Hugging Face's logo

nwliu
/
ProDVa-CAMEO

Safetensors
dva
Model card Files
xet
Community
nwliu commited on
Commit
284ab40
·
verified ·
1 Parent(s): 6ced356

Upload folder using huggingface_hub

Browse files
Files changed (20) hide show
  1. config.json +303 -0
  2. generation_config.json +4 -0
  3. lm_tokenizer/merges.txt +0 -0
  4. lm_tokenizer/special_tokens_map.json +24 -0
  5. lm_tokenizer/tokenizer.json +0 -0
  6. lm_tokenizer/tokenizer_config.json +21 -0
  7. lm_tokenizer/vocab.json +0 -0
  8. model-00001-of-00002.safetensors +3 -0
  9. model-00002-of-00002.safetensors +3 -0
  10. model.safetensors.index.json +1037 -0
  11. phrase_tokenizer/merges.txt +0 -0
  12. phrase_tokenizer/special_tokens_map.json +24 -0
  13. phrase_tokenizer/tokenizer.json +0 -0
  14. phrase_tokenizer/tokenizer_config.json +21 -0
  15. phrase_tokenizer/vocab.json +0 -0
  16. text_tokenizer/merges.txt +0 -0
  17. text_tokenizer/special_tokens_map.json +6 -0
  18. text_tokenizer/tokenizer.json +0 -0
  19. text_tokenizer/tokenizer_config.json +21 -0
  20. text_tokenizer/vocab.json +0 -0
config.json ADDED
@@ -0,0 +1,303 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DVAModel"
4
+ ],
5
+ "description_loss_weight": 1.0,
6
+ "description_proj_act": "tanh",
7
+ "description_proj_pdrop": 0.1,
8
+ "language_model_config": {
9
+ "_attn_implementation_autoset": false,
10
+ "_name_or_path": "nferruz/ProtGPT2",
11
+ "activation_function": "gelu_new",
12
+ "add_cross_attention": false,
13
+ "architectures": [
14
+ "GPT2LMHeadModel"
15
+ ],
16
+ "attn_pdrop": 0.1,
17
+ "bad_words_ids": null,
18
+ "begin_suppress_tokens": null,
19
+ "bos_token_id": 0,
20
+ "chunk_size_feed_forward": 0,
21
+ "cross_attention_hidden_size": null,
22
+ "decoder_start_token_id": null,
23
+ "diversity_penalty": 0.0,
24
+ "do_sample": false,
25
+ "early_stopping": false,
26
+ "embd_pdrop": 0.1,
27
+ "encoder_no_repeat_ngram_size": 0,
28
+ "eos_token_id": 0,
29
+ "exponential_decay_length_penalty": null,
30
+ "finetuning_task": null,
31
+ "forced_bos_token_id": null,
32
+ "forced_eos_token_id": null,
33
+ "id2label": {
34
+ "0": "LABEL_0",
35
+ "1": "LABEL_1"
36
+ },
37
+ "initializer_range": 0.02,
38
+ "is_decoder": false,
39
+ "is_encoder_decoder": false,
40
+ "label2id": {
41
+ "LABEL_0": 0,
42
+ "LABEL_1": 1
43
+ },
44
+ "layer_norm_epsilon": 1e-05,
45
+ "length_penalty": 1.0,
46
+ "max_length": 20,
47
+ "min_length": 0,
48
+ "model_type": "gpt2",
49
+ "n_ctx": 1024,
50
+ "n_embd": 1280,
51
+ "n_head": 20,
52
+ "n_inner": null,
53
+ "n_layer": 36,
54
+ "n_positions": 1024,
55
+ "no_repeat_ngram_size": 0,
56
+ "num_beam_groups": 1,
57
+ "num_beams": 1,
58
+ "num_return_sequences": 1,
59
+ "output_attentions": false,
60
+ "output_hidden_states": false,
61
+ "output_scores": false,
62
+ "pad_token_id": null,
63
+ "prefix": null,
64
+ "problem_type": null,
65
+ "pruned_heads": {},
66
+ "remove_invalid_values": false,
67
+ "reorder_and_upcast_attn": false,
68
+ "repetition_penalty": 1.0,
69
+ "resid_pdrop": 0.1,
70
+ "return_dict": true,
71
+ "return_dict_in_generate": false,
72
+ "scale_attn_by_inverse_layer_idx": false,
73
+ "scale_attn_weights": true,
74
+ "sep_token_id": null,
75
+ "summary_activation": null,
76
+ "summary_first_dropout": 0.1,
77
+ "summary_proj_to_labels": true,
78
+ "summary_type": "cls_index",
79
+ "summary_use_proj": true,
80
+ "suppress_tokens": null,
81
+ "task_specific_params": {
82
+ "text-generation": {
83
+ "do_sample": true,
84
+ "max_length": 50
85
+ }
86
+ },
87
+ "temperature": 1.0,
88
+ "tf_legacy_loss": false,
89
+ "tie_encoder_decoder": false,
90
+ "tie_word_embeddings": true,
91
+ "tokenizer_class": null,
92
+ "top_k": 50,
93
+ "top_p": 1.0,
94
+ "torch_dtype": null,
95
+ "torchscript": false,
96
+ "typical_p": 1.0,
97
+ "use_bfloat16": false,
98
+ "use_cache": true,
99
+ "vocab_size": 50257
100
+ },
101
+ "model_type": "dva",
102
+ "phrase_encoder_batch_size": 64,
103
+ "phrase_encoder_config": {
104
+ "_attn_implementation_autoset": false,
105
+ "_name_or_path": "nferruz/ProtGPT2",
106
+ "activation_function": "gelu_new",
107
+ "add_cross_attention": false,
108
+ "architectures": [
109
+ "GPT2LMHeadModel"
110
+ ],
111
+ "attn_pdrop": 0.1,
112
+ "bad_words_ids": null,
113
+ "begin_suppress_tokens": null,
114
+ "bos_token_id": 0,
115
+ "chunk_size_feed_forward": 0,
116
+ "cross_attention_hidden_size": null,
117
+ "decoder_start_token_id": null,
118
+ "diversity_penalty": 0.0,
119
+ "do_sample": false,
120
+ "early_stopping": false,
121
+ "embd_pdrop": 0.1,
122
+ "encoder_no_repeat_ngram_size": 0,
123
+ "eos_token_id": 0,
124
+ "exponential_decay_length_penalty": null,
125
+ "finetuning_task": null,
126
+ "forced_bos_token_id": null,
127
+ "forced_eos_token_id": null,
128
+ "id2label": {
129
+ "0": "LABEL_0",
130
+ "1": "LABEL_1"
131
+ },
132
+ "initializer_range": 0.02,
133
+ "is_decoder": false,
134
+ "is_encoder_decoder": false,
135
+ "label2id": {
136
+ "LABEL_0": 0,
137
+ "LABEL_1": 1
138
+ },
139
+ "layer_norm_epsilon": 1e-05,
140
+ "length_penalty": 1.0,
141
+ "max_length": 20,
142
+ "min_length": 0,
143
+ "model_type": "gpt2",
144
+ "n_ctx": 1024,
145
+ "n_embd": 1280,
146
+ "n_head": 20,
147
+ "n_inner": null,
148
+ "n_layer": 36,
149
+ "n_positions": 1024,
150
+ "no_repeat_ngram_size": 0,
151
+ "num_beam_groups": 1,
152
+ "num_beams": 1,
153
+ "num_return_sequences": 1,
154
+ "output_attentions": false,
155
+ "output_hidden_states": false,
156
+ "output_scores": false,
157
+ "pad_token_id": null,
158
+ "prefix": null,
159
+ "problem_type": null,
160
+ "pruned_heads": {},
161
+ "remove_invalid_values": false,
162
+ "reorder_and_upcast_attn": false,
163
+ "repetition_penalty": 1.0,
164
+ "resid_pdrop": 0.1,
165
+ "return_dict": true,
166
+ "return_dict_in_generate": false,
167
+ "scale_attn_by_inverse_layer_idx": false,
168
+ "scale_attn_weights": true,
169
+ "sep_token_id": null,
170
+ "summary_activation": null,
171
+ "summary_first_dropout": 0.1,
172
+ "summary_proj_to_labels": true,
173
+ "summary_type": "cls_index",
174
+ "summary_use_proj": true,
175
+ "suppress_tokens": null,
176
+ "task_specific_params": {
177
+ "text-generation": {
178
+ "do_sample": true,
179
+ "max_length": 50
180
+ }
181
+ },
182
+ "temperature": 1.0,
183
+ "tf_legacy_loss": false,
184
+ "tie_encoder_decoder": false,
185
+ "tie_word_embeddings": true,
186
+ "tokenizer_class": null,
187
+ "top_k": 50,
188
+ "top_p": 1.0,
189
+ "torch_dtype": null,
190
+ "torchscript": false,
191
+ "typical_p": 1.0,
192
+ "use_bfloat16": false,
193
+ "use_cache": true,
194
+ "vocab_size": 50257
195
+ },
196
+ "phrase_encoder_proj_act": "tanh",
197
+ "phrase_encoder_proj_pdrop": 0.1,
198
+ "text_encoder_config": {
199
+ "_attn_implementation_autoset": false,
200
+ "_name_or_path": "gpt2",
201
+ "activation_function": "gelu_new",
202
+ "add_cross_attention": false,
203
+ "architectures": [
204
+ "GPT2LMHeadModel"
205
+ ],
206
+ "attn_pdrop": 0.1,
207
+ "bad_words_ids": null,
208
+ "begin_suppress_tokens": null,
209
+ "bos_token_id": 50256,
210
+ "chunk_size_feed_forward": 0,
211
+ "cross_attention_hidden_size": null,
212
+ "decoder_start_token_id": null,
213
+ "diversity_penalty": 0.0,
214
+ "do_sample": false,
215
+ "early_stopping": false,
216
+ "embd_pdrop": 0.1,
217
+ "encoder_no_repeat_ngram_size": 0,
218
+ "eos_token_id": 50256,
219
+ "exponential_decay_length_penalty": null,
220
+ "finetuning_task": null,
221
+ "forced_bos_token_id": null,
222
+ "forced_eos_token_id": null,
223
+ "id2label": {
224
+ "0": "LABEL_0",
225
+ "1": "LABEL_1"
226
+ },
227
+ "initializer_range": 0.02,
228
+ "is_decoder": false,
229
+ "is_encoder_decoder": false,
230
+ "label2id": {
231
+ "LABEL_0": 0,
232
+ "LABEL_1": 1
233
+ },
234
+ "layer_norm_epsilon": 1e-05,
235
+ "length_penalty": 1.0,
236
+ "max_length": 20,
237
+ "min_length": 0,
238
+ "model_type": "gpt2",
239
+ "n_ctx": 1024,
240
+ "n_embd": 768,
241
+ "n_head": 12,
242
+ "n_inner": null,
243
+ "n_layer": 12,
244
+ "n_positions": 1024,
245
+ "no_repeat_ngram_size": 0,
246
+ "num_beam_groups": 1,
247
+ "num_beams": 1,
248
+ "num_return_sequences": 1,
249
+ "output_attentions": false,
250
+ "output_hidden_states": false,
251
+ "output_scores": false,
252
+ "pad_token_id": null,
253
+ "prefix": null,
254
+ "problem_type": null,
255
+ "pruned_heads": {},
256
+ "remove_invalid_values": false,
257
+ "reorder_and_upcast_attn": false,
258
+ "repetition_penalty": 1.0,
259
+ "resid_pdrop": 0.1,
260
+ "return_dict": true,
261
+ "return_dict_in_generate": false,
262
+ "scale_attn_by_inverse_layer_idx": false,
263
+ "scale_attn_weights": true,
264
+ "sep_token_id": null,
265
+ "summary_activation": null,
266
+ "summary_first_dropout": 0.1,
267
+ "summary_proj_to_labels": true,
268
+ "summary_type": "cls_index",
269
+ "summary_use_proj": true,
270
+ "suppress_tokens": null,
271
+ "task_specific_params": {
272
+ "text-generation": {
273
+ "do_sample": true,
274
+ "max_length": 50
275
+ }
276
+ },
277
+ "temperature": 1.0,
278
+ "tf_legacy_loss": false,
279
+ "tie_encoder_decoder": false,
280
+ "tie_word_embeddings": true,
281
+ "tokenizer_class": null,
282
+ "top_k": 50,
283
+ "top_p": 1.0,
284
+ "torch_dtype": null,
285
+ "torchscript": false,
286
+ "typical_p": 1.0,
287
+ "use_bfloat16": false,
288
+ "use_cache": true,
289
+ "vocab_size": 50257
290
+ },
291
+ "text_encoder_proj_act": "tanh",
292
+ "text_encoder_proj_pdrop": 0.1,
293
+ "torch_dtype": "float32",
294
+ "transformers_version": "4.51.3",
295
+ "type_classification_head_act": "tanh",
296
+ "type_classification_head_num_classes": 6,
297
+ "type_classification_head_pdrop": 0.1,
298
+ "type_loss_weight": 1.0,
299
+ "use_description_loss": true,
300
+ "use_phrase_encoder_proj": true,
301
+ "use_text_encoder_proj": true,
302
+ "use_type_loss": true
303
+ }
generation_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "transformers_version": "4.51.3"
4
+ }
lm_tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
lm_tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<|endoftext|>",
17
+ "unk_token": {
18
+ "content": "<|endoftext|>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
lm_tokenizer/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
lm_tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1000000000000000019884624838656,
18
+ "pad_token": "<|endoftext|>",
19
+ "tokenizer_class": "GPT2Tokenizer",
20
+ "unk_token": "<|endoftext|>"
21
+ }
lm_tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e25bef735e676638634f4bd1265e713ed39f5fdf7dc51be1045cc566b95241f1
3
+ size 4974606680
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:117f238908da12b1d6ac30f71e0cdb6a1e3a652405fe610a8944fd46237ab9a0
3
+ size 2244604464
model.safetensors.index.json ADDED
@@ -0,0 +1,1037 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 7219095576
4
+ },
5
+ "weight_map": {
6
+ "description_proj.2.bias": "model-00002-of-00002.safetensors",
7
+ "description_proj.2.weight": "model-00002-of-00002.safetensors",
8
+ "language_model.transformer.h.0.attn.c_attn.bias": "model-00001-of-00002.safetensors",
9
+ "language_model.transformer.h.0.attn.c_attn.weight": "model-00001-of-00002.safetensors",
10
+ "language_model.transformer.h.0.attn.c_proj.bias": "model-00001-of-00002.safetensors",
11
+ "language_model.transformer.h.0.attn.c_proj.weight": "model-00001-of-00002.safetensors",
12
+ "language_model.transformer.h.0.ln_1.bias": "model-00001-of-00002.safetensors",
13
+ "language_model.transformer.h.0.ln_1.weight": "model-00001-of-00002.safetensors",
14
+ "language_model.transformer.h.0.ln_2.bias": "model-00001-of-00002.safetensors",
15
+ "language_model.transformer.h.0.ln_2.weight": "model-00001-of-00002.safetensors",
16
+ "language_model.transformer.h.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
17
+ "language_model.transformer.h.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
18
+ "language_model.transformer.h.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
19
+ "language_model.transformer.h.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
20
+ "language_model.transformer.h.1.attn.c_attn.bias": "model-00001-of-00002.safetensors",
21
+ "language_model.transformer.h.1.attn.c_attn.weight": "model-00001-of-00002.safetensors",
22
+ "language_model.transformer.h.1.attn.c_proj.bias": "model-00001-of-00002.safetensors",
23
+ "language_model.transformer.h.1.attn.c_proj.weight": "model-00001-of-00002.safetensors",
24
+ "language_model.transformer.h.1.ln_1.bias": "model-00001-of-00002.safetensors",
25
+ "language_model.transformer.h.1.ln_1.weight": "model-00001-of-00002.safetensors",
26
+ "language_model.transformer.h.1.ln_2.bias": "model-00001-of-00002.safetensors",
27
+ "language_model.transformer.h.1.ln_2.weight": "model-00001-of-00002.safetensors",
28
+ "language_model.transformer.h.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
29
+ "language_model.transformer.h.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
30
+ "language_model.transformer.h.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
31
+ "language_model.transformer.h.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
32
+ "language_model.transformer.h.10.attn.c_attn.bias": "model-00001-of-00002.safetensors",
33
+ "language_model.transformer.h.10.attn.c_attn.weight": "model-00001-of-00002.safetensors",
34
+ "language_model.transformer.h.10.attn.c_proj.bias": "model-00001-of-00002.safetensors",
35
+ "language_model.transformer.h.10.attn.c_proj.weight": "model-00001-of-00002.safetensors",
36
+ "language_model.transformer.h.10.ln_1.bias": "model-00001-of-00002.safetensors",
37
+ "language_model.transformer.h.10.ln_1.weight": "model-00001-of-00002.safetensors",
38
+ "language_model.transformer.h.10.ln_2.bias": "model-00001-of-00002.safetensors",
39
+ "language_model.transformer.h.10.ln_2.weight": "model-00001-of-00002.safetensors",
40
+ "language_model.transformer.h.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
41
+ "language_model.transformer.h.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
42
+ "language_model.transformer.h.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
43
+ "language_model.transformer.h.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
44
+ "language_model.transformer.h.11.attn.c_attn.bias": "model-00001-of-00002.safetensors",
45
+ "language_model.transformer.h.11.attn.c_attn.weight": "model-00001-of-00002.safetensors",
46
+ "language_model.transformer.h.11.attn.c_proj.bias": "model-00001-of-00002.safetensors",
47
+ "language_model.transformer.h.11.attn.c_proj.weight": "model-00001-of-00002.safetensors",
48
+ "language_model.transformer.h.11.ln_1.bias": "model-00001-of-00002.safetensors",
49
+ "language_model.transformer.h.11.ln_1.weight": "model-00001-of-00002.safetensors",
50
+ "language_model.transformer.h.11.ln_2.bias": "model-00001-of-00002.safetensors",
51
+ "language_model.transformer.h.11.ln_2.weight": "model-00001-of-00002.safetensors",
52
+ "language_model.transformer.h.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
53
+ "language_model.transformer.h.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
54
+ "language_model.transformer.h.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
55
+ "language_model.transformer.h.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
56
+ "language_model.transformer.h.12.attn.c_attn.bias": "model-00001-of-00002.safetensors",
57
+ "language_model.transformer.h.12.attn.c_attn.weight": "model-00001-of-00002.safetensors",
58
+ "language_model.transformer.h.12.attn.c_proj.bias": "model-00001-of-00002.safetensors",
59
+ "language_model.transformer.h.12.attn.c_proj.weight": "model-00001-of-00002.safetensors",
60
+ "language_model.transformer.h.12.ln_1.bias": "model-00001-of-00002.safetensors",
61
+ "language_model.transformer.h.12.ln_1.weight": "model-00001-of-00002.safetensors",
62
+ "language_model.transformer.h.12.ln_2.bias": "model-00001-of-00002.safetensors",
63
+ "language_model.transformer.h.12.ln_2.weight": "model-00001-of-00002.safetensors",
64
+ "language_model.transformer.h.12.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
65
+ "language_model.transformer.h.12.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
66
+ "language_model.transformer.h.12.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
67
+ "language_model.transformer.h.12.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
68
+ "language_model.transformer.h.13.attn.c_attn.bias": "model-00001-of-00002.safetensors",
69
+ "language_model.transformer.h.13.attn.c_attn.weight": "model-00001-of-00002.safetensors",
70
+ "language_model.transformer.h.13.attn.c_proj.bias": "model-00001-of-00002.safetensors",
71
+ "language_model.transformer.h.13.attn.c_proj.weight": "model-00001-of-00002.safetensors",
72
+ "language_model.transformer.h.13.ln_1.bias": "model-00001-of-00002.safetensors",
73
+ "language_model.transformer.h.13.ln_1.weight": "model-00001-of-00002.safetensors",
74
+ "language_model.transformer.h.13.ln_2.bias": "model-00001-of-00002.safetensors",
75
+ "language_model.transformer.h.13.ln_2.weight": "model-00001-of-00002.safetensors",
76
+ "language_model.transformer.h.13.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
77
+ "language_model.transformer.h.13.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
78
+ "language_model.transformer.h.13.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
79
+ "language_model.transformer.h.13.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
80
+ "language_model.transformer.h.14.attn.c_attn.bias": "model-00001-of-00002.safetensors",
81
+ "language_model.transformer.h.14.attn.c_attn.weight": "model-00001-of-00002.safetensors",
82
+ "language_model.transformer.h.14.attn.c_proj.bias": "model-00001-of-00002.safetensors",
83
+ "language_model.transformer.h.14.attn.c_proj.weight": "model-00001-of-00002.safetensors",
84
+ "language_model.transformer.h.14.ln_1.bias": "model-00001-of-00002.safetensors",
85
+ "language_model.transformer.h.14.ln_1.weight": "model-00001-of-00002.safetensors",
86
+ "language_model.transformer.h.14.ln_2.bias": "model-00001-of-00002.safetensors",
87
+ "language_model.transformer.h.14.ln_2.weight": "model-00001-of-00002.safetensors",
88
+ "language_model.transformer.h.14.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
89
+ "language_model.transformer.h.14.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
90
+ "language_model.transformer.h.14.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
91
+ "language_model.transformer.h.14.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
92
+ "language_model.transformer.h.15.attn.c_attn.bias": "model-00001-of-00002.safetensors",
93
+ "language_model.transformer.h.15.attn.c_attn.weight": "model-00001-of-00002.safetensors",
94
+ "language_model.transformer.h.15.attn.c_proj.bias": "model-00001-of-00002.safetensors",
95
+ "language_model.transformer.h.15.attn.c_proj.weight": "model-00001-of-00002.safetensors",
96
+ "language_model.transformer.h.15.ln_1.bias": "model-00001-of-00002.safetensors",
97
+ "language_model.transformer.h.15.ln_1.weight": "model-00001-of-00002.safetensors",
98
+ "language_model.transformer.h.15.ln_2.bias": "model-00001-of-00002.safetensors",
99
+ "language_model.transformer.h.15.ln_2.weight": "model-00001-of-00002.safetensors",
100
+ "language_model.transformer.h.15.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
101
+ "language_model.transformer.h.15.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
102
+ "language_model.transformer.h.15.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
103
+ "language_model.transformer.h.15.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
104
+ "language_model.transformer.h.16.attn.c_attn.bias": "model-00001-of-00002.safetensors",
105
+ "language_model.transformer.h.16.attn.c_attn.weight": "model-00001-of-00002.safetensors",
106
+ "language_model.transformer.h.16.attn.c_proj.bias": "model-00001-of-00002.safetensors",
107
+ "language_model.transformer.h.16.attn.c_proj.weight": "model-00001-of-00002.safetensors",
108
+ "language_model.transformer.h.16.ln_1.bias": "model-00001-of-00002.safetensors",
109
+ "language_model.transformer.h.16.ln_1.weight": "model-00001-of-00002.safetensors",
110
+ "language_model.transformer.h.16.ln_2.bias": "model-00001-of-00002.safetensors",
111
+ "language_model.transformer.h.16.ln_2.weight": "model-00001-of-00002.safetensors",
112
+ "language_model.transformer.h.16.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
113
+ "language_model.transformer.h.16.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
114
+ "language_model.transformer.h.16.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
115
+ "language_model.transformer.h.16.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
116
+ "language_model.transformer.h.17.attn.c_attn.bias": "model-00001-of-00002.safetensors",
117
+ "language_model.transformer.h.17.attn.c_attn.weight": "model-00001-of-00002.safetensors",
118
+ "language_model.transformer.h.17.attn.c_proj.bias": "model-00001-of-00002.safetensors",
119
+ "language_model.transformer.h.17.attn.c_proj.weight": "model-00001-of-00002.safetensors",
120
+ "language_model.transformer.h.17.ln_1.bias": "model-00001-of-00002.safetensors",
121
+ "language_model.transformer.h.17.ln_1.weight": "model-00001-of-00002.safetensors",
122
+ "language_model.transformer.h.17.ln_2.bias": "model-00001-of-00002.safetensors",
123
+ "language_model.transformer.h.17.ln_2.weight": "model-00001-of-00002.safetensors",
124
+ "language_model.transformer.h.17.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
125
+ "language_model.transformer.h.17.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
126
+ "language_model.transformer.h.17.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
127
+ "language_model.transformer.h.17.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
128
+ "language_model.transformer.h.18.attn.c_attn.bias": "model-00001-of-00002.safetensors",
129
+ "language_model.transformer.h.18.attn.c_attn.weight": "model-00001-of-00002.safetensors",
130
+ "language_model.transformer.h.18.attn.c_proj.bias": "model-00001-of-00002.safetensors",
131
+ "language_model.transformer.h.18.attn.c_proj.weight": "model-00001-of-00002.safetensors",
132
+ "language_model.transformer.h.18.ln_1.bias": "model-00001-of-00002.safetensors",
133
+ "language_model.transformer.h.18.ln_1.weight": "model-00001-of-00002.safetensors",
134
+ "language_model.transformer.h.18.ln_2.bias": "model-00001-of-00002.safetensors",
135
+ "language_model.transformer.h.18.ln_2.weight": "model-00001-of-00002.safetensors",
136
+ "language_model.transformer.h.18.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
137
+ "language_model.transformer.h.18.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
138
+ "language_model.transformer.h.18.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
139
+ "language_model.transformer.h.18.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
140
+ "language_model.transformer.h.19.attn.c_attn.bias": "model-00001-of-00002.safetensors",
141
+ "language_model.transformer.h.19.attn.c_attn.weight": "model-00001-of-00002.safetensors",
142
+ "language_model.transformer.h.19.attn.c_proj.bias": "model-00001-of-00002.safetensors",
143
+ "language_model.transformer.h.19.attn.c_proj.weight": "model-00001-of-00002.safetensors",
144
+ "language_model.transformer.h.19.ln_1.bias": "model-00001-of-00002.safetensors",
145
+ "language_model.transformer.h.19.ln_1.weight": "model-00001-of-00002.safetensors",
146
+ "language_model.transformer.h.19.ln_2.bias": "model-00001-of-00002.safetensors",
147
+ "language_model.transformer.h.19.ln_2.weight": "model-00001-of-00002.safetensors",
148
+ "language_model.transformer.h.19.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
149
+ "language_model.transformer.h.19.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
150
+ "language_model.transformer.h.19.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
151
+ "language_model.transformer.h.19.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
152
+ "language_model.transformer.h.2.attn.c_attn.bias": "model-00001-of-00002.safetensors",
153
+ "language_model.transformer.h.2.attn.c_attn.weight": "model-00001-of-00002.safetensors",
154
+ "language_model.transformer.h.2.attn.c_proj.bias": "model-00001-of-00002.safetensors",
155
+ "language_model.transformer.h.2.attn.c_proj.weight": "model-00001-of-00002.safetensors",
156
+ "language_model.transformer.h.2.ln_1.bias": "model-00001-of-00002.safetensors",
157
+ "language_model.transformer.h.2.ln_1.weight": "model-00001-of-00002.safetensors",
158
+ "language_model.transformer.h.2.ln_2.bias": "model-00001-of-00002.safetensors",
159
+ "language_model.transformer.h.2.ln_2.weight": "model-00001-of-00002.safetensors",
160
+ "language_model.transformer.h.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
161
+ "language_model.transformer.h.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
162
+ "language_model.transformer.h.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
163
+ "language_model.transformer.h.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
164
+ "language_model.transformer.h.20.attn.c_attn.bias": "model-00001-of-00002.safetensors",
165
+ "language_model.transformer.h.20.attn.c_attn.weight": "model-00001-of-00002.safetensors",
166
+ "language_model.transformer.h.20.attn.c_proj.bias": "model-00001-of-00002.safetensors",
167
+ "language_model.transformer.h.20.attn.c_proj.weight": "model-00001-of-00002.safetensors",
168
+ "language_model.transformer.h.20.ln_1.bias": "model-00001-of-00002.safetensors",
169
+ "language_model.transformer.h.20.ln_1.weight": "model-00001-of-00002.safetensors",
170
+ "language_model.transformer.h.20.ln_2.bias": "model-00001-of-00002.safetensors",
171
+ "language_model.transformer.h.20.ln_2.weight": "model-00001-of-00002.safetensors",
172
+ "language_model.transformer.h.20.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
173
+ "language_model.transformer.h.20.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
174
+ "language_model.transformer.h.20.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
175
+ "language_model.transformer.h.20.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
176
+ "language_model.transformer.h.21.attn.c_attn.bias": "model-00001-of-00002.safetensors",
177
+ "language_model.transformer.h.21.attn.c_attn.weight": "model-00001-of-00002.safetensors",
178
+ "language_model.transformer.h.21.attn.c_proj.bias": "model-00001-of-00002.safetensors",
179
+ "language_model.transformer.h.21.attn.c_proj.weight": "model-00001-of-00002.safetensors",
180
+ "language_model.transformer.h.21.ln_1.bias": "model-00001-of-00002.safetensors",
181
+ "language_model.transformer.h.21.ln_1.weight": "model-00001-of-00002.safetensors",
182
+ "language_model.transformer.h.21.ln_2.bias": "model-00001-of-00002.safetensors",
183
+ "language_model.transformer.h.21.ln_2.weight": "model-00001-of-00002.safetensors",
184
+ "language_model.transformer.h.21.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
185
+ "language_model.transformer.h.21.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
186
+ "language_model.transformer.h.21.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
187
+ "language_model.transformer.h.21.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
188
+ "language_model.transformer.h.22.attn.c_attn.bias": "model-00001-of-00002.safetensors",
189
+ "language_model.transformer.h.22.attn.c_attn.weight": "model-00001-of-00002.safetensors",
190
+ "language_model.transformer.h.22.attn.c_proj.bias": "model-00001-of-00002.safetensors",
191
+ "language_model.transformer.h.22.attn.c_proj.weight": "model-00001-of-00002.safetensors",
192
+ "language_model.transformer.h.22.ln_1.bias": "model-00001-of-00002.safetensors",
193
+ "language_model.transformer.h.22.ln_1.weight": "model-00001-of-00002.safetensors",
194
+ "language_model.transformer.h.22.ln_2.bias": "model-00001-of-00002.safetensors",
195
+ "language_model.transformer.h.22.ln_2.weight": "model-00001-of-00002.safetensors",
196
+ "language_model.transformer.h.22.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
197
+ "language_model.transformer.h.22.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
198
+ "language_model.transformer.h.22.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
199
+ "language_model.transformer.h.22.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
200
+ "language_model.transformer.h.23.attn.c_attn.bias": "model-00001-of-00002.safetensors",
201
+ "language_model.transformer.h.23.attn.c_attn.weight": "model-00001-of-00002.safetensors",
202
+ "language_model.transformer.h.23.attn.c_proj.bias": "model-00001-of-00002.safetensors",
203
+ "language_model.transformer.h.23.attn.c_proj.weight": "model-00001-of-00002.safetensors",
204
+ "language_model.transformer.h.23.ln_1.bias": "model-00001-of-00002.safetensors",
205
+ "language_model.transformer.h.23.ln_1.weight": "model-00001-of-00002.safetensors",
206
+ "language_model.transformer.h.23.ln_2.bias": "model-00001-of-00002.safetensors",
207
+ "language_model.transformer.h.23.ln_2.weight": "model-00001-of-00002.safetensors",
208
+ "language_model.transformer.h.23.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
209
+ "language_model.transformer.h.23.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
210
+ "language_model.transformer.h.23.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
211
+ "language_model.transformer.h.23.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
212
+ "language_model.transformer.h.24.attn.c_attn.bias": "model-00001-of-00002.safetensors",
213
+ "language_model.transformer.h.24.attn.c_attn.weight": "model-00001-of-00002.safetensors",
214
+ "language_model.transformer.h.24.attn.c_proj.bias": "model-00001-of-00002.safetensors",
215
+ "language_model.transformer.h.24.attn.c_proj.weight": "model-00001-of-00002.safetensors",
216
+ "language_model.transformer.h.24.ln_1.bias": "model-00001-of-00002.safetensors",
217
+ "language_model.transformer.h.24.ln_1.weight": "model-00001-of-00002.safetensors",
218
+ "language_model.transformer.h.24.ln_2.bias": "model-00001-of-00002.safetensors",
219
+ "language_model.transformer.h.24.ln_2.weight": "model-00001-of-00002.safetensors",
220
+ "language_model.transformer.h.24.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
221
+ "language_model.transformer.h.24.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
222
+ "language_model.transformer.h.24.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
223
+ "language_model.transformer.h.24.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
224
+ "language_model.transformer.h.25.attn.c_attn.bias": "model-00001-of-00002.safetensors",
225
+ "language_model.transformer.h.25.attn.c_attn.weight": "model-00001-of-00002.safetensors",
226
+ "language_model.transformer.h.25.attn.c_proj.bias": "model-00001-of-00002.safetensors",
227
+ "language_model.transformer.h.25.attn.c_proj.weight": "model-00001-of-00002.safetensors",
228
+ "language_model.transformer.h.25.ln_1.bias": "model-00001-of-00002.safetensors",
229
+ "language_model.transformer.h.25.ln_1.weight": "model-00001-of-00002.safetensors",
230
+ "language_model.transformer.h.25.ln_2.bias": "model-00001-of-00002.safetensors",
231
+ "language_model.transformer.h.25.ln_2.weight": "model-00001-of-00002.safetensors",
232
+ "language_model.transformer.h.25.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
233
+ "language_model.transformer.h.25.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
234
+ "language_model.transformer.h.25.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
235
+ "language_model.transformer.h.25.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
236
+ "language_model.transformer.h.26.attn.c_attn.bias": "model-00001-of-00002.safetensors",
237
+ "language_model.transformer.h.26.attn.c_attn.weight": "model-00001-of-00002.safetensors",
238
+ "language_model.transformer.h.26.attn.c_proj.bias": "model-00001-of-00002.safetensors",
239
+ "language_model.transformer.h.26.attn.c_proj.weight": "model-00001-of-00002.safetensors",
240
+ "language_model.transformer.h.26.ln_1.bias": "model-00001-of-00002.safetensors",
241
+ "language_model.transformer.h.26.ln_1.weight": "model-00001-of-00002.safetensors",
242
+ "language_model.transformer.h.26.ln_2.bias": "model-00001-of-00002.safetensors",
243
+ "language_model.transformer.h.26.ln_2.weight": "model-00001-of-00002.safetensors",
244
+ "language_model.transformer.h.26.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
245
+ "language_model.transformer.h.26.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
246
+ "language_model.transformer.h.26.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
247
+ "language_model.transformer.h.26.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
248
+ "language_model.transformer.h.27.attn.c_attn.bias": "model-00001-of-00002.safetensors",
249
+ "language_model.transformer.h.27.attn.c_attn.weight": "model-00001-of-00002.safetensors",
250
+ "language_model.transformer.h.27.attn.c_proj.bias": "model-00001-of-00002.safetensors",
251
+ "language_model.transformer.h.27.attn.c_proj.weight": "model-00001-of-00002.safetensors",
252
+ "language_model.transformer.h.27.ln_1.bias": "model-00001-of-00002.safetensors",
253
+ "language_model.transformer.h.27.ln_1.weight": "model-00001-of-00002.safetensors",
254
+ "language_model.transformer.h.27.ln_2.bias": "model-00001-of-00002.safetensors",
255
+ "language_model.transformer.h.27.ln_2.weight": "model-00001-of-00002.safetensors",
256
+ "language_model.transformer.h.27.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
257
+ "language_model.transformer.h.27.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
258
+ "language_model.transformer.h.27.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
259
+ "language_model.transformer.h.27.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
260
+ "language_model.transformer.h.28.attn.c_attn.bias": "model-00001-of-00002.safetensors",
261
+ "language_model.transformer.h.28.attn.c_attn.weight": "model-00001-of-00002.safetensors",
262
+ "language_model.transformer.h.28.attn.c_proj.bias": "model-00001-of-00002.safetensors",
263
+ "language_model.transformer.h.28.attn.c_proj.weight": "model-00001-of-00002.safetensors",
264
+ "language_model.transformer.h.28.ln_1.bias": "model-00001-of-00002.safetensors",
265
+ "language_model.transformer.h.28.ln_1.weight": "model-00001-of-00002.safetensors",
266
+ "language_model.transformer.h.28.ln_2.bias": "model-00001-of-00002.safetensors",
267
+ "language_model.transformer.h.28.ln_2.weight": "model-00001-of-00002.safetensors",
268
+ "language_model.transformer.h.28.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
269
+ "language_model.transformer.h.28.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
270
+ "language_model.transformer.h.28.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
271
+ "language_model.transformer.h.28.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
272
+ "language_model.transformer.h.29.attn.c_attn.bias": "model-00001-of-00002.safetensors",
273
+ "language_model.transformer.h.29.attn.c_attn.weight": "model-00001-of-00002.safetensors",
274
+ "language_model.transformer.h.29.attn.c_proj.bias": "model-00001-of-00002.safetensors",
275
+ "language_model.transformer.h.29.attn.c_proj.weight": "model-00001-of-00002.safetensors",
276
+ "language_model.transformer.h.29.ln_1.bias": "model-00001-of-00002.safetensors",
277
+ "language_model.transformer.h.29.ln_1.weight": "model-00001-of-00002.safetensors",
278
+ "language_model.transformer.h.29.ln_2.bias": "model-00001-of-00002.safetensors",
279
+ "language_model.transformer.h.29.ln_2.weight": "model-00001-of-00002.safetensors",
280
+ "language_model.transformer.h.29.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
281
+ "language_model.transformer.h.29.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
282
+ "language_model.transformer.h.29.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
283
+ "language_model.transformer.h.29.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
284
+ "language_model.transformer.h.3.attn.c_attn.bias": "model-00001-of-00002.safetensors",
285
+ "language_model.transformer.h.3.attn.c_attn.weight": "model-00001-of-00002.safetensors",
286
+ "language_model.transformer.h.3.attn.c_proj.bias": "model-00001-of-00002.safetensors",
287
+ "language_model.transformer.h.3.attn.c_proj.weight": "model-00001-of-00002.safetensors",
288
+ "language_model.transformer.h.3.ln_1.bias": "model-00001-of-00002.safetensors",
289
+ "language_model.transformer.h.3.ln_1.weight": "model-00001-of-00002.safetensors",
290
+ "language_model.transformer.h.3.ln_2.bias": "model-00001-of-00002.safetensors",
291
+ "language_model.transformer.h.3.ln_2.weight": "model-00001-of-00002.safetensors",
292
+ "language_model.transformer.h.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
293
+ "language_model.transformer.h.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
294
+ "language_model.transformer.h.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
295
+ "language_model.transformer.h.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
296
+ "language_model.transformer.h.30.attn.c_attn.bias": "model-00001-of-00002.safetensors",
297
+ "language_model.transformer.h.30.attn.c_attn.weight": "model-00001-of-00002.safetensors",
298
+ "language_model.transformer.h.30.attn.c_proj.bias": "model-00001-of-00002.safetensors",
299
+ "language_model.transformer.h.30.attn.c_proj.weight": "model-00001-of-00002.safetensors",
300
+ "language_model.transformer.h.30.ln_1.bias": "model-00001-of-00002.safetensors",
301
+ "language_model.transformer.h.30.ln_1.weight": "model-00001-of-00002.safetensors",
302
+ "language_model.transformer.h.30.ln_2.bias": "model-00001-of-00002.safetensors",
303
+ "language_model.transformer.h.30.ln_2.weight": "model-00001-of-00002.safetensors",
304
+ "language_model.transformer.h.30.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
305
+ "language_model.transformer.h.30.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
306
+ "language_model.transformer.h.30.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
307
+ "language_model.transformer.h.30.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
308
+ "language_model.transformer.h.31.attn.c_attn.bias": "model-00001-of-00002.safetensors",
309
+ "language_model.transformer.h.31.attn.c_attn.weight": "model-00001-of-00002.safetensors",
310
+ "language_model.transformer.h.31.attn.c_proj.bias": "model-00001-of-00002.safetensors",
311
+ "language_model.transformer.h.31.attn.c_proj.weight": "model-00001-of-00002.safetensors",
312
+ "language_model.transformer.h.31.ln_1.bias": "model-00001-of-00002.safetensors",
313
+ "language_model.transformer.h.31.ln_1.weight": "model-00001-of-00002.safetensors",
314
+ "language_model.transformer.h.31.ln_2.bias": "model-00001-of-00002.safetensors",
315
+ "language_model.transformer.h.31.ln_2.weight": "model-00001-of-00002.safetensors",
316
+ "language_model.transformer.h.31.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
317
+ "language_model.transformer.h.31.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
318
+ "language_model.transformer.h.31.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
319
+ "language_model.transformer.h.31.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
320
+ "language_model.transformer.h.32.attn.c_attn.bias": "model-00001-of-00002.safetensors",
321
+ "language_model.transformer.h.32.attn.c_attn.weight": "model-00001-of-00002.safetensors",
322
+ "language_model.transformer.h.32.attn.c_proj.bias": "model-00001-of-00002.safetensors",
323
+ "language_model.transformer.h.32.attn.c_proj.weight": "model-00001-of-00002.safetensors",
324
+ "language_model.transformer.h.32.ln_1.bias": "model-00001-of-00002.safetensors",
325
+ "language_model.transformer.h.32.ln_1.weight": "model-00001-of-00002.safetensors",
326
+ "language_model.transformer.h.32.ln_2.bias": "model-00001-of-00002.safetensors",
327
+ "language_model.transformer.h.32.ln_2.weight": "model-00001-of-00002.safetensors",
328
+ "language_model.transformer.h.32.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
329
+ "language_model.transformer.h.32.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
330
+ "language_model.transformer.h.32.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
331
+ "language_model.transformer.h.32.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
332
+ "language_model.transformer.h.33.attn.c_attn.bias": "model-00001-of-00002.safetensors",
333
+ "language_model.transformer.h.33.attn.c_attn.weight": "model-00001-of-00002.safetensors",
334
+ "language_model.transformer.h.33.attn.c_proj.bias": "model-00001-of-00002.safetensors",
335
+ "language_model.transformer.h.33.attn.c_proj.weight": "model-00001-of-00002.safetensors",
336
+ "language_model.transformer.h.33.ln_1.bias": "model-00001-of-00002.safetensors",
337
+ "language_model.transformer.h.33.ln_1.weight": "model-00001-of-00002.safetensors",
338
+ "language_model.transformer.h.33.ln_2.bias": "model-00001-of-00002.safetensors",
339
+ "language_model.transformer.h.33.ln_2.weight": "model-00001-of-00002.safetensors",
340
+ "language_model.transformer.h.33.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
341
+ "language_model.transformer.h.33.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
342
+ "language_model.transformer.h.33.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
343
+ "language_model.transformer.h.33.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
344
+ "language_model.transformer.h.34.attn.c_attn.bias": "model-00001-of-00002.safetensors",
345
+ "language_model.transformer.h.34.attn.c_attn.weight": "model-00001-of-00002.safetensors",
346
+ "language_model.transformer.h.34.attn.c_proj.bias": "model-00001-of-00002.safetensors",
347
+ "language_model.transformer.h.34.attn.c_proj.weight": "model-00001-of-00002.safetensors",
348
+ "language_model.transformer.h.34.ln_1.bias": "model-00001-of-00002.safetensors",
349
+ "language_model.transformer.h.34.ln_1.weight": "model-00001-of-00002.safetensors",
350
+ "language_model.transformer.h.34.ln_2.bias": "model-00001-of-00002.safetensors",
351
+ "language_model.transformer.h.34.ln_2.weight": "model-00001-of-00002.safetensors",
352
+ "language_model.transformer.h.34.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
353
+ "language_model.transformer.h.34.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
354
+ "language_model.transformer.h.34.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
355
+ "language_model.transformer.h.34.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
356
+ "language_model.transformer.h.35.attn.c_attn.bias": "model-00001-of-00002.safetensors",
357
+ "language_model.transformer.h.35.attn.c_attn.weight": "model-00001-of-00002.safetensors",
358
+ "language_model.transformer.h.35.attn.c_proj.bias": "model-00001-of-00002.safetensors",
359
+ "language_model.transformer.h.35.attn.c_proj.weight": "model-00001-of-00002.safetensors",
360
+ "language_model.transformer.h.35.ln_1.bias": "model-00001-of-00002.safetensors",
361
+ "language_model.transformer.h.35.ln_1.weight": "model-00001-of-00002.safetensors",
362
+ "language_model.transformer.h.35.ln_2.bias": "model-00001-of-00002.safetensors",
363
+ "language_model.transformer.h.35.ln_2.weight": "model-00001-of-00002.safetensors",
364
+ "language_model.transformer.h.35.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
365
+ "language_model.transformer.h.35.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
366
+ "language_model.transformer.h.35.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
367
+ "language_model.transformer.h.35.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
368
+ "language_model.transformer.h.4.attn.c_attn.bias": "model-00001-of-00002.safetensors",
369
+ "language_model.transformer.h.4.attn.c_attn.weight": "model-00001-of-00002.safetensors",
370
+ "language_model.transformer.h.4.attn.c_proj.bias": "model-00001-of-00002.safetensors",
371
+ "language_model.transformer.h.4.attn.c_proj.weight": "model-00001-of-00002.safetensors",
372
+ "language_model.transformer.h.4.ln_1.bias": "model-00001-of-00002.safetensors",
373
+ "language_model.transformer.h.4.ln_1.weight": "model-00001-of-00002.safetensors",
374
+ "language_model.transformer.h.4.ln_2.bias": "model-00001-of-00002.safetensors",
375
+ "language_model.transformer.h.4.ln_2.weight": "model-00001-of-00002.safetensors",
376
+ "language_model.transformer.h.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
377
+ "language_model.transformer.h.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
378
+ "language_model.transformer.h.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
379
+ "language_model.transformer.h.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
380
+ "language_model.transformer.h.5.attn.c_attn.bias": "model-00001-of-00002.safetensors",
381
+ "language_model.transformer.h.5.attn.c_attn.weight": "model-00001-of-00002.safetensors",
382
+ "language_model.transformer.h.5.attn.c_proj.bias": "model-00001-of-00002.safetensors",
383
+ "language_model.transformer.h.5.attn.c_proj.weight": "model-00001-of-00002.safetensors",
384
+ "language_model.transformer.h.5.ln_1.bias": "model-00001-of-00002.safetensors",
385
+ "language_model.transformer.h.5.ln_1.weight": "model-00001-of-00002.safetensors",
386
+ "language_model.transformer.h.5.ln_2.bias": "model-00001-of-00002.safetensors",
387
+ "language_model.transformer.h.5.ln_2.weight": "model-00001-of-00002.safetensors",
388
+ "language_model.transformer.h.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
389
+ "language_model.transformer.h.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
390
+ "language_model.transformer.h.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
391
+ "language_model.transformer.h.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
392
+ "language_model.transformer.h.6.attn.c_attn.bias": "model-00001-of-00002.safetensors",
393
+ "language_model.transformer.h.6.attn.c_attn.weight": "model-00001-of-00002.safetensors",
394
+ "language_model.transformer.h.6.attn.c_proj.bias": "model-00001-of-00002.safetensors",
395
+ "language_model.transformer.h.6.attn.c_proj.weight": "model-00001-of-00002.safetensors",
396
+ "language_model.transformer.h.6.ln_1.bias": "model-00001-of-00002.safetensors",
397
+ "language_model.transformer.h.6.ln_1.weight": "model-00001-of-00002.safetensors",
398
+ "language_model.transformer.h.6.ln_2.bias": "model-00001-of-00002.safetensors",
399
+ "language_model.transformer.h.6.ln_2.weight": "model-00001-of-00002.safetensors",
400
+ "language_model.transformer.h.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
401
+ "language_model.transformer.h.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
402
+ "language_model.transformer.h.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
403
+ "language_model.transformer.h.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
404
+ "language_model.transformer.h.7.attn.c_attn.bias": "model-00001-of-00002.safetensors",
405
+ "language_model.transformer.h.7.attn.c_attn.weight": "model-00001-of-00002.safetensors",
406
+ "language_model.transformer.h.7.attn.c_proj.bias": "model-00001-of-00002.safetensors",
407
+ "language_model.transformer.h.7.attn.c_proj.weight": "model-00001-of-00002.safetensors",
408
+ "language_model.transformer.h.7.ln_1.bias": "model-00001-of-00002.safetensors",
409
+ "language_model.transformer.h.7.ln_1.weight": "model-00001-of-00002.safetensors",
410
+ "language_model.transformer.h.7.ln_2.bias": "model-00001-of-00002.safetensors",
411
+ "language_model.transformer.h.7.ln_2.weight": "model-00001-of-00002.safetensors",
412
+ "language_model.transformer.h.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
413
+ "language_model.transformer.h.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
414
+ "language_model.transformer.h.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
415
+ "language_model.transformer.h.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
416
+ "language_model.transformer.h.8.attn.c_attn.bias": "model-00001-of-00002.safetensors",
417
+ "language_model.transformer.h.8.attn.c_attn.weight": "model-00001-of-00002.safetensors",
418
+ "language_model.transformer.h.8.attn.c_proj.bias": "model-00001-of-00002.safetensors",
419
+ "language_model.transformer.h.8.attn.c_proj.weight": "model-00001-of-00002.safetensors",
420
+ "language_model.transformer.h.8.ln_1.bias": "model-00001-of-00002.safetensors",
421
+ "language_model.transformer.h.8.ln_1.weight": "model-00001-of-00002.safetensors",
422
+ "language_model.transformer.h.8.ln_2.bias": "model-00001-of-00002.safetensors",
423
+ "language_model.transformer.h.8.ln_2.weight": "model-00001-of-00002.safetensors",
424
+ "language_model.transformer.h.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
425
+ "language_model.transformer.h.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
426
+ "language_model.transformer.h.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
427
+ "language_model.transformer.h.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
428
+ "language_model.transformer.h.9.attn.c_attn.bias": "model-00001-of-00002.safetensors",
429
+ "language_model.transformer.h.9.attn.c_attn.weight": "model-00001-of-00002.safetensors",
430
+ "language_model.transformer.h.9.attn.c_proj.bias": "model-00001-of-00002.safetensors",
431
+ "language_model.transformer.h.9.attn.c_proj.weight": "model-00001-of-00002.safetensors",
432
+ "language_model.transformer.h.9.ln_1.bias": "model-00001-of-00002.safetensors",
433
+ "language_model.transformer.h.9.ln_1.weight": "model-00001-of-00002.safetensors",
434
+ "language_model.transformer.h.9.ln_2.bias": "model-00001-of-00002.safetensors",
435
+ "language_model.transformer.h.9.ln_2.weight": "model-00001-of-00002.safetensors",
436
+ "language_model.transformer.h.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
437
+ "language_model.transformer.h.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
438
+ "language_model.transformer.h.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
439
+ "language_model.transformer.h.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
440
+ "language_model.transformer.ln_f.bias": "model-00001-of-00002.safetensors",
441
+ "language_model.transformer.ln_f.weight": "model-00001-of-00002.safetensors",
442
+ "language_model.transformer.wpe.weight": "model-00001-of-00002.safetensors",
443
+ "language_model.transformer.wte.weight": "model-00001-of-00002.safetensors",
444
+ "phrase_encoder.h.0.attn.c_attn.bias": "model-00001-of-00002.safetensors",
445
+ "phrase_encoder.h.0.attn.c_attn.weight": "model-00001-of-00002.safetensors",
446
+ "phrase_encoder.h.0.attn.c_proj.bias": "model-00001-of-00002.safetensors",
447
+ "phrase_encoder.h.0.attn.c_proj.weight": "model-00001-of-00002.safetensors",
448
+ "phrase_encoder.h.0.ln_1.bias": "model-00001-of-00002.safetensors",
449
+ "phrase_encoder.h.0.ln_1.weight": "model-00001-of-00002.safetensors",
450
+ "phrase_encoder.h.0.ln_2.bias": "model-00001-of-00002.safetensors",
451
+ "phrase_encoder.h.0.ln_2.weight": "model-00001-of-00002.safetensors",
452
+ "phrase_encoder.h.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
453
+ "phrase_encoder.h.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
454
+ "phrase_encoder.h.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
455
+ "phrase_encoder.h.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
456
+ "phrase_encoder.h.1.attn.c_attn.bias": "model-00001-of-00002.safetensors",
457
+ "phrase_encoder.h.1.attn.c_attn.weight": "model-00001-of-00002.safetensors",
458
+ "phrase_encoder.h.1.attn.c_proj.bias": "model-00001-of-00002.safetensors",
459
+ "phrase_encoder.h.1.attn.c_proj.weight": "model-00001-of-00002.safetensors",
460
+ "phrase_encoder.h.1.ln_1.bias": "model-00001-of-00002.safetensors",
461
+ "phrase_encoder.h.1.ln_1.weight": "model-00001-of-00002.safetensors",
462
+ "phrase_encoder.h.1.ln_2.bias": "model-00001-of-00002.safetensors",
463
+ "phrase_encoder.h.1.ln_2.weight": "model-00001-of-00002.safetensors",
464
+ "phrase_encoder.h.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
465
+ "phrase_encoder.h.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
466
+ "phrase_encoder.h.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
467
+ "phrase_encoder.h.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
468
+ "phrase_encoder.h.10.attn.c_attn.bias": "model-00002-of-00002.safetensors",
469
+ "phrase_encoder.h.10.attn.c_attn.weight": "model-00002-of-00002.safetensors",
470
+ "phrase_encoder.h.10.attn.c_proj.bias": "model-00002-of-00002.safetensors",
471
+ "phrase_encoder.h.10.attn.c_proj.weight": "model-00002-of-00002.safetensors",
472
+ "phrase_encoder.h.10.ln_1.bias": "model-00002-of-00002.safetensors",
473
+ "phrase_encoder.h.10.ln_1.weight": "model-00002-of-00002.safetensors",
474
+ "phrase_encoder.h.10.ln_2.bias": "model-00002-of-00002.safetensors",
475
+ "phrase_encoder.h.10.ln_2.weight": "model-00002-of-00002.safetensors",
476
+ "phrase_encoder.h.10.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
477
+ "phrase_encoder.h.10.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
478
+ "phrase_encoder.h.10.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
479
+ "phrase_encoder.h.10.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
480
+ "phrase_encoder.h.11.attn.c_attn.bias": "model-00002-of-00002.safetensors",
481
+ "phrase_encoder.h.11.attn.c_attn.weight": "model-00002-of-00002.safetensors",
482
+ "phrase_encoder.h.11.attn.c_proj.bias": "model-00002-of-00002.safetensors",
483
+ "phrase_encoder.h.11.attn.c_proj.weight": "model-00002-of-00002.safetensors",
484
+ "phrase_encoder.h.11.ln_1.bias": "model-00002-of-00002.safetensors",
485
+ "phrase_encoder.h.11.ln_1.weight": "model-00002-of-00002.safetensors",
486
+ "phrase_encoder.h.11.ln_2.bias": "model-00002-of-00002.safetensors",
487
+ "phrase_encoder.h.11.ln_2.weight": "model-00002-of-00002.safetensors",
488
+ "phrase_encoder.h.11.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
489
+ "phrase_encoder.h.11.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
490
+ "phrase_encoder.h.11.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
491
+ "phrase_encoder.h.11.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
492
+ "phrase_encoder.h.12.attn.c_attn.bias": "model-00002-of-00002.safetensors",
493
+ "phrase_encoder.h.12.attn.c_attn.weight": "model-00002-of-00002.safetensors",
494
+ "phrase_encoder.h.12.attn.c_proj.bias": "model-00002-of-00002.safetensors",
495
+ "phrase_encoder.h.12.attn.c_proj.weight": "model-00002-of-00002.safetensors",
496
+ "phrase_encoder.h.12.ln_1.bias": "model-00002-of-00002.safetensors",
497
+ "phrase_encoder.h.12.ln_1.weight": "model-00002-of-00002.safetensors",
498
+ "phrase_encoder.h.12.ln_2.bias": "model-00002-of-00002.safetensors",
499
+ "phrase_encoder.h.12.ln_2.weight": "model-00002-of-00002.safetensors",
500
+ "phrase_encoder.h.12.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
501
+ "phrase_encoder.h.12.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
502
+ "phrase_encoder.h.12.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
503
+ "phrase_encoder.h.12.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
504
+ "phrase_encoder.h.13.attn.c_attn.bias": "model-00002-of-00002.safetensors",
505
+ "phrase_encoder.h.13.attn.c_attn.weight": "model-00002-of-00002.safetensors",
506
+ "phrase_encoder.h.13.attn.c_proj.bias": "model-00002-of-00002.safetensors",
507
+ "phrase_encoder.h.13.attn.c_proj.weight": "model-00002-of-00002.safetensors",
508
+ "phrase_encoder.h.13.ln_1.bias": "model-00002-of-00002.safetensors",
509
+ "phrase_encoder.h.13.ln_1.weight": "model-00002-of-00002.safetensors",
510
+ "phrase_encoder.h.13.ln_2.bias": "model-00002-of-00002.safetensors",
511
+ "phrase_encoder.h.13.ln_2.weight": "model-00002-of-00002.safetensors",
512
+ "phrase_encoder.h.13.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
513
+ "phrase_encoder.h.13.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
514
+ "phrase_encoder.h.13.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
515
+ "phrase_encoder.h.13.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
516
+ "phrase_encoder.h.14.attn.c_attn.bias": "model-00002-of-00002.safetensors",
517
+ "phrase_encoder.h.14.attn.c_attn.weight": "model-00002-of-00002.safetensors",
518
+ "phrase_encoder.h.14.attn.c_proj.bias": "model-00002-of-00002.safetensors",
519
+ "phrase_encoder.h.14.attn.c_proj.weight": "model-00002-of-00002.safetensors",
520
+ "phrase_encoder.h.14.ln_1.bias": "model-00002-of-00002.safetensors",
521
+ "phrase_encoder.h.14.ln_1.weight": "model-00002-of-00002.safetensors",
522
+ "phrase_encoder.h.14.ln_2.bias": "model-00002-of-00002.safetensors",
523
+ "phrase_encoder.h.14.ln_2.weight": "model-00002-of-00002.safetensors",
524
+ "phrase_encoder.h.14.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
525
+ "phrase_encoder.h.14.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
526
+ "phrase_encoder.h.14.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
527
+ "phrase_encoder.h.14.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
528
+ "phrase_encoder.h.15.attn.c_attn.bias": "model-00002-of-00002.safetensors",
529
+ "phrase_encoder.h.15.attn.c_attn.weight": "model-00002-of-00002.safetensors",
530
+ "phrase_encoder.h.15.attn.c_proj.bias": "model-00002-of-00002.safetensors",
531
+ "phrase_encoder.h.15.attn.c_proj.weight": "model-00002-of-00002.safetensors",
532
+ "phrase_encoder.h.15.ln_1.bias": "model-00002-of-00002.safetensors",
533
+ "phrase_encoder.h.15.ln_1.weight": "model-00002-of-00002.safetensors",
534
+ "phrase_encoder.h.15.ln_2.bias": "model-00002-of-00002.safetensors",
535
+ "phrase_encoder.h.15.ln_2.weight": "model-00002-of-00002.safetensors",
536
+ "phrase_encoder.h.15.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
537
+ "phrase_encoder.h.15.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
538
+ "phrase_encoder.h.15.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
539
+ "phrase_encoder.h.15.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
540
+ "phrase_encoder.h.16.attn.c_attn.bias": "model-00002-of-00002.safetensors",
541
+ "phrase_encoder.h.16.attn.c_attn.weight": "model-00002-of-00002.safetensors",
542
+ "phrase_encoder.h.16.attn.c_proj.bias": "model-00002-of-00002.safetensors",
543
+ "phrase_encoder.h.16.attn.c_proj.weight": "model-00002-of-00002.safetensors",
544
+ "phrase_encoder.h.16.ln_1.bias": "model-00002-of-00002.safetensors",
545
+ "phrase_encoder.h.16.ln_1.weight": "model-00002-of-00002.safetensors",
546
+ "phrase_encoder.h.16.ln_2.bias": "model-00002-of-00002.safetensors",
547
+ "phrase_encoder.h.16.ln_2.weight": "model-00002-of-00002.safetensors",
548
+ "phrase_encoder.h.16.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
549
+ "phrase_encoder.h.16.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
550
+ "phrase_encoder.h.16.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
551
+ "phrase_encoder.h.16.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
552
+ "phrase_encoder.h.17.attn.c_attn.bias": "model-00002-of-00002.safetensors",
553
+ "phrase_encoder.h.17.attn.c_attn.weight": "model-00002-of-00002.safetensors",
554
+ "phrase_encoder.h.17.attn.c_proj.bias": "model-00002-of-00002.safetensors",
555
+ "phrase_encoder.h.17.attn.c_proj.weight": "model-00002-of-00002.safetensors",
556
+ "phrase_encoder.h.17.ln_1.bias": "model-00002-of-00002.safetensors",
557
+ "phrase_encoder.h.17.ln_1.weight": "model-00002-of-00002.safetensors",
558
+ "phrase_encoder.h.17.ln_2.bias": "model-00002-of-00002.safetensors",
559
+ "phrase_encoder.h.17.ln_2.weight": "model-00002-of-00002.safetensors",
560
+ "phrase_encoder.h.17.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
561
+ "phrase_encoder.h.17.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
562
+ "phrase_encoder.h.17.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
563
+ "phrase_encoder.h.17.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
564
+ "phrase_encoder.h.18.attn.c_attn.bias": "model-00002-of-00002.safetensors",
565
+ "phrase_encoder.h.18.attn.c_attn.weight": "model-00002-of-00002.safetensors",
566
+ "phrase_encoder.h.18.attn.c_proj.bias": "model-00002-of-00002.safetensors",
567
+ "phrase_encoder.h.18.attn.c_proj.weight": "model-00002-of-00002.safetensors",
568
+ "phrase_encoder.h.18.ln_1.bias": "model-00002-of-00002.safetensors",
569
+ "phrase_encoder.h.18.ln_1.weight": "model-00002-of-00002.safetensors",
570
+ "phrase_encoder.h.18.ln_2.bias": "model-00002-of-00002.safetensors",
571
+ "phrase_encoder.h.18.ln_2.weight": "model-00002-of-00002.safetensors",
572
+ "phrase_encoder.h.18.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
573
+ "phrase_encoder.h.18.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
574
+ "phrase_encoder.h.18.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
575
+ "phrase_encoder.h.18.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
576
+ "phrase_encoder.h.19.attn.c_attn.bias": "model-00002-of-00002.safetensors",
577
+ "phrase_encoder.h.19.attn.c_attn.weight": "model-00002-of-00002.safetensors",
578
+ "phrase_encoder.h.19.attn.c_proj.bias": "model-00002-of-00002.safetensors",
579
+ "phrase_encoder.h.19.attn.c_proj.weight": "model-00002-of-00002.safetensors",
580
+ "phrase_encoder.h.19.ln_1.bias": "model-00002-of-00002.safetensors",
581
+ "phrase_encoder.h.19.ln_1.weight": "model-00002-of-00002.safetensors",
582
+ "phrase_encoder.h.19.ln_2.bias": "model-00002-of-00002.safetensors",
583
+ "phrase_encoder.h.19.ln_2.weight": "model-00002-of-00002.safetensors",
584
+ "phrase_encoder.h.19.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
585
+ "phrase_encoder.h.19.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
586
+ "phrase_encoder.h.19.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
587
+ "phrase_encoder.h.19.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
588
+ "phrase_encoder.h.2.attn.c_attn.bias": "model-00001-of-00002.safetensors",
589
+ "phrase_encoder.h.2.attn.c_attn.weight": "model-00001-of-00002.safetensors",
590
+ "phrase_encoder.h.2.attn.c_proj.bias": "model-00001-of-00002.safetensors",
591
+ "phrase_encoder.h.2.attn.c_proj.weight": "model-00001-of-00002.safetensors",
592
+ "phrase_encoder.h.2.ln_1.bias": "model-00001-of-00002.safetensors",
593
+ "phrase_encoder.h.2.ln_1.weight": "model-00001-of-00002.safetensors",
594
+ "phrase_encoder.h.2.ln_2.bias": "model-00001-of-00002.safetensors",
595
+ "phrase_encoder.h.2.ln_2.weight": "model-00001-of-00002.safetensors",
596
+ "phrase_encoder.h.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
597
+ "phrase_encoder.h.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
598
+ "phrase_encoder.h.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
599
+ "phrase_encoder.h.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
600
+ "phrase_encoder.h.20.attn.c_attn.bias": "model-00002-of-00002.safetensors",
601
+ "phrase_encoder.h.20.attn.c_attn.weight": "model-00002-of-00002.safetensors",
602
+ "phrase_encoder.h.20.attn.c_proj.bias": "model-00002-of-00002.safetensors",
603
+ "phrase_encoder.h.20.attn.c_proj.weight": "model-00002-of-00002.safetensors",
604
+ "phrase_encoder.h.20.ln_1.bias": "model-00002-of-00002.safetensors",
605
+ "phrase_encoder.h.20.ln_1.weight": "model-00002-of-00002.safetensors",
606
+ "phrase_encoder.h.20.ln_2.bias": "model-00002-of-00002.safetensors",
607
+ "phrase_encoder.h.20.ln_2.weight": "model-00002-of-00002.safetensors",
608
+ "phrase_encoder.h.20.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
609
+ "phrase_encoder.h.20.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
610
+ "phrase_encoder.h.20.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
611
+ "phrase_encoder.h.20.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
612
+ "phrase_encoder.h.21.attn.c_attn.bias": "model-00002-of-00002.safetensors",
613
+ "phrase_encoder.h.21.attn.c_attn.weight": "model-00002-of-00002.safetensors",
614
+ "phrase_encoder.h.21.attn.c_proj.bias": "model-00002-of-00002.safetensors",
615
+ "phrase_encoder.h.21.attn.c_proj.weight": "model-00002-of-00002.safetensors",
616
+ "phrase_encoder.h.21.ln_1.bias": "model-00002-of-00002.safetensors",
617
+ "phrase_encoder.h.21.ln_1.weight": "model-00002-of-00002.safetensors",
618
+ "phrase_encoder.h.21.ln_2.bias": "model-00002-of-00002.safetensors",
619
+ "phrase_encoder.h.21.ln_2.weight": "model-00002-of-00002.safetensors",
620
+ "phrase_encoder.h.21.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
621
+ "phrase_encoder.h.21.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
622
+ "phrase_encoder.h.21.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
623
+ "phrase_encoder.h.21.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
624
+ "phrase_encoder.h.22.attn.c_attn.bias": "model-00002-of-00002.safetensors",
625
+ "phrase_encoder.h.22.attn.c_attn.weight": "model-00002-of-00002.safetensors",
626
+ "phrase_encoder.h.22.attn.c_proj.bias": "model-00002-of-00002.safetensors",
627
+ "phrase_encoder.h.22.attn.c_proj.weight": "model-00002-of-00002.safetensors",
628
+ "phrase_encoder.h.22.ln_1.bias": "model-00002-of-00002.safetensors",
629
+ "phrase_encoder.h.22.ln_1.weight": "model-00002-of-00002.safetensors",
630
+ "phrase_encoder.h.22.ln_2.bias": "model-00002-of-00002.safetensors",
631
+ "phrase_encoder.h.22.ln_2.weight": "model-00002-of-00002.safetensors",
632
+ "phrase_encoder.h.22.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
633
+ "phrase_encoder.h.22.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
634
+ "phrase_encoder.h.22.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
635
+ "phrase_encoder.h.22.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
636
+ "phrase_encoder.h.23.attn.c_attn.bias": "model-00002-of-00002.safetensors",
637
+ "phrase_encoder.h.23.attn.c_attn.weight": "model-00002-of-00002.safetensors",
638
+ "phrase_encoder.h.23.attn.c_proj.bias": "model-00002-of-00002.safetensors",
639
+ "phrase_encoder.h.23.attn.c_proj.weight": "model-00002-of-00002.safetensors",
640
+ "phrase_encoder.h.23.ln_1.bias": "model-00002-of-00002.safetensors",
641
+ "phrase_encoder.h.23.ln_1.weight": "model-00002-of-00002.safetensors",
642
+ "phrase_encoder.h.23.ln_2.bias": "model-00002-of-00002.safetensors",
643
+ "phrase_encoder.h.23.ln_2.weight": "model-00002-of-00002.safetensors",
644
+ "phrase_encoder.h.23.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
645
+ "phrase_encoder.h.23.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
646
+ "phrase_encoder.h.23.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
647
+ "phrase_encoder.h.23.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
648
+ "phrase_encoder.h.24.attn.c_attn.bias": "model-00002-of-00002.safetensors",
649
+ "phrase_encoder.h.24.attn.c_attn.weight": "model-00002-of-00002.safetensors",
650
+ "phrase_encoder.h.24.attn.c_proj.bias": "model-00002-of-00002.safetensors",
651
+ "phrase_encoder.h.24.attn.c_proj.weight": "model-00002-of-00002.safetensors",
652
+ "phrase_encoder.h.24.ln_1.bias": "model-00002-of-00002.safetensors",
653
+ "phrase_encoder.h.24.ln_1.weight": "model-00002-of-00002.safetensors",
654
+ "phrase_encoder.h.24.ln_2.bias": "model-00002-of-00002.safetensors",
655
+ "phrase_encoder.h.24.ln_2.weight": "model-00002-of-00002.safetensors",
656
+ "phrase_encoder.h.24.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
657
+ "phrase_encoder.h.24.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
658
+ "phrase_encoder.h.24.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
659
+ "phrase_encoder.h.24.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
660
+ "phrase_encoder.h.25.attn.c_attn.bias": "model-00002-of-00002.safetensors",
661
+ "phrase_encoder.h.25.attn.c_attn.weight": "model-00002-of-00002.safetensors",
662
+ "phrase_encoder.h.25.attn.c_proj.bias": "model-00002-of-00002.safetensors",
663
+ "phrase_encoder.h.25.attn.c_proj.weight": "model-00002-of-00002.safetensors",
664
+ "phrase_encoder.h.25.ln_1.bias": "model-00002-of-00002.safetensors",
665
+ "phrase_encoder.h.25.ln_1.weight": "model-00002-of-00002.safetensors",
666
+ "phrase_encoder.h.25.ln_2.bias": "model-00002-of-00002.safetensors",
667
+ "phrase_encoder.h.25.ln_2.weight": "model-00002-of-00002.safetensors",
668
+ "phrase_encoder.h.25.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
669
+ "phrase_encoder.h.25.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
670
+ "phrase_encoder.h.25.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
671
+ "phrase_encoder.h.25.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
672
+ "phrase_encoder.h.26.attn.c_attn.bias": "model-00002-of-00002.safetensors",
673
+ "phrase_encoder.h.26.attn.c_attn.weight": "model-00002-of-00002.safetensors",
674
+ "phrase_encoder.h.26.attn.c_proj.bias": "model-00002-of-00002.safetensors",
675
+ "phrase_encoder.h.26.attn.c_proj.weight": "model-00002-of-00002.safetensors",
676
+ "phrase_encoder.h.26.ln_1.bias": "model-00002-of-00002.safetensors",
677
+ "phrase_encoder.h.26.ln_1.weight": "model-00002-of-00002.safetensors",
678
+ "phrase_encoder.h.26.ln_2.bias": "model-00002-of-00002.safetensors",
679
+ "phrase_encoder.h.26.ln_2.weight": "model-00002-of-00002.safetensors",
680
+ "phrase_encoder.h.26.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
681
+ "phrase_encoder.h.26.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
682
+ "phrase_encoder.h.26.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
683
+ "phrase_encoder.h.26.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
684
+ "phrase_encoder.h.27.attn.c_attn.bias": "model-00002-of-00002.safetensors",
685
+ "phrase_encoder.h.27.attn.c_attn.weight": "model-00002-of-00002.safetensors",
686
+ "phrase_encoder.h.27.attn.c_proj.bias": "model-00002-of-00002.safetensors",
687
+ "phrase_encoder.h.27.attn.c_proj.weight": "model-00002-of-00002.safetensors",
688
+ "phrase_encoder.h.27.ln_1.bias": "model-00002-of-00002.safetensors",
689
+ "phrase_encoder.h.27.ln_1.weight": "model-00002-of-00002.safetensors",
690
+ "phrase_encoder.h.27.ln_2.bias": "model-00002-of-00002.safetensors",
691
+ "phrase_encoder.h.27.ln_2.weight": "model-00002-of-00002.safetensors",
692
+ "phrase_encoder.h.27.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
693
+ "phrase_encoder.h.27.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
694
+ "phrase_encoder.h.27.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
695
+ "phrase_encoder.h.27.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
696
+ "phrase_encoder.h.28.attn.c_attn.bias": "model-00002-of-00002.safetensors",
697
+ "phrase_encoder.h.28.attn.c_attn.weight": "model-00002-of-00002.safetensors",
698
+ "phrase_encoder.h.28.attn.c_proj.bias": "model-00002-of-00002.safetensors",
699
+ "phrase_encoder.h.28.attn.c_proj.weight": "model-00002-of-00002.safetensors",
700
+ "phrase_encoder.h.28.ln_1.bias": "model-00002-of-00002.safetensors",
701
+ "phrase_encoder.h.28.ln_1.weight": "model-00002-of-00002.safetensors",
702
+ "phrase_encoder.h.28.ln_2.bias": "model-00002-of-00002.safetensors",
703
+ "phrase_encoder.h.28.ln_2.weight": "model-00002-of-00002.safetensors",
704
+ "phrase_encoder.h.28.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
705
+ "phrase_encoder.h.28.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
706
+ "phrase_encoder.h.28.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
707
+ "phrase_encoder.h.28.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
708
+ "phrase_encoder.h.29.attn.c_attn.bias": "model-00002-of-00002.safetensors",
709
+ "phrase_encoder.h.29.attn.c_attn.weight": "model-00002-of-00002.safetensors",
710
+ "phrase_encoder.h.29.attn.c_proj.bias": "model-00002-of-00002.safetensors",
711
+ "phrase_encoder.h.29.attn.c_proj.weight": "model-00002-of-00002.safetensors",
712
+ "phrase_encoder.h.29.ln_1.bias": "model-00002-of-00002.safetensors",
713
+ "phrase_encoder.h.29.ln_1.weight": "model-00002-of-00002.safetensors",
714
+ "phrase_encoder.h.29.ln_2.bias": "model-00002-of-00002.safetensors",
715
+ "phrase_encoder.h.29.ln_2.weight": "model-00002-of-00002.safetensors",
716
+ "phrase_encoder.h.29.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
717
+ "phrase_encoder.h.29.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
718
+ "phrase_encoder.h.29.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
719
+ "phrase_encoder.h.29.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
720
+ "phrase_encoder.h.3.attn.c_attn.bias": "model-00001-of-00002.safetensors",
721
+ "phrase_encoder.h.3.attn.c_attn.weight": "model-00001-of-00002.safetensors",
722
+ "phrase_encoder.h.3.attn.c_proj.bias": "model-00001-of-00002.safetensors",
723
+ "phrase_encoder.h.3.attn.c_proj.weight": "model-00001-of-00002.safetensors",
724
+ "phrase_encoder.h.3.ln_1.bias": "model-00001-of-00002.safetensors",
725
+ "phrase_encoder.h.3.ln_1.weight": "model-00001-of-00002.safetensors",
726
+ "phrase_encoder.h.3.ln_2.bias": "model-00001-of-00002.safetensors",
727
+ "phrase_encoder.h.3.ln_2.weight": "model-00001-of-00002.safetensors",
728
+ "phrase_encoder.h.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
729
+ "phrase_encoder.h.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
730
+ "phrase_encoder.h.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
731
+ "phrase_encoder.h.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
732
+ "phrase_encoder.h.30.attn.c_attn.bias": "model-00002-of-00002.safetensors",
733
+ "phrase_encoder.h.30.attn.c_attn.weight": "model-00002-of-00002.safetensors",
734
+ "phrase_encoder.h.30.attn.c_proj.bias": "model-00002-of-00002.safetensors",
735
+ "phrase_encoder.h.30.attn.c_proj.weight": "model-00002-of-00002.safetensors",
736
+ "phrase_encoder.h.30.ln_1.bias": "model-00002-of-00002.safetensors",
737
+ "phrase_encoder.h.30.ln_1.weight": "model-00002-of-00002.safetensors",
738
+ "phrase_encoder.h.30.ln_2.bias": "model-00002-of-00002.safetensors",
739
+ "phrase_encoder.h.30.ln_2.weight": "model-00002-of-00002.safetensors",
740
+ "phrase_encoder.h.30.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
741
+ "phrase_encoder.h.30.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
742
+ "phrase_encoder.h.30.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
743
+ "phrase_encoder.h.30.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
744
+ "phrase_encoder.h.31.attn.c_attn.bias": "model-00002-of-00002.safetensors",
745
+ "phrase_encoder.h.31.attn.c_attn.weight": "model-00002-of-00002.safetensors",
746
+ "phrase_encoder.h.31.attn.c_proj.bias": "model-00002-of-00002.safetensors",
747
+ "phrase_encoder.h.31.attn.c_proj.weight": "model-00002-of-00002.safetensors",
748
+ "phrase_encoder.h.31.ln_1.bias": "model-00002-of-00002.safetensors",
749
+ "phrase_encoder.h.31.ln_1.weight": "model-00002-of-00002.safetensors",
750
+ "phrase_encoder.h.31.ln_2.bias": "model-00002-of-00002.safetensors",
751
+ "phrase_encoder.h.31.ln_2.weight": "model-00002-of-00002.safetensors",
752
+ "phrase_encoder.h.31.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
753
+ "phrase_encoder.h.31.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
754
+ "phrase_encoder.h.31.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
755
+ "phrase_encoder.h.31.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
756
+ "phrase_encoder.h.32.attn.c_attn.bias": "model-00002-of-00002.safetensors",
757
+ "phrase_encoder.h.32.attn.c_attn.weight": "model-00002-of-00002.safetensors",
758
+ "phrase_encoder.h.32.attn.c_proj.bias": "model-00002-of-00002.safetensors",
759
+ "phrase_encoder.h.32.attn.c_proj.weight": "model-00002-of-00002.safetensors",
760
+ "phrase_encoder.h.32.ln_1.bias": "model-00002-of-00002.safetensors",
761
+ "phrase_encoder.h.32.ln_1.weight": "model-00002-of-00002.safetensors",
762
+ "phrase_encoder.h.32.ln_2.bias": "model-00002-of-00002.safetensors",
763
+ "phrase_encoder.h.32.ln_2.weight": "model-00002-of-00002.safetensors",
764
+ "phrase_encoder.h.32.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
765
+ "phrase_encoder.h.32.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
766
+ "phrase_encoder.h.32.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
767
+ "phrase_encoder.h.32.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
768
+ "phrase_encoder.h.33.attn.c_attn.bias": "model-00002-of-00002.safetensors",
769
+ "phrase_encoder.h.33.attn.c_attn.weight": "model-00002-of-00002.safetensors",
770
+ "phrase_encoder.h.33.attn.c_proj.bias": "model-00002-of-00002.safetensors",
771
+ "phrase_encoder.h.33.attn.c_proj.weight": "model-00002-of-00002.safetensors",
772
+ "phrase_encoder.h.33.ln_1.bias": "model-00002-of-00002.safetensors",
773
+ "phrase_encoder.h.33.ln_1.weight": "model-00002-of-00002.safetensors",
774
+ "phrase_encoder.h.33.ln_2.bias": "model-00002-of-00002.safetensors",
775
+ "phrase_encoder.h.33.ln_2.weight": "model-00002-of-00002.safetensors",
776
+ "phrase_encoder.h.33.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
777
+ "phrase_encoder.h.33.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
778
+ "phrase_encoder.h.33.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
779
+ "phrase_encoder.h.33.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
780
+ "phrase_encoder.h.34.attn.c_attn.bias": "model-00002-of-00002.safetensors",
781
+ "phrase_encoder.h.34.attn.c_attn.weight": "model-00002-of-00002.safetensors",
782
+ "phrase_encoder.h.34.attn.c_proj.bias": "model-00002-of-00002.safetensors",
783
+ "phrase_encoder.h.34.attn.c_proj.weight": "model-00002-of-00002.safetensors",
784
+ "phrase_encoder.h.34.ln_1.bias": "model-00002-of-00002.safetensors",
785
+ "phrase_encoder.h.34.ln_1.weight": "model-00002-of-00002.safetensors",
786
+ "phrase_encoder.h.34.ln_2.bias": "model-00002-of-00002.safetensors",
787
+ "phrase_encoder.h.34.ln_2.weight": "model-00002-of-00002.safetensors",
788
+ "phrase_encoder.h.34.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
789
+ "phrase_encoder.h.34.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
790
+ "phrase_encoder.h.34.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
791
+ "phrase_encoder.h.34.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
792
+ "phrase_encoder.h.35.attn.c_attn.bias": "model-00002-of-00002.safetensors",
793
+ "phrase_encoder.h.35.attn.c_attn.weight": "model-00002-of-00002.safetensors",
794
+ "phrase_encoder.h.35.attn.c_proj.bias": "model-00002-of-00002.safetensors",
795
+ "phrase_encoder.h.35.attn.c_proj.weight": "model-00002-of-00002.safetensors",
796
+ "phrase_encoder.h.35.ln_1.bias": "model-00002-of-00002.safetensors",
797
+ "phrase_encoder.h.35.ln_1.weight": "model-00002-of-00002.safetensors",
798
+ "phrase_encoder.h.35.ln_2.bias": "model-00002-of-00002.safetensors",
799
+ "phrase_encoder.h.35.ln_2.weight": "model-00002-of-00002.safetensors",
800
+ "phrase_encoder.h.35.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
801
+ "phrase_encoder.h.35.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
802
+ "phrase_encoder.h.35.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
803
+ "phrase_encoder.h.35.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
804
+ "phrase_encoder.h.4.attn.c_attn.bias": "model-00001-of-00002.safetensors",
805
+ "phrase_encoder.h.4.attn.c_attn.weight": "model-00001-of-00002.safetensors",
806
+ "phrase_encoder.h.4.attn.c_proj.bias": "model-00001-of-00002.safetensors",
807
+ "phrase_encoder.h.4.attn.c_proj.weight": "model-00001-of-00002.safetensors",
808
+ "phrase_encoder.h.4.ln_1.bias": "model-00001-of-00002.safetensors",
809
+ "phrase_encoder.h.4.ln_1.weight": "model-00001-of-00002.safetensors",
810
+ "phrase_encoder.h.4.ln_2.bias": "model-00001-of-00002.safetensors",
811
+ "phrase_encoder.h.4.ln_2.weight": "model-00001-of-00002.safetensors",
812
+ "phrase_encoder.h.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
813
+ "phrase_encoder.h.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
814
+ "phrase_encoder.h.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
815
+ "phrase_encoder.h.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
816
+ "phrase_encoder.h.5.attn.c_attn.bias": "model-00001-of-00002.safetensors",
817
+ "phrase_encoder.h.5.attn.c_attn.weight": "model-00001-of-00002.safetensors",
818
+ "phrase_encoder.h.5.attn.c_proj.bias": "model-00001-of-00002.safetensors",
819
+ "phrase_encoder.h.5.attn.c_proj.weight": "model-00001-of-00002.safetensors",
820
+ "phrase_encoder.h.5.ln_1.bias": "model-00001-of-00002.safetensors",
821
+ "phrase_encoder.h.5.ln_1.weight": "model-00001-of-00002.safetensors",
822
+ "phrase_encoder.h.5.ln_2.bias": "model-00001-of-00002.safetensors",
823
+ "phrase_encoder.h.5.ln_2.weight": "model-00001-of-00002.safetensors",
824
+ "phrase_encoder.h.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
825
+ "phrase_encoder.h.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
826
+ "phrase_encoder.h.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
827
+ "phrase_encoder.h.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
828
+ "phrase_encoder.h.6.attn.c_attn.bias": "model-00001-of-00002.safetensors",
829
+ "phrase_encoder.h.6.attn.c_attn.weight": "model-00001-of-00002.safetensors",
830
+ "phrase_encoder.h.6.attn.c_proj.bias": "model-00001-of-00002.safetensors",
831
+ "phrase_encoder.h.6.attn.c_proj.weight": "model-00001-of-00002.safetensors",
832
+ "phrase_encoder.h.6.ln_1.bias": "model-00001-of-00002.safetensors",
833
+ "phrase_encoder.h.6.ln_1.weight": "model-00001-of-00002.safetensors",
834
+ "phrase_encoder.h.6.ln_2.bias": "model-00001-of-00002.safetensors",
835
+ "phrase_encoder.h.6.ln_2.weight": "model-00001-of-00002.safetensors",
836
+ "phrase_encoder.h.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
837
+ "phrase_encoder.h.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
838
+ "phrase_encoder.h.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
839
+ "phrase_encoder.h.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
840
+ "phrase_encoder.h.7.attn.c_attn.bias": "model-00001-of-00002.safetensors",
841
+ "phrase_encoder.h.7.attn.c_attn.weight": "model-00001-of-00002.safetensors",
842
+ "phrase_encoder.h.7.attn.c_proj.bias": "model-00001-of-00002.safetensors",
843
+ "phrase_encoder.h.7.attn.c_proj.weight": "model-00001-of-00002.safetensors",
844
+ "phrase_encoder.h.7.ln_1.bias": "model-00001-of-00002.safetensors",
845
+ "phrase_encoder.h.7.ln_1.weight": "model-00001-of-00002.safetensors",
846
+ "phrase_encoder.h.7.ln_2.bias": "model-00001-of-00002.safetensors",
847
+ "phrase_encoder.h.7.ln_2.weight": "model-00001-of-00002.safetensors",
848
+ "phrase_encoder.h.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
849
+ "phrase_encoder.h.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
850
+ "phrase_encoder.h.7.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
851
+ "phrase_encoder.h.7.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
852
+ "phrase_encoder.h.8.attn.c_attn.bias": "model-00002-of-00002.safetensors",
853
+ "phrase_encoder.h.8.attn.c_attn.weight": "model-00002-of-00002.safetensors",
854
+ "phrase_encoder.h.8.attn.c_proj.bias": "model-00002-of-00002.safetensors",
855
+ "phrase_encoder.h.8.attn.c_proj.weight": "model-00002-of-00002.safetensors",
856
+ "phrase_encoder.h.8.ln_1.bias": "model-00002-of-00002.safetensors",
857
+ "phrase_encoder.h.8.ln_1.weight": "model-00002-of-00002.safetensors",
858
+ "phrase_encoder.h.8.ln_2.bias": "model-00002-of-00002.safetensors",
859
+ "phrase_encoder.h.8.ln_2.weight": "model-00002-of-00002.safetensors",
860
+ "phrase_encoder.h.8.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
861
+ "phrase_encoder.h.8.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
862
+ "phrase_encoder.h.8.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
863
+ "phrase_encoder.h.8.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
864
+ "phrase_encoder.h.9.attn.c_attn.bias": "model-00002-of-00002.safetensors",
865
+ "phrase_encoder.h.9.attn.c_attn.weight": "model-00002-of-00002.safetensors",
866
+ "phrase_encoder.h.9.attn.c_proj.bias": "model-00002-of-00002.safetensors",
867
+ "phrase_encoder.h.9.attn.c_proj.weight": "model-00002-of-00002.safetensors",
868
+ "phrase_encoder.h.9.ln_1.bias": "model-00002-of-00002.safetensors",
869
+ "phrase_encoder.h.9.ln_1.weight": "model-00002-of-00002.safetensors",
870
+ "phrase_encoder.h.9.ln_2.bias": "model-00002-of-00002.safetensors",
871
+ "phrase_encoder.h.9.ln_2.weight": "model-00002-of-00002.safetensors",
872
+ "phrase_encoder.h.9.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
873
+ "phrase_encoder.h.9.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
874
+ "phrase_encoder.h.9.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
875
+ "phrase_encoder.h.9.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
876
+ "phrase_encoder.ln_f.bias": "model-00002-of-00002.safetensors",
877
+ "phrase_encoder.ln_f.weight": "model-00002-of-00002.safetensors",
878
+ "phrase_encoder.wpe.weight": "model-00001-of-00002.safetensors",
879
+ "phrase_encoder.wte.weight": "model-00001-of-00002.safetensors",
880
+ "phrase_encoder_proj.2.bias": "model-00002-of-00002.safetensors",
881
+ "phrase_encoder_proj.2.weight": "model-00002-of-00002.safetensors",
882
+ "sv_input_embeddings": "model-00001-of-00002.safetensors",
883
+ "sv_output_embeddings": "model-00001-of-00002.safetensors",
884
+ "text_encoder.h.0.attn.c_attn.bias": "model-00001-of-00002.safetensors",
885
+ "text_encoder.h.0.attn.c_attn.weight": "model-00001-of-00002.safetensors",
886
+ "text_encoder.h.0.attn.c_proj.bias": "model-00001-of-00002.safetensors",
887
+ "text_encoder.h.0.attn.c_proj.weight": "model-00001-of-00002.safetensors",
888
+ "text_encoder.h.0.ln_1.bias": "model-00001-of-00002.safetensors",
889
+ "text_encoder.h.0.ln_1.weight": "model-00001-of-00002.safetensors",
890
+ "text_encoder.h.0.ln_2.bias": "model-00001-of-00002.safetensors",
891
+ "text_encoder.h.0.ln_2.weight": "model-00001-of-00002.safetensors",
892
+ "text_encoder.h.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
893
+ "text_encoder.h.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
894
+ "text_encoder.h.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
895
+ "text_encoder.h.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
896
+ "text_encoder.h.1.attn.c_attn.bias": "model-00001-of-00002.safetensors",
897
+ "text_encoder.h.1.attn.c_attn.weight": "model-00001-of-00002.safetensors",
898
+ "text_encoder.h.1.attn.c_proj.bias": "model-00001-of-00002.safetensors",
899
+ "text_encoder.h.1.attn.c_proj.weight": "model-00001-of-00002.safetensors",
900
+ "text_encoder.h.1.ln_1.bias": "model-00001-of-00002.safetensors",
901
+ "text_encoder.h.1.ln_1.weight": "model-00001-of-00002.safetensors",
902
+ "text_encoder.h.1.ln_2.bias": "model-00001-of-00002.safetensors",
903
+ "text_encoder.h.1.ln_2.weight": "model-00001-of-00002.safetensors",
904
+ "text_encoder.h.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
905
+ "text_encoder.h.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
906
+ "text_encoder.h.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
907
+ "text_encoder.h.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
908
+ "text_encoder.h.10.attn.c_attn.bias": "model-00001-of-00002.safetensors",
909
+ "text_encoder.h.10.attn.c_attn.weight": "model-00001-of-00002.safetensors",
910
+ "text_encoder.h.10.attn.c_proj.bias": "model-00001-of-00002.safetensors",
911
+ "text_encoder.h.10.attn.c_proj.weight": "model-00001-of-00002.safetensors",
912
+ "text_encoder.h.10.ln_1.bias": "model-00001-of-00002.safetensors",
913
+ "text_encoder.h.10.ln_1.weight": "model-00001-of-00002.safetensors",
914
+ "text_encoder.h.10.ln_2.bias": "model-00001-of-00002.safetensors",
915
+ "text_encoder.h.10.ln_2.weight": "model-00001-of-00002.safetensors",
916
+ "text_encoder.h.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
917
+ "text_encoder.h.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
918
+ "text_encoder.h.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
919
+ "text_encoder.h.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
920
+ "text_encoder.h.11.attn.c_attn.bias": "model-00001-of-00002.safetensors",
921
+ "text_encoder.h.11.attn.c_attn.weight": "model-00001-of-00002.safetensors",
922
+ "text_encoder.h.11.attn.c_proj.bias": "model-00001-of-00002.safetensors",
923
+ "text_encoder.h.11.attn.c_proj.weight": "model-00001-of-00002.safetensors",
924
+ "text_encoder.h.11.ln_1.bias": "model-00001-of-00002.safetensors",
925
+ "text_encoder.h.11.ln_1.weight": "model-00001-of-00002.safetensors",
926
+ "text_encoder.h.11.ln_2.bias": "model-00001-of-00002.safetensors",
927
+ "text_encoder.h.11.ln_2.weight": "model-00001-of-00002.safetensors",
928
+ "text_encoder.h.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
929
+ "text_encoder.h.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
930
+ "text_encoder.h.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
931
+ "text_encoder.h.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
932
+ "text_encoder.h.2.attn.c_attn.bias": "model-00001-of-00002.safetensors",
933
+ "text_encoder.h.2.attn.c_attn.weight": "model-00001-of-00002.safetensors",
934
+ "text_encoder.h.2.attn.c_proj.bias": "model-00001-of-00002.safetensors",
935
+ "text_encoder.h.2.attn.c_proj.weight": "model-00001-of-00002.safetensors",
936
+ "text_encoder.h.2.ln_1.bias": "model-00001-of-00002.safetensors",
937
+ "text_encoder.h.2.ln_1.weight": "model-00001-of-00002.safetensors",
938
+ "text_encoder.h.2.ln_2.bias": "model-00001-of-00002.safetensors",
939
+ "text_encoder.h.2.ln_2.weight": "model-00001-of-00002.safetensors",
940
+ "text_encoder.h.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
941
+ "text_encoder.h.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
942
+ "text_encoder.h.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
943
+ "text_encoder.h.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
944
+ "text_encoder.h.3.attn.c_attn.bias": "model-00001-of-00002.safetensors",
945
+ "text_encoder.h.3.attn.c_attn.weight": "model-00001-of-00002.safetensors",
946
+ "text_encoder.h.3.attn.c_proj.bias": "model-00001-of-00002.safetensors",
947
+ "text_encoder.h.3.attn.c_proj.weight": "model-00001-of-00002.safetensors",
948
+ "text_encoder.h.3.ln_1.bias": "model-00001-of-00002.safetensors",
949
+ "text_encoder.h.3.ln_1.weight": "model-00001-of-00002.safetensors",
950
+ "text_encoder.h.3.ln_2.bias": "model-00001-of-00002.safetensors",
951
+ "text_encoder.h.3.ln_2.weight": "model-00001-of-00002.safetensors",
952
+ "text_encoder.h.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
953
+ "text_encoder.h.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
954
+ "text_encoder.h.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
955
+ "text_encoder.h.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
956
+ "text_encoder.h.4.attn.c_attn.bias": "model-00001-of-00002.safetensors",
957
+ "text_encoder.h.4.attn.c_attn.weight": "model-00001-of-00002.safetensors",
958
+ "text_encoder.h.4.attn.c_proj.bias": "model-00001-of-00002.safetensors",
959
+ "text_encoder.h.4.attn.c_proj.weight": "model-00001-of-00002.safetensors",
960
+ "text_encoder.h.4.ln_1.bias": "model-00001-of-00002.safetensors",
961
+ "text_encoder.h.4.ln_1.weight": "model-00001-of-00002.safetensors",
962
+ "text_encoder.h.4.ln_2.bias": "model-00001-of-00002.safetensors",
963
+ "text_encoder.h.4.ln_2.weight": "model-00001-of-00002.safetensors",
964
+ "text_encoder.h.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
965
+ "text_encoder.h.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
966
+ "text_encoder.h.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
967
+ "text_encoder.h.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
968
+ "text_encoder.h.5.attn.c_attn.bias": "model-00001-of-00002.safetensors",
969
+ "text_encoder.h.5.attn.c_attn.weight": "model-00001-of-00002.safetensors",
970
+ "text_encoder.h.5.attn.c_proj.bias": "model-00001-of-00002.safetensors",
971
+ "text_encoder.h.5.attn.c_proj.weight": "model-00001-of-00002.safetensors",
972
+ "text_encoder.h.5.ln_1.bias": "model-00001-of-00002.safetensors",
973
+ "text_encoder.h.5.ln_1.weight": "model-00001-of-00002.safetensors",
974
+ "text_encoder.h.5.ln_2.bias": "model-00001-of-00002.safetensors",
975
+ "text_encoder.h.5.ln_2.weight": "model-00001-of-00002.safetensors",
976
+ "text_encoder.h.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
977
+ "text_encoder.h.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
978
+ "text_encoder.h.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
979
+ "text_encoder.h.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
980
+ "text_encoder.h.6.attn.c_attn.bias": "model-00001-of-00002.safetensors",
981
+ "text_encoder.h.6.attn.c_attn.weight": "model-00001-of-00002.safetensors",
982
+ "text_encoder.h.6.attn.c_proj.bias": "model-00001-of-00002.safetensors",
983
+ "text_encoder.h.6.attn.c_proj.weight": "model-00001-of-00002.safetensors",
984
+ "text_encoder.h.6.ln_1.bias": "model-00001-of-00002.safetensors",
985
+ "text_encoder.h.6.ln_1.weight": "model-00001-of-00002.safetensors",
986
+ "text_encoder.h.6.ln_2.bias": "model-00001-of-00002.safetensors",
987
+ "text_encoder.h.6.ln_2.weight": "model-00001-of-00002.safetensors",
988
+ "text_encoder.h.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
989
+ "text_encoder.h.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
990
+ "text_encoder.h.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
991
+ "text_encoder.h.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
992
+ "text_encoder.h.7.attn.c_attn.bias": "model-00001-of-00002.safetensors",
993
+ "text_encoder.h.7.attn.c_attn.weight": "model-00001-of-00002.safetensors",
994
+ "text_encoder.h.7.attn.c_proj.bias": "model-00001-of-00002.safetensors",
995
+ "text_encoder.h.7.attn.c_proj.weight": "model-00001-of-00002.safetensors",
996
+ "text_encoder.h.7.ln_1.bias": "model-00001-of-00002.safetensors",
997
+ "text_encoder.h.7.ln_1.weight": "model-00001-of-00002.safetensors",
998
+ "text_encoder.h.7.ln_2.bias": "model-00001-of-00002.safetensors",
999
+ "text_encoder.h.7.ln_2.weight": "model-00001-of-00002.safetensors",
1000
+ "text_encoder.h.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1001
+ "text_encoder.h.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1002
+ "text_encoder.h.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1003
+ "text_encoder.h.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1004
+ "text_encoder.h.8.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1005
+ "text_encoder.h.8.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1006
+ "text_encoder.h.8.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1007
+ "text_encoder.h.8.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1008
+ "text_encoder.h.8.ln_1.bias": "model-00001-of-00002.safetensors",
1009
+ "text_encoder.h.8.ln_1.weight": "model-00001-of-00002.safetensors",
1010
+ "text_encoder.h.8.ln_2.bias": "model-00001-of-00002.safetensors",
1011
+ "text_encoder.h.8.ln_2.weight": "model-00001-of-00002.safetensors",
1012
+ "text_encoder.h.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1013
+ "text_encoder.h.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1014
+ "text_encoder.h.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1015
+ "text_encoder.h.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1016
+ "text_encoder.h.9.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1017
+ "text_encoder.h.9.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1018
+ "text_encoder.h.9.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1019
+ "text_encoder.h.9.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1020
+ "text_encoder.h.9.ln_1.bias": "model-00001-of-00002.safetensors",
1021
+ "text_encoder.h.9.ln_1.weight": "model-00001-of-00002.safetensors",
1022
+ "text_encoder.h.9.ln_2.bias": "model-00001-of-00002.safetensors",
1023
+ "text_encoder.h.9.ln_2.weight": "model-00001-of-00002.safetensors",
1024
+ "text_encoder.h.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1025
+ "text_encoder.h.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1026
+ "text_encoder.h.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1027
+ "text_encoder.h.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1028
+ "text_encoder.ln_f.bias": "model-00001-of-00002.safetensors",
1029
+ "text_encoder.ln_f.weight": "model-00001-of-00002.safetensors",
1030
+ "text_encoder.wpe.weight": "model-00001-of-00002.safetensors",
1031
+ "text_encoder.wte.weight": "model-00001-of-00002.safetensors",
1032
+ "text_encoder_proj.2.bias": "model-00002-of-00002.safetensors",
1033
+ "text_encoder_proj.2.weight": "model-00002-of-00002.safetensors",
1034
+ "type_classification_head.2.bias": "model-00002-of-00002.safetensors",
1035
+ "type_classification_head.2.weight": "model-00002-of-00002.safetensors"
1036
+ }
1037
+ }
phrase_tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
phrase_tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "<|endoftext|>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "<|endoftext|>",
17
+ "unk_token": {
18
+ "content": "<|endoftext|>",
19
+ "lstrip": false,
20
+ "normalized": true,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
phrase_tokenizer/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
phrase_tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1000000000000000019884624838656,
18
+ "pad_token": "<|endoftext|>",
19
+ "tokenizer_class": "GPT2Tokenizer",
20
+ "unk_token": "<|endoftext|>"
21
+ }
phrase_tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
text_tokenizer/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
text_tokenizer/special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "pad_token": "<|endoftext|>",
5
+ "unk_token": "<|endoftext|>"
6
+ }
text_tokenizer/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
text_tokenizer/tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "50256": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "bos_token": "<|endoftext|>",
14
+ "clean_up_tokenization_spaces": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "extra_special_tokens": {},
17
+ "model_max_length": 1000000000000000019884624838656,
18
+ "pad_token": "<|endoftext|>",
19
+ "tokenizer_class": "GPT2Tokenizer",
20
+ "unk_token": "<|endoftext|>"
21
+ }
text_tokenizer/vocab.json ADDED
The diff for this file is too large to render. See raw diff