
- "min_length": 0,
217
- "model_type": "clip_vision_model",
218
- "no_repeat_ngram_size": 0,
219
- "num_attention_heads": 16,
220
- "num_beam_groups": 1,
221
- "num_beams": 1,
222
- "num_channels": 3,
223
- "num_hidden_layers": 24,
224
- "num_return_sequences": 1,
225
- "output_attentions": false,
226
- "output_hidden_states": false,
227
- "output_scores": false,
228
- "pad_token_id": null,
229
- "patch_size": 14,
230
- "prefix": null,
231
- "pretrained_vision_name_or_path": "/hbox2dir/clip-vit-large-patch14-336",
232
- "problem_type": null,
233
- "projection_dim": 768,
234
- "pruned_heads": {},
235
- "remove_invalid_values": false,
236
- "repetition_penalty": 1.0,
237
- "return_dict": true,
238
- "return_dict_in_generate": false,
239
- "sep_token_id": null,
240
- "suppress_tokens": null,
241
- "task_specific_params": null,
242
- "temperature": 1.0,
243
- "tf_legacy_loss": false,
244
- "tie_encoder_decoder": false,
245
- "tie_word_embeddings": true,
246
- "tokenizer_class": null,
247
- "top_k": 50,
248
- "top_p": 1.0,
249
- "torch_dtype": null,
250
- "torchscript": false,
251
- "transformers_version": "4.30.2",
252
- "typical_p": 1.0,
253
- "use_bfloat16": false
254
- },
255
- "visual_projector_config": {
256
- "_name_or_path": "",
257
- "add_cross_attention": false,
258
- "architectures": null,
259
- "attention_probs_dropout_prob": 0.1,
260
- "bad_words_ids": null,
261
- "begin_suppress_tokens": null,
262
- "bos_token_id": null,
263
- "chunk_size_feed_forward": 0,
264
- "cross_attention_hidden_size": null,
265
- "decoder_start_token_id": null,
266
- "depth": 3,
267
- "diversity_penalty": 0.0,
268
- "do_sample": false,
269
- "early_stopping": false,
270
- "encoder_hidden_size": 1024,
271
- "encoder_no_repeat_ngram_size": 0,
272
- "eos_token_id": null,
273
- "exponential_decay_length_penalty": null,
274
- "feature_layer_index": -2,
275
- "finetuning_task": null,
276
- "forced_bos_token_id": null,
277
- "forced_eos_token_id": null,
278
- "hidden_size": 1024,
279
- "id2label": {
280
- "0": "LABEL_0",
281
- "1": "LABEL_1"
282
- },
283
- "initializer_range": 0.02,
284
- "intermediate_size": 4096,
285
- "is_decoder": false,
286
- "is_encoder_decoder": false,
287
- "label2id": {
288
- "LABEL_0": 0,
289
- "LABEL_1": 1
290
- },
291
- "layer_norm_eps": 1e-06,
292
- "length_penalty": 1.0,
293
- "max_length": 20,
294
- "min_length": 0,
295
- "mlp_depth": 2,
296
- "model_type": "mllm_visual_projector",
297
- "no_repeat_ngram_size": 0,
298
- "num_attention_heads": 16,
299
- "num_beam_groups": 1,
300
- "num_beams": 1,
301
- "num_eos_tokens": 0,
302
- "num_hidden_layers": 6,
303
- "num_queries": 576,
304
- "num_return_sequences": 1,
305
- "output_attentions": false,
306
- "output_hidden_states": false,
307
- "output_scores": false,
308
- "pad_token_id": null,
309
- "pos_emb": true,
310
- "prefix": null,
311
- "prenorm": false,
312
- "problem_type": null,
313
- "projector_type": "c-abs",
314
- "pruned_heads": {},
315
- "remove_invalid_values": false,
316
- "repetition_penalty": 1.0,
317
- "return_dict": true,
318
- "return_dict_in_generate": false,
319
- "sep_token_id": null,
320
- "suppress_tokens": null,
321
- "task_specific_params": null,
322
- "temperature": 1.0,
323
- "tf_legacy_loss": false,
324
- "tie_encoder_decoder": false,
325
- "tie_word_embeddings": true,
326
- "tokenizer_class": null,
327
- "top_k": 50,
328
- "top_p": 1.0,
329
- "torch_dtype": null,
330
- "torchscript": false,
331
- "transformers_version": "4.30.2",
332
- "typical_p": 1.0,
333
- "use_bfloat16": false,
334
- "use_cls": true
335
- },
336
- "vocab_size": 32000
337
- }