Hugging Face's logo

tomvoelker
/
gpt22gpt2-gpt2-large-cnn-dailymail-seed42

Text2Text Generation
Transformers
TensorBoard
Safetensors
encoder-decoder
Generated from Trainer
Model card Files Metrics Community
tomvoelker commited on
Commit
b637d6e
·
verified ·
1 Parent(s): ff4ea0c

Training in progress, step 2000

Browse files
Files changed (11) hide show
  1. config.json +203 -0
  2. merges.txt +0 -0
  3. model-00001-of-00002.safetensors +3 -0
  4. model-00002-of-00002.safetensors +3 -0
  5. model.safetensors.index.json +1167 -0
  6. runs/Mar13_12-57-59_gpu09gmz.benchmarkcenter.megware.com/events.out.tfevents.1741870786.gpu09gmz.benchmarkcenter.megware.com.12123.0 +3 -0
  7. special_tokens_map.json +6 -0
  8. tokenizer.json +0 -0
  9. tokenizer_config.json +20 -0
  10. training_args.bin +3 -0
  11. vocab.json +0 -0
config.json ADDED
@@ -0,0 +1,203 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "EncoderDecoderModel"
4
+ ],
5
+ "decoder": {
6
+ "_name_or_path": "gpt2-large",
7
+ "activation_function": "gelu_new",
8
+ "add_cross_attention": true,
9
+ "architectures": [
10
+ "GPT2LMHeadModel"
11
+ ],
12
+ "attn_pdrop": 0.1,
13
+ "bad_words_ids": null,
14
+ "begin_suppress_tokens": null,
15
+ "bos_token_id": 50256,
16
+ "chunk_size_feed_forward": 0,
17
+ "cross_attention_hidden_size": null,
18
+ "decoder_start_token_id": null,
19
+ "diversity_penalty": 0.0,
20
+ "do_sample": false,
21
+ "early_stopping": false,
22
+ "embd_pdrop": 0.1,
23
+ "encoder_no_repeat_ngram_size": 0,
24
+ "eos_token_id": 50256,
25
+ "exponential_decay_length_penalty": null,
26
+ "finetuning_task": null,
27
+ "forced_bos_token_id": null,
28
+ "forced_eos_token_id": null,
29
+ "id2label": {
30
+ "0": "LABEL_0",
31
+ "1": "LABEL_1"
32
+ },
33
+ "initializer_range": 0.02,
34
+ "is_decoder": true,
35
+ "is_encoder_decoder": false,
36
+ "label2id": {
37
+ "LABEL_0": 0,
38
+ "LABEL_1": 1
39
+ },
40
+ "layer_norm_epsilon": 1e-05,
41
+ "length_penalty": 1.0,
42
+ "max_length": 20,
43
+ "min_length": 0,
44
+ "model_type": "gpt2",
45
+ "n_ctx": 1024,
46
+ "n_embd": 1280,
47
+ "n_head": 20,
48
+ "n_inner": null,
49
+ "n_layer": 36,
50
+ "n_positions": 1024,
51
+ "no_repeat_ngram_size": 0,
52
+ "num_beam_groups": 1,
53
+ "num_beams": 1,
54
+ "num_return_sequences": 1,
55
+ "output_attentions": false,
56
+ "output_hidden_states": false,
57
+ "output_scores": false,
58
+ "pad_token_id": null,
59
+ "prefix": null,
60
+ "problem_type": null,
61
+ "pruned_heads": {},
62
+ "remove_invalid_values": false,
63
+ "reorder_and_upcast_attn": false,
64
+ "repetition_penalty": 1.0,
65
+ "resid_pdrop": 0.1,
66
+ "return_dict": true,
67
+ "return_dict_in_generate": false,
68
+ "scale_attn_by_inverse_layer_idx": false,
69
+ "scale_attn_weights": true,
70
+ "sep_token_id": null,
71
+ "summary_activation": null,
72
+ "summary_first_dropout": 0.1,
73
+ "summary_proj_to_labels": true,
74
+ "summary_type": "cls_index",
75
+ "summary_use_proj": true,
76
+ "suppress_tokens": null,
77
+ "task_specific_params": {
78
+ "text-generation": {
79
+ "do_sample": true,
80
+ "max_length": 50
81
+ }
82
+ },
83
+ "temperature": 1.0,
84
+ "tf_legacy_loss": false,
85
+ "tie_encoder_decoder": false,
86
+ "tie_word_embeddings": true,
87
+ "tokenizer_class": null,
88
+ "top_k": 50,
89
+ "top_p": 1.0,
90
+ "torch_dtype": null,
91
+ "torchscript": false,
92
+ "typical_p": 1.0,
93
+ "use_bfloat16": false,
94
+ "use_cache": true,
95
+ "vocab_size": 50257
96
+ },
97
+ "decoder_start_token_id": 50256,
98
+ "early_stopping": true,
99
+ "encoder": {
100
+ "_name_or_path": "gpt2-large",
101
+ "activation_function": "gelu_new",
102
+ "add_cross_attention": false,
103
+ "architectures": [
104
+ "GPT2LMHeadModel"
105
+ ],
106
+ "attn_pdrop": 0.1,
107
+ "bad_words_ids": null,
108
+ "begin_suppress_tokens": null,
109
+ "bos_token_id": 50256,
110
+ "chunk_size_feed_forward": 0,
111
+ "cross_attention_hidden_size": null,
112
+ "decoder_start_token_id": null,
113
+ "diversity_penalty": 0.0,
114
+ "do_sample": false,
115
+ "early_stopping": false,
116
+ "embd_pdrop": 0.1,
117
+ "encoder_no_repeat_ngram_size": 0,
118
+ "eos_token_id": 50256,
119
+ "exponential_decay_length_penalty": null,
120
+ "finetuning_task": null,
121
+ "forced_bos_token_id": null,
122
+ "forced_eos_token_id": null,
123
+ "id2label": {
124
+ "0": "LABEL_0",
125
+ "1": "LABEL_1"
126
+ },
127
+ "initializer_range": 0.02,
128
+ "is_decoder": false,
129
+ "is_encoder_decoder": false,
130
+ "label2id": {
131
+ "LABEL_0": 0,
132
+ "LABEL_1": 1
133
+ },
134
+ "layer_norm_epsilon": 1e-05,
135
+ "length_penalty": 1.0,
136
+ "max_length": 20,
137
+ "min_length": 0,
138
+ "model_type": "gpt2",
139
+ "n_ctx": 1024,
140
+ "n_embd": 1280,
141
+ "n_head": 20,
142
+ "n_inner": null,
143
+ "n_layer": 36,
144
+ "n_positions": 1024,
145
+ "no_repeat_ngram_size": 0,
146
+ "num_beam_groups": 1,
147
+ "num_beams": 1,
148
+ "num_return_sequences": 1,
149
+ "output_attentions": false,
150
+ "output_hidden_states": false,
151
+ "output_scores": false,
152
+ "pad_token_id": null,
153
+ "prefix": null,
154
+ "problem_type": null,
155
+ "pruned_heads": {},
156
+ "remove_invalid_values": false,
157
+ "reorder_and_upcast_attn": false,
158
+ "repetition_penalty": 1.0,
159
+ "resid_pdrop": 0.1,
160
+ "return_dict": true,
161
+ "return_dict_in_generate": false,
162
+ "scale_attn_by_inverse_layer_idx": false,
163
+ "scale_attn_weights": true,
164
+ "sep_token_id": null,
165
+ "summary_activation": null,
166
+ "summary_first_dropout": 0.1,
167
+ "summary_proj_to_labels": true,
168
+ "summary_type": "cls_index",
169
+ "summary_use_proj": true,
170
+ "suppress_tokens": null,
171
+ "task_specific_params": {
172
+ "text-generation": {
173
+ "do_sample": true,
174
+ "max_length": 50
175
+ }
176
+ },
177
+ "temperature": 1.0,
178
+ "tf_legacy_loss": false,
179
+ "tie_encoder_decoder": false,
180
+ "tie_word_embeddings": true,
181
+ "tokenizer_class": null,
182
+ "top_k": 50,
183
+ "top_p": 1.0,
184
+ "torch_dtype": null,
185
+ "torchscript": false,
186
+ "typical_p": 1.0,
187
+ "use_bfloat16": false,
188
+ "use_cache": true,
189
+ "vocab_size": 50257
190
+ },
191
+ "eos_token_id": 50256,
192
+ "is_encoder_decoder": true,
193
+ "length_penalty": 2.0,
194
+ "max_length": 142,
195
+ "min_length": 56,
196
+ "model_type": "encoder-decoder",
197
+ "no_repeat_ngram_size": 3,
198
+ "num_beams": 4,
199
+ "pad_token_id": 50256,
200
+ "torch_dtype": "float32",
201
+ "transformers_version": "4.44.2",
202
+ "vocab_size": 50257
203
+ }
merges.txt 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:b9cb5ff175a6d841dba068337d32f4f439d68c6f2dc2815be5f3414ba02205d6
3
+ size 4985583856
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb5b2c0d9959aa63fac78e83bcd3357a427b9472f3518d5dfbbf3a14979af416
3
+ size 2151609800
model.safetensors.index.json ADDED
@@ -0,0 +1,1167 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 7137064960
4
+ },
5
+ "weight_map": {
6
+ "decoder.transformer.h.0.attn.c_attn.bias": "model-00001-of-00002.safetensors",
7
+ "decoder.transformer.h.0.attn.c_attn.weight": "model-00001-of-00002.safetensors",
8
+ "decoder.transformer.h.0.attn.c_proj.bias": "model-00001-of-00002.safetensors",
9
+ "decoder.transformer.h.0.attn.c_proj.weight": "model-00001-of-00002.safetensors",
10
+ "decoder.transformer.h.0.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
11
+ "decoder.transformer.h.0.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
12
+ "decoder.transformer.h.0.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
13
+ "decoder.transformer.h.0.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
14
+ "decoder.transformer.h.0.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
15
+ "decoder.transformer.h.0.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
16
+ "decoder.transformer.h.0.ln_1.bias": "model-00001-of-00002.safetensors",
17
+ "decoder.transformer.h.0.ln_1.weight": "model-00001-of-00002.safetensors",
18
+ "decoder.transformer.h.0.ln_2.bias": "model-00001-of-00002.safetensors",
19
+ "decoder.transformer.h.0.ln_2.weight": "model-00001-of-00002.safetensors",
20
+ "decoder.transformer.h.0.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
21
+ "decoder.transformer.h.0.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
22
+ "decoder.transformer.h.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
23
+ "decoder.transformer.h.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
24
+ "decoder.transformer.h.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
25
+ "decoder.transformer.h.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
26
+ "decoder.transformer.h.1.attn.c_attn.bias": "model-00001-of-00002.safetensors",
27
+ "decoder.transformer.h.1.attn.c_attn.weight": "model-00001-of-00002.safetensors",
28
+ "decoder.transformer.h.1.attn.c_proj.bias": "model-00001-of-00002.safetensors",
29
+ "decoder.transformer.h.1.attn.c_proj.weight": "model-00001-of-00002.safetensors",
30
+ "decoder.transformer.h.1.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
31
+ "decoder.transformer.h.1.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
32
+ "decoder.transformer.h.1.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
33
+ "decoder.transformer.h.1.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
34
+ "decoder.transformer.h.1.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
35
+ "decoder.transformer.h.1.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
36
+ "decoder.transformer.h.1.ln_1.bias": "model-00001-of-00002.safetensors",
37
+ "decoder.transformer.h.1.ln_1.weight": "model-00001-of-00002.safetensors",
38
+ "decoder.transformer.h.1.ln_2.bias": "model-00001-of-00002.safetensors",
39
+ "decoder.transformer.h.1.ln_2.weight": "model-00001-of-00002.safetensors",
40
+ "decoder.transformer.h.1.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
41
+ "decoder.transformer.h.1.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
42
+ "decoder.transformer.h.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
43
+ "decoder.transformer.h.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
44
+ "decoder.transformer.h.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
45
+ "decoder.transformer.h.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
46
+ "decoder.transformer.h.10.attn.c_attn.bias": "model-00001-of-00002.safetensors",
47
+ "decoder.transformer.h.10.attn.c_attn.weight": "model-00001-of-00002.safetensors",
48
+ "decoder.transformer.h.10.attn.c_proj.bias": "model-00001-of-00002.safetensors",
49
+ "decoder.transformer.h.10.attn.c_proj.weight": "model-00001-of-00002.safetensors",
50
+ "decoder.transformer.h.10.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
51
+ "decoder.transformer.h.10.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
52
+ "decoder.transformer.h.10.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
53
+ "decoder.transformer.h.10.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
54
+ "decoder.transformer.h.10.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
55
+ "decoder.transformer.h.10.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
56
+ "decoder.transformer.h.10.ln_1.bias": "model-00001-of-00002.safetensors",
57
+ "decoder.transformer.h.10.ln_1.weight": "model-00001-of-00002.safetensors",
58
+ "decoder.transformer.h.10.ln_2.bias": "model-00001-of-00002.safetensors",
59
+ "decoder.transformer.h.10.ln_2.weight": "model-00001-of-00002.safetensors",
60
+ "decoder.transformer.h.10.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
61
+ "decoder.transformer.h.10.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
62
+ "decoder.transformer.h.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
63
+ "decoder.transformer.h.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
64
+ "decoder.transformer.h.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
65
+ "decoder.transformer.h.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
66
+ "decoder.transformer.h.11.attn.c_attn.bias": "model-00001-of-00002.safetensors",
67
+ "decoder.transformer.h.11.attn.c_attn.weight": "model-00001-of-00002.safetensors",
68
+ "decoder.transformer.h.11.attn.c_proj.bias": "model-00001-of-00002.safetensors",
69
+ "decoder.transformer.h.11.attn.c_proj.weight": "model-00001-of-00002.safetensors",
70
+ "decoder.transformer.h.11.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
71
+ "decoder.transformer.h.11.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
72
+ "decoder.transformer.h.11.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
73
+ "decoder.transformer.h.11.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
74
+ "decoder.transformer.h.11.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
75
+ "decoder.transformer.h.11.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
76
+ "decoder.transformer.h.11.ln_1.bias": "model-00001-of-00002.safetensors",
77
+ "decoder.transformer.h.11.ln_1.weight": "model-00001-of-00002.safetensors",
78
+ "decoder.transformer.h.11.ln_2.bias": "model-00001-of-00002.safetensors",
79
+ "decoder.transformer.h.11.ln_2.weight": "model-00001-of-00002.safetensors",
80
+ "decoder.transformer.h.11.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
81
+ "decoder.transformer.h.11.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
82
+ "decoder.transformer.h.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
83
+ "decoder.transformer.h.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
84
+ "decoder.transformer.h.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
85
+ "decoder.transformer.h.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
86
+ "decoder.transformer.h.12.attn.c_attn.bias": "model-00001-of-00002.safetensors",
87
+ "decoder.transformer.h.12.attn.c_attn.weight": "model-00001-of-00002.safetensors",
88
+ "decoder.transformer.h.12.attn.c_proj.bias": "model-00001-of-00002.safetensors",
89
+ "decoder.transformer.h.12.attn.c_proj.weight": "model-00001-of-00002.safetensors",
90
+ "decoder.transformer.h.12.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
91
+ "decoder.transformer.h.12.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
92
+ "decoder.transformer.h.12.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
93
+ "decoder.transformer.h.12.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
94
+ "decoder.transformer.h.12.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
95
+ "decoder.transformer.h.12.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
96
+ "decoder.transformer.h.12.ln_1.bias": "model-00001-of-00002.safetensors",
97
+ "decoder.transformer.h.12.ln_1.weight": "model-00001-of-00002.safetensors",
98
+ "decoder.transformer.h.12.ln_2.bias": "model-00001-of-00002.safetensors",
99
+ "decoder.transformer.h.12.ln_2.weight": "model-00001-of-00002.safetensors",
100
+ "decoder.transformer.h.12.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
101
+ "decoder.transformer.h.12.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
102
+ "decoder.transformer.h.12.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
103
+ "decoder.transformer.h.12.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
104
+ "decoder.transformer.h.12.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
105
+ "decoder.transformer.h.12.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
106
+ "decoder.transformer.h.13.attn.c_attn.bias": "model-00001-of-00002.safetensors",
107
+ "decoder.transformer.h.13.attn.c_attn.weight": "model-00001-of-00002.safetensors",
108
+ "decoder.transformer.h.13.attn.c_proj.bias": "model-00001-of-00002.safetensors",
109
+ "decoder.transformer.h.13.attn.c_proj.weight": "model-00001-of-00002.safetensors",
110
+ "decoder.transformer.h.13.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
111
+ "decoder.transformer.h.13.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
112
+ "decoder.transformer.h.13.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
113
+ "decoder.transformer.h.13.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
114
+ "decoder.transformer.h.13.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
115
+ "decoder.transformer.h.13.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
116
+ "decoder.transformer.h.13.ln_1.bias": "model-00001-of-00002.safetensors",
117
+ "decoder.transformer.h.13.ln_1.weight": "model-00001-of-00002.safetensors",
118
+ "decoder.transformer.h.13.ln_2.bias": "model-00001-of-00002.safetensors",
119
+ "decoder.transformer.h.13.ln_2.weight": "model-00001-of-00002.safetensors",
120
+ "decoder.transformer.h.13.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
121
+ "decoder.transformer.h.13.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
122
+ "decoder.transformer.h.13.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
123
+ "decoder.transformer.h.13.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
124
+ "decoder.transformer.h.13.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
125
+ "decoder.transformer.h.13.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
126
+ "decoder.transformer.h.14.attn.c_attn.bias": "model-00001-of-00002.safetensors",
127
+ "decoder.transformer.h.14.attn.c_attn.weight": "model-00001-of-00002.safetensors",
128
+ "decoder.transformer.h.14.attn.c_proj.bias": "model-00001-of-00002.safetensors",
129
+ "decoder.transformer.h.14.attn.c_proj.weight": "model-00001-of-00002.safetensors",
130
+ "decoder.transformer.h.14.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
131
+ "decoder.transformer.h.14.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
132
+ "decoder.transformer.h.14.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
133
+ "decoder.transformer.h.14.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
134
+ "decoder.transformer.h.14.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
135
+ "decoder.transformer.h.14.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
136
+ "decoder.transformer.h.14.ln_1.bias": "model-00001-of-00002.safetensors",
137
+ "decoder.transformer.h.14.ln_1.weight": "model-00001-of-00002.safetensors",
138
+ "decoder.transformer.h.14.ln_2.bias": "model-00001-of-00002.safetensors",
139
+ "decoder.transformer.h.14.ln_2.weight": "model-00001-of-00002.safetensors",
140
+ "decoder.transformer.h.14.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
141
+ "decoder.transformer.h.14.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
142
+ "decoder.transformer.h.14.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
143
+ "decoder.transformer.h.14.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
144
+ "decoder.transformer.h.14.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
145
+ "decoder.transformer.h.14.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
146
+ "decoder.transformer.h.15.attn.c_attn.bias": "model-00001-of-00002.safetensors",
147
+ "decoder.transformer.h.15.attn.c_attn.weight": "model-00001-of-00002.safetensors",
148
+ "decoder.transformer.h.15.attn.c_proj.bias": "model-00001-of-00002.safetensors",
149
+ "decoder.transformer.h.15.attn.c_proj.weight": "model-00001-of-00002.safetensors",
150
+ "decoder.transformer.h.15.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
151
+ "decoder.transformer.h.15.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
152
+ "decoder.transformer.h.15.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
153
+ "decoder.transformer.h.15.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
154
+ "decoder.transformer.h.15.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
155
+ "decoder.transformer.h.15.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
156
+ "decoder.transformer.h.15.ln_1.bias": "model-00001-of-00002.safetensors",
157
+ "decoder.transformer.h.15.ln_1.weight": "model-00001-of-00002.safetensors",
158
+ "decoder.transformer.h.15.ln_2.bias": "model-00001-of-00002.safetensors",
159
+ "decoder.transformer.h.15.ln_2.weight": "model-00001-of-00002.safetensors",
160
+ "decoder.transformer.h.15.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
161
+ "decoder.transformer.h.15.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
162
+ "decoder.transformer.h.15.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
163
+ "decoder.transformer.h.15.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
164
+ "decoder.transformer.h.15.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
165
+ "decoder.transformer.h.15.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
166
+ "decoder.transformer.h.16.attn.c_attn.bias": "model-00002-of-00002.safetensors",
167
+ "decoder.transformer.h.16.attn.c_attn.weight": "model-00002-of-00002.safetensors",
168
+ "decoder.transformer.h.16.attn.c_proj.bias": "model-00002-of-00002.safetensors",
169
+ "decoder.transformer.h.16.attn.c_proj.weight": "model-00002-of-00002.safetensors",
170
+ "decoder.transformer.h.16.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
171
+ "decoder.transformer.h.16.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
172
+ "decoder.transformer.h.16.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
173
+ "decoder.transformer.h.16.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
174
+ "decoder.transformer.h.16.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
175
+ "decoder.transformer.h.16.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
176
+ "decoder.transformer.h.16.ln_1.bias": "model-00002-of-00002.safetensors",
177
+ "decoder.transformer.h.16.ln_1.weight": "model-00002-of-00002.safetensors",
178
+ "decoder.transformer.h.16.ln_2.bias": "model-00002-of-00002.safetensors",
179
+ "decoder.transformer.h.16.ln_2.weight": "model-00002-of-00002.safetensors",
180
+ "decoder.transformer.h.16.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
181
+ "decoder.transformer.h.16.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
182
+ "decoder.transformer.h.16.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
183
+ "decoder.transformer.h.16.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
184
+ "decoder.transformer.h.16.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
185
+ "decoder.transformer.h.16.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
186
+ "decoder.transformer.h.17.attn.c_attn.bias": "model-00002-of-00002.safetensors",
187
+ "decoder.transformer.h.17.attn.c_attn.weight": "model-00002-of-00002.safetensors",
188
+ "decoder.transformer.h.17.attn.c_proj.bias": "model-00002-of-00002.safetensors",
189
+ "decoder.transformer.h.17.attn.c_proj.weight": "model-00002-of-00002.safetensors",
190
+ "decoder.transformer.h.17.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
191
+ "decoder.transformer.h.17.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
192
+ "decoder.transformer.h.17.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
193
+ "decoder.transformer.h.17.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
194
+ "decoder.transformer.h.17.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
195
+ "decoder.transformer.h.17.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
196
+ "decoder.transformer.h.17.ln_1.bias": "model-00002-of-00002.safetensors",
197
+ "decoder.transformer.h.17.ln_1.weight": "model-00002-of-00002.safetensors",
198
+ "decoder.transformer.h.17.ln_2.bias": "model-00002-of-00002.safetensors",
199
+ "decoder.transformer.h.17.ln_2.weight": "model-00002-of-00002.safetensors",
200
+ "decoder.transformer.h.17.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
201
+ "decoder.transformer.h.17.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
202
+ "decoder.transformer.h.17.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
203
+ "decoder.transformer.h.17.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
204
+ "decoder.transformer.h.17.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
205
+ "decoder.transformer.h.17.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
206
+ "decoder.transformer.h.18.attn.c_attn.bias": "model-00002-of-00002.safetensors",
207
+ "decoder.transformer.h.18.attn.c_attn.weight": "model-00002-of-00002.safetensors",
208
+ "decoder.transformer.h.18.attn.c_proj.bias": "model-00002-of-00002.safetensors",
209
+ "decoder.transformer.h.18.attn.c_proj.weight": "model-00002-of-00002.safetensors",
210
+ "decoder.transformer.h.18.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
211
+ "decoder.transformer.h.18.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
212
+ "decoder.transformer.h.18.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
213
+ "decoder.transformer.h.18.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
214
+ "decoder.transformer.h.18.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
215
+ "decoder.transformer.h.18.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
216
+ "decoder.transformer.h.18.ln_1.bias": "model-00002-of-00002.safetensors",
217
+ "decoder.transformer.h.18.ln_1.weight": "model-00002-of-00002.safetensors",
218
+ "decoder.transformer.h.18.ln_2.bias": "model-00002-of-00002.safetensors",
219
+ "decoder.transformer.h.18.ln_2.weight": "model-00002-of-00002.safetensors",
220
+ "decoder.transformer.h.18.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
221
+ "decoder.transformer.h.18.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
222
+ "decoder.transformer.h.18.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
223
+ "decoder.transformer.h.18.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
224
+ "decoder.transformer.h.18.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
225
+ "decoder.transformer.h.18.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
226
+ "decoder.transformer.h.19.attn.c_attn.bias": "model-00002-of-00002.safetensors",
227
+ "decoder.transformer.h.19.attn.c_attn.weight": "model-00002-of-00002.safetensors",
228
+ "decoder.transformer.h.19.attn.c_proj.bias": "model-00002-of-00002.safetensors",
229
+ "decoder.transformer.h.19.attn.c_proj.weight": "model-00002-of-00002.safetensors",
230
+ "decoder.transformer.h.19.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
231
+ "decoder.transformer.h.19.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
232
+ "decoder.transformer.h.19.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
233
+ "decoder.transformer.h.19.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
234
+ "decoder.transformer.h.19.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
235
+ "decoder.transformer.h.19.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
236
+ "decoder.transformer.h.19.ln_1.bias": "model-00002-of-00002.safetensors",
237
+ "decoder.transformer.h.19.ln_1.weight": "model-00002-of-00002.safetensors",
238
+ "decoder.transformer.h.19.ln_2.bias": "model-00002-of-00002.safetensors",
239
+ "decoder.transformer.h.19.ln_2.weight": "model-00002-of-00002.safetensors",
240
+ "decoder.transformer.h.19.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
241
+ "decoder.transformer.h.19.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
242
+ "decoder.transformer.h.19.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
243
+ "decoder.transformer.h.19.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
244
+ "decoder.transformer.h.19.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
245
+ "decoder.transformer.h.19.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
246
+ "decoder.transformer.h.2.attn.c_attn.bias": "model-00001-of-00002.safetensors",
247
+ "decoder.transformer.h.2.attn.c_attn.weight": "model-00001-of-00002.safetensors",
248
+ "decoder.transformer.h.2.attn.c_proj.bias": "model-00001-of-00002.safetensors",
249
+ "decoder.transformer.h.2.attn.c_proj.weight": "model-00001-of-00002.safetensors",
250
+ "decoder.transformer.h.2.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
251
+ "decoder.transformer.h.2.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
252
+ "decoder.transformer.h.2.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
253
+ "decoder.transformer.h.2.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
254
+ "decoder.transformer.h.2.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
255
+ "decoder.transformer.h.2.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
256
+ "decoder.transformer.h.2.ln_1.bias": "model-00001-of-00002.safetensors",
257
+ "decoder.transformer.h.2.ln_1.weight": "model-00001-of-00002.safetensors",
258
+ "decoder.transformer.h.2.ln_2.bias": "model-00001-of-00002.safetensors",
259
+ "decoder.transformer.h.2.ln_2.weight": "model-00001-of-00002.safetensors",
260
+ "decoder.transformer.h.2.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
261
+ "decoder.transformer.h.2.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
262
+ "decoder.transformer.h.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
263
+ "decoder.transformer.h.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
264
+ "decoder.transformer.h.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
265
+ "decoder.transformer.h.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
266
+ "decoder.transformer.h.20.attn.c_attn.bias": "model-00002-of-00002.safetensors",
267
+ "decoder.transformer.h.20.attn.c_attn.weight": "model-00002-of-00002.safetensors",
268
+ "decoder.transformer.h.20.attn.c_proj.bias": "model-00002-of-00002.safetensors",
269
+ "decoder.transformer.h.20.attn.c_proj.weight": "model-00002-of-00002.safetensors",
270
+ "decoder.transformer.h.20.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
271
+ "decoder.transformer.h.20.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
272
+ "decoder.transformer.h.20.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
273
+ "decoder.transformer.h.20.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
274
+ "decoder.transformer.h.20.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
275
+ "decoder.transformer.h.20.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
276
+ "decoder.transformer.h.20.ln_1.bias": "model-00002-of-00002.safetensors",
277
+ "decoder.transformer.h.20.ln_1.weight": "model-00002-of-00002.safetensors",
278
+ "decoder.transformer.h.20.ln_2.bias": "model-00002-of-00002.safetensors",
279
+ "decoder.transformer.h.20.ln_2.weight": "model-00002-of-00002.safetensors",
280
+ "decoder.transformer.h.20.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
281
+ "decoder.transformer.h.20.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
282
+ "decoder.transformer.h.20.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
283
+ "decoder.transformer.h.20.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
284
+ "decoder.transformer.h.20.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
285
+ "decoder.transformer.h.20.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
286
+ "decoder.transformer.h.21.attn.c_attn.bias": "model-00002-of-00002.safetensors",
287
+ "decoder.transformer.h.21.attn.c_attn.weight": "model-00002-of-00002.safetensors",
288
+ "decoder.transformer.h.21.attn.c_proj.bias": "model-00002-of-00002.safetensors",
289
+ "decoder.transformer.h.21.attn.c_proj.weight": "model-00002-of-00002.safetensors",
290
+ "decoder.transformer.h.21.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
291
+ "decoder.transformer.h.21.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
292
+ "decoder.transformer.h.21.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
293
+ "decoder.transformer.h.21.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
294
+ "decoder.transformer.h.21.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
295
+ "decoder.transformer.h.21.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
296
+ "decoder.transformer.h.21.ln_1.bias": "model-00002-of-00002.safetensors",
297
+ "decoder.transformer.h.21.ln_1.weight": "model-00002-of-00002.safetensors",
298
+ "decoder.transformer.h.21.ln_2.bias": "model-00002-of-00002.safetensors",
299
+ "decoder.transformer.h.21.ln_2.weight": "model-00002-of-00002.safetensors",
300
+ "decoder.transformer.h.21.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
301
+ "decoder.transformer.h.21.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
302
+ "decoder.transformer.h.21.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
303
+ "decoder.transformer.h.21.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
304
+ "decoder.transformer.h.21.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
305
+ "decoder.transformer.h.21.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
306
+ "decoder.transformer.h.22.attn.c_attn.bias": "model-00002-of-00002.safetensors",
307
+ "decoder.transformer.h.22.attn.c_attn.weight": "model-00002-of-00002.safetensors",
308
+ "decoder.transformer.h.22.attn.c_proj.bias": "model-00002-of-00002.safetensors",
309
+ "decoder.transformer.h.22.attn.c_proj.weight": "model-00002-of-00002.safetensors",
310
+ "decoder.transformer.h.22.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
311
+ "decoder.transformer.h.22.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
312
+ "decoder.transformer.h.22.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
313
+ "decoder.transformer.h.22.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
314
+ "decoder.transformer.h.22.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
315
+ "decoder.transformer.h.22.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
316
+ "decoder.transformer.h.22.ln_1.bias": "model-00002-of-00002.safetensors",
317
+ "decoder.transformer.h.22.ln_1.weight": "model-00002-of-00002.safetensors",
318
+ "decoder.transformer.h.22.ln_2.bias": "model-00002-of-00002.safetensors",
319
+ "decoder.transformer.h.22.ln_2.weight": "model-00002-of-00002.safetensors",
320
+ "decoder.transformer.h.22.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
321
+ "decoder.transformer.h.22.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
322
+ "decoder.transformer.h.22.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
323
+ "decoder.transformer.h.22.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
324
+ "decoder.transformer.h.22.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
325
+ "decoder.transformer.h.22.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
326
+ "decoder.transformer.h.23.attn.c_attn.bias": "model-00002-of-00002.safetensors",
327
+ "decoder.transformer.h.23.attn.c_attn.weight": "model-00002-of-00002.safetensors",
328
+ "decoder.transformer.h.23.attn.c_proj.bias": "model-00002-of-00002.safetensors",
329
+ "decoder.transformer.h.23.attn.c_proj.weight": "model-00002-of-00002.safetensors",
330
+ "decoder.transformer.h.23.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
331
+ "decoder.transformer.h.23.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
332
+ "decoder.transformer.h.23.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
333
+ "decoder.transformer.h.23.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
334
+ "decoder.transformer.h.23.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
335
+ "decoder.transformer.h.23.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
336
+ "decoder.transformer.h.23.ln_1.bias": "model-00002-of-00002.safetensors",
337
+ "decoder.transformer.h.23.ln_1.weight": "model-00002-of-00002.safetensors",
338
+ "decoder.transformer.h.23.ln_2.bias": "model-00002-of-00002.safetensors",
339
+ "decoder.transformer.h.23.ln_2.weight": "model-00002-of-00002.safetensors",
340
+ "decoder.transformer.h.23.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
341
+ "decoder.transformer.h.23.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
342
+ "decoder.transformer.h.23.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
343
+ "decoder.transformer.h.23.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
344
+ "decoder.transformer.h.23.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
345
+ "decoder.transformer.h.23.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
346
+ "decoder.transformer.h.24.attn.c_attn.bias": "model-00002-of-00002.safetensors",
347
+ "decoder.transformer.h.24.attn.c_attn.weight": "model-00002-of-00002.safetensors",
348
+ "decoder.transformer.h.24.attn.c_proj.bias": "model-00002-of-00002.safetensors",
349
+ "decoder.transformer.h.24.attn.c_proj.weight": "model-00002-of-00002.safetensors",
350
+ "decoder.transformer.h.24.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
351
+ "decoder.transformer.h.24.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
352
+ "decoder.transformer.h.24.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
353
+ "decoder.transformer.h.24.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
354
+ "decoder.transformer.h.24.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
355
+ "decoder.transformer.h.24.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
356
+ "decoder.transformer.h.24.ln_1.bias": "model-00002-of-00002.safetensors",
357
+ "decoder.transformer.h.24.ln_1.weight": "model-00002-of-00002.safetensors",
358
+ "decoder.transformer.h.24.ln_2.bias": "model-00002-of-00002.safetensors",
359
+ "decoder.transformer.h.24.ln_2.weight": "model-00002-of-00002.safetensors",
360
+ "decoder.transformer.h.24.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
361
+ "decoder.transformer.h.24.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
362
+ "decoder.transformer.h.24.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
363
+ "decoder.transformer.h.24.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
364
+ "decoder.transformer.h.24.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
365
+ "decoder.transformer.h.24.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
366
+ "decoder.transformer.h.25.attn.c_attn.bias": "model-00002-of-00002.safetensors",
367
+ "decoder.transformer.h.25.attn.c_attn.weight": "model-00002-of-00002.safetensors",
368
+ "decoder.transformer.h.25.attn.c_proj.bias": "model-00002-of-00002.safetensors",
369
+ "decoder.transformer.h.25.attn.c_proj.weight": "model-00002-of-00002.safetensors",
370
+ "decoder.transformer.h.25.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
371
+ "decoder.transformer.h.25.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
372
+ "decoder.transformer.h.25.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
373
+ "decoder.transformer.h.25.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
374
+ "decoder.transformer.h.25.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
375
+ "decoder.transformer.h.25.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
376
+ "decoder.transformer.h.25.ln_1.bias": "model-00002-of-00002.safetensors",
377
+ "decoder.transformer.h.25.ln_1.weight": "model-00002-of-00002.safetensors",
378
+ "decoder.transformer.h.25.ln_2.bias": "model-00002-of-00002.safetensors",
379
+ "decoder.transformer.h.25.ln_2.weight": "model-00002-of-00002.safetensors",
380
+ "decoder.transformer.h.25.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
381
+ "decoder.transformer.h.25.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
382
+ "decoder.transformer.h.25.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
383
+ "decoder.transformer.h.25.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
384
+ "decoder.transformer.h.25.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
385
+ "decoder.transformer.h.25.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
386
+ "decoder.transformer.h.26.attn.c_attn.bias": "model-00002-of-00002.safetensors",
387
+ "decoder.transformer.h.26.attn.c_attn.weight": "model-00002-of-00002.safetensors",
388
+ "decoder.transformer.h.26.attn.c_proj.bias": "model-00002-of-00002.safetensors",
389
+ "decoder.transformer.h.26.attn.c_proj.weight": "model-00002-of-00002.safetensors",
390
+ "decoder.transformer.h.26.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
391
+ "decoder.transformer.h.26.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
392
+ "decoder.transformer.h.26.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
393
+ "decoder.transformer.h.26.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
394
+ "decoder.transformer.h.26.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
395
+ "decoder.transformer.h.26.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
396
+ "decoder.transformer.h.26.ln_1.bias": "model-00002-of-00002.safetensors",
397
+ "decoder.transformer.h.26.ln_1.weight": "model-00002-of-00002.safetensors",
398
+ "decoder.transformer.h.26.ln_2.bias": "model-00002-of-00002.safetensors",
399
+ "decoder.transformer.h.26.ln_2.weight": "model-00002-of-00002.safetensors",
400
+ "decoder.transformer.h.26.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
401
+ "decoder.transformer.h.26.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
402
+ "decoder.transformer.h.26.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
403
+ "decoder.transformer.h.26.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
404
+ "decoder.transformer.h.26.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
405
+ "decoder.transformer.h.26.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
406
+ "decoder.transformer.h.27.attn.c_attn.bias": "model-00002-of-00002.safetensors",
407
+ "decoder.transformer.h.27.attn.c_attn.weight": "model-00002-of-00002.safetensors",
408
+ "decoder.transformer.h.27.attn.c_proj.bias": "model-00002-of-00002.safetensors",
409
+ "decoder.transformer.h.27.attn.c_proj.weight": "model-00002-of-00002.safetensors",
410
+ "decoder.transformer.h.27.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
411
+ "decoder.transformer.h.27.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
412
+ "decoder.transformer.h.27.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
413
+ "decoder.transformer.h.27.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
414
+ "decoder.transformer.h.27.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
415
+ "decoder.transformer.h.27.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
416
+ "decoder.transformer.h.27.ln_1.bias": "model-00002-of-00002.safetensors",
417
+ "decoder.transformer.h.27.ln_1.weight": "model-00002-of-00002.safetensors",
418
+ "decoder.transformer.h.27.ln_2.bias": "model-00002-of-00002.safetensors",
419
+ "decoder.transformer.h.27.ln_2.weight": "model-00002-of-00002.safetensors",
420
+ "decoder.transformer.h.27.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
421
+ "decoder.transformer.h.27.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
422
+ "decoder.transformer.h.27.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
423
+ "decoder.transformer.h.27.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
424
+ "decoder.transformer.h.27.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
425
+ "decoder.transformer.h.27.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
426
+ "decoder.transformer.h.28.attn.c_attn.bias": "model-00002-of-00002.safetensors",
427
+ "decoder.transformer.h.28.attn.c_attn.weight": "model-00002-of-00002.safetensors",
428
+ "decoder.transformer.h.28.attn.c_proj.bias": "model-00002-of-00002.safetensors",
429
+ "decoder.transformer.h.28.attn.c_proj.weight": "model-00002-of-00002.safetensors",
430
+ "decoder.transformer.h.28.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
431
+ "decoder.transformer.h.28.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
432
+ "decoder.transformer.h.28.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
433
+ "decoder.transformer.h.28.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
434
+ "decoder.transformer.h.28.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
435
+ "decoder.transformer.h.28.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
436
+ "decoder.transformer.h.28.ln_1.bias": "model-00002-of-00002.safetensors",
437
+ "decoder.transformer.h.28.ln_1.weight": "model-00002-of-00002.safetensors",
438
+ "decoder.transformer.h.28.ln_2.bias": "model-00002-of-00002.safetensors",
439
+ "decoder.transformer.h.28.ln_2.weight": "model-00002-of-00002.safetensors",
440
+ "decoder.transformer.h.28.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
441
+ "decoder.transformer.h.28.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
442
+ "decoder.transformer.h.28.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
443
+ "decoder.transformer.h.28.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
444
+ "decoder.transformer.h.28.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
445
+ "decoder.transformer.h.28.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
446
+ "decoder.transformer.h.29.attn.c_attn.bias": "model-00002-of-00002.safetensors",
447
+ "decoder.transformer.h.29.attn.c_attn.weight": "model-00002-of-00002.safetensors",
448
+ "decoder.transformer.h.29.attn.c_proj.bias": "model-00002-of-00002.safetensors",
449
+ "decoder.transformer.h.29.attn.c_proj.weight": "model-00002-of-00002.safetensors",
450
+ "decoder.transformer.h.29.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
451
+ "decoder.transformer.h.29.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
452
+ "decoder.transformer.h.29.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
453
+ "decoder.transformer.h.29.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
454
+ "decoder.transformer.h.29.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
455
+ "decoder.transformer.h.29.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
456
+ "decoder.transformer.h.29.ln_1.bias": "model-00002-of-00002.safetensors",
457
+ "decoder.transformer.h.29.ln_1.weight": "model-00002-of-00002.safetensors",
458
+ "decoder.transformer.h.29.ln_2.bias": "model-00002-of-00002.safetensors",
459
+ "decoder.transformer.h.29.ln_2.weight": "model-00002-of-00002.safetensors",
460
+ "decoder.transformer.h.29.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
461
+ "decoder.transformer.h.29.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
462
+ "decoder.transformer.h.29.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
463
+ "decoder.transformer.h.29.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
464
+ "decoder.transformer.h.29.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
465
+ "decoder.transformer.h.29.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
466
+ "decoder.transformer.h.3.attn.c_attn.bias": "model-00001-of-00002.safetensors",
467
+ "decoder.transformer.h.3.attn.c_attn.weight": "model-00001-of-00002.safetensors",
468
+ "decoder.transformer.h.3.attn.c_proj.bias": "model-00001-of-00002.safetensors",
469
+ "decoder.transformer.h.3.attn.c_proj.weight": "model-00001-of-00002.safetensors",
470
+ "decoder.transformer.h.3.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
471
+ "decoder.transformer.h.3.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
472
+ "decoder.transformer.h.3.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
473
+ "decoder.transformer.h.3.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
474
+ "decoder.transformer.h.3.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
475
+ "decoder.transformer.h.3.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
476
+ "decoder.transformer.h.3.ln_1.bias": "model-00001-of-00002.safetensors",
477
+ "decoder.transformer.h.3.ln_1.weight": "model-00001-of-00002.safetensors",
478
+ "decoder.transformer.h.3.ln_2.bias": "model-00001-of-00002.safetensors",
479
+ "decoder.transformer.h.3.ln_2.weight": "model-00001-of-00002.safetensors",
480
+ "decoder.transformer.h.3.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
481
+ "decoder.transformer.h.3.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
482
+ "decoder.transformer.h.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
483
+ "decoder.transformer.h.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
484
+ "decoder.transformer.h.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
485
+ "decoder.transformer.h.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
486
+ "decoder.transformer.h.30.attn.c_attn.bias": "model-00002-of-00002.safetensors",
487
+ "decoder.transformer.h.30.attn.c_attn.weight": "model-00002-of-00002.safetensors",
488
+ "decoder.transformer.h.30.attn.c_proj.bias": "model-00002-of-00002.safetensors",
489
+ "decoder.transformer.h.30.attn.c_proj.weight": "model-00002-of-00002.safetensors",
490
+ "decoder.transformer.h.30.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
491
+ "decoder.transformer.h.30.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
492
+ "decoder.transformer.h.30.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
493
+ "decoder.transformer.h.30.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
494
+ "decoder.transformer.h.30.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
495
+ "decoder.transformer.h.30.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
496
+ "decoder.transformer.h.30.ln_1.bias": "model-00002-of-00002.safetensors",
497
+ "decoder.transformer.h.30.ln_1.weight": "model-00002-of-00002.safetensors",
498
+ "decoder.transformer.h.30.ln_2.bias": "model-00002-of-00002.safetensors",
499
+ "decoder.transformer.h.30.ln_2.weight": "model-00002-of-00002.safetensors",
500
+ "decoder.transformer.h.30.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
501
+ "decoder.transformer.h.30.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
502
+ "decoder.transformer.h.30.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
503
+ "decoder.transformer.h.30.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
504
+ "decoder.transformer.h.30.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
505
+ "decoder.transformer.h.30.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
506
+ "decoder.transformer.h.31.attn.c_attn.bias": "model-00002-of-00002.safetensors",
507
+ "decoder.transformer.h.31.attn.c_attn.weight": "model-00002-of-00002.safetensors",
508
+ "decoder.transformer.h.31.attn.c_proj.bias": "model-00002-of-00002.safetensors",
509
+ "decoder.transformer.h.31.attn.c_proj.weight": "model-00002-of-00002.safetensors",
510
+ "decoder.transformer.h.31.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
511
+ "decoder.transformer.h.31.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
512
+ "decoder.transformer.h.31.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
513
+ "decoder.transformer.h.31.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
514
+ "decoder.transformer.h.31.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
515
+ "decoder.transformer.h.31.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
516
+ "decoder.transformer.h.31.ln_1.bias": "model-00002-of-00002.safetensors",
517
+ "decoder.transformer.h.31.ln_1.weight": "model-00002-of-00002.safetensors",
518
+ "decoder.transformer.h.31.ln_2.bias": "model-00002-of-00002.safetensors",
519
+ "decoder.transformer.h.31.ln_2.weight": "model-00002-of-00002.safetensors",
520
+ "decoder.transformer.h.31.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
521
+ "decoder.transformer.h.31.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
522
+ "decoder.transformer.h.31.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
523
+ "decoder.transformer.h.31.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
524
+ "decoder.transformer.h.31.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
525
+ "decoder.transformer.h.31.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
526
+ "decoder.transformer.h.32.attn.c_attn.bias": "model-00002-of-00002.safetensors",
527
+ "decoder.transformer.h.32.attn.c_attn.weight": "model-00002-of-00002.safetensors",
528
+ "decoder.transformer.h.32.attn.c_proj.bias": "model-00002-of-00002.safetensors",
529
+ "decoder.transformer.h.32.attn.c_proj.weight": "model-00002-of-00002.safetensors",
530
+ "decoder.transformer.h.32.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
531
+ "decoder.transformer.h.32.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
532
+ "decoder.transformer.h.32.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
533
+ "decoder.transformer.h.32.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
534
+ "decoder.transformer.h.32.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
535
+ "decoder.transformer.h.32.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
536
+ "decoder.transformer.h.32.ln_1.bias": "model-00002-of-00002.safetensors",
537
+ "decoder.transformer.h.32.ln_1.weight": "model-00002-of-00002.safetensors",
538
+ "decoder.transformer.h.32.ln_2.bias": "model-00002-of-00002.safetensors",
539
+ "decoder.transformer.h.32.ln_2.weight": "model-00002-of-00002.safetensors",
540
+ "decoder.transformer.h.32.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
541
+ "decoder.transformer.h.32.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
542
+ "decoder.transformer.h.32.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
543
+ "decoder.transformer.h.32.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
544
+ "decoder.transformer.h.32.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
545
+ "decoder.transformer.h.32.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
546
+ "decoder.transformer.h.33.attn.c_attn.bias": "model-00002-of-00002.safetensors",
547
+ "decoder.transformer.h.33.attn.c_attn.weight": "model-00002-of-00002.safetensors",
548
+ "decoder.transformer.h.33.attn.c_proj.bias": "model-00002-of-00002.safetensors",
549
+ "decoder.transformer.h.33.attn.c_proj.weight": "model-00002-of-00002.safetensors",
550
+ "decoder.transformer.h.33.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
551
+ "decoder.transformer.h.33.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
552
+ "decoder.transformer.h.33.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
553
+ "decoder.transformer.h.33.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
554
+ "decoder.transformer.h.33.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
555
+ "decoder.transformer.h.33.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
556
+ "decoder.transformer.h.33.ln_1.bias": "model-00002-of-00002.safetensors",
557
+ "decoder.transformer.h.33.ln_1.weight": "model-00002-of-00002.safetensors",
558
+ "decoder.transformer.h.33.ln_2.bias": "model-00002-of-00002.safetensors",
559
+ "decoder.transformer.h.33.ln_2.weight": "model-00002-of-00002.safetensors",
560
+ "decoder.transformer.h.33.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
561
+ "decoder.transformer.h.33.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
562
+ "decoder.transformer.h.33.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
563
+ "decoder.transformer.h.33.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
564
+ "decoder.transformer.h.33.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
565
+ "decoder.transformer.h.33.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
566
+ "decoder.transformer.h.34.attn.c_attn.bias": "model-00002-of-00002.safetensors",
567
+ "decoder.transformer.h.34.attn.c_attn.weight": "model-00002-of-00002.safetensors",
568
+ "decoder.transformer.h.34.attn.c_proj.bias": "model-00002-of-00002.safetensors",
569
+ "decoder.transformer.h.34.attn.c_proj.weight": "model-00002-of-00002.safetensors",
570
+ "decoder.transformer.h.34.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
571
+ "decoder.transformer.h.34.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
572
+ "decoder.transformer.h.34.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
573
+ "decoder.transformer.h.34.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
574
+ "decoder.transformer.h.34.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
575
+ "decoder.transformer.h.34.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
576
+ "decoder.transformer.h.34.ln_1.bias": "model-00002-of-00002.safetensors",
577
+ "decoder.transformer.h.34.ln_1.weight": "model-00002-of-00002.safetensors",
578
+ "decoder.transformer.h.34.ln_2.bias": "model-00002-of-00002.safetensors",
579
+ "decoder.transformer.h.34.ln_2.weight": "model-00002-of-00002.safetensors",
580
+ "decoder.transformer.h.34.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
581
+ "decoder.transformer.h.34.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
582
+ "decoder.transformer.h.34.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
583
+ "decoder.transformer.h.34.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
584
+ "decoder.transformer.h.34.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
585
+ "decoder.transformer.h.34.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
586
+ "decoder.transformer.h.35.attn.c_attn.bias": "model-00002-of-00002.safetensors",
587
+ "decoder.transformer.h.35.attn.c_attn.weight": "model-00002-of-00002.safetensors",
588
+ "decoder.transformer.h.35.attn.c_proj.bias": "model-00002-of-00002.safetensors",
589
+ "decoder.transformer.h.35.attn.c_proj.weight": "model-00002-of-00002.safetensors",
590
+ "decoder.transformer.h.35.crossattention.c_attn.bias": "model-00002-of-00002.safetensors",
591
+ "decoder.transformer.h.35.crossattention.c_attn.weight": "model-00002-of-00002.safetensors",
592
+ "decoder.transformer.h.35.crossattention.c_proj.bias": "model-00002-of-00002.safetensors",
593
+ "decoder.transformer.h.35.crossattention.c_proj.weight": "model-00002-of-00002.safetensors",
594
+ "decoder.transformer.h.35.crossattention.q_attn.bias": "model-00002-of-00002.safetensors",
595
+ "decoder.transformer.h.35.crossattention.q_attn.weight": "model-00002-of-00002.safetensors",
596
+ "decoder.transformer.h.35.ln_1.bias": "model-00002-of-00002.safetensors",
597
+ "decoder.transformer.h.35.ln_1.weight": "model-00002-of-00002.safetensors",
598
+ "decoder.transformer.h.35.ln_2.bias": "model-00002-of-00002.safetensors",
599
+ "decoder.transformer.h.35.ln_2.weight": "model-00002-of-00002.safetensors",
600
+ "decoder.transformer.h.35.ln_cross_attn.bias": "model-00002-of-00002.safetensors",
601
+ "decoder.transformer.h.35.ln_cross_attn.weight": "model-00002-of-00002.safetensors",
602
+ "decoder.transformer.h.35.mlp.c_fc.bias": "model-00002-of-00002.safetensors",
603
+ "decoder.transformer.h.35.mlp.c_fc.weight": "model-00002-of-00002.safetensors",
604
+ "decoder.transformer.h.35.mlp.c_proj.bias": "model-00002-of-00002.safetensors",
605
+ "decoder.transformer.h.35.mlp.c_proj.weight": "model-00002-of-00002.safetensors",
606
+ "decoder.transformer.h.4.attn.c_attn.bias": "model-00001-of-00002.safetensors",
607
+ "decoder.transformer.h.4.attn.c_attn.weight": "model-00001-of-00002.safetensors",
608
+ "decoder.transformer.h.4.attn.c_proj.bias": "model-00001-of-00002.safetensors",
609
+ "decoder.transformer.h.4.attn.c_proj.weight": "model-00001-of-00002.safetensors",
610
+ "decoder.transformer.h.4.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
611
+ "decoder.transformer.h.4.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
612
+ "decoder.transformer.h.4.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
613
+ "decoder.transformer.h.4.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
614
+ "decoder.transformer.h.4.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
615
+ "decoder.transformer.h.4.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
616
+ "decoder.transformer.h.4.ln_1.bias": "model-00001-of-00002.safetensors",
617
+ "decoder.transformer.h.4.ln_1.weight": "model-00001-of-00002.safetensors",
618
+ "decoder.transformer.h.4.ln_2.bias": "model-00001-of-00002.safetensors",
619
+ "decoder.transformer.h.4.ln_2.weight": "model-00001-of-00002.safetensors",
620
+ "decoder.transformer.h.4.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
621
+ "decoder.transformer.h.4.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
622
+ "decoder.transformer.h.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
623
+ "decoder.transformer.h.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
624
+ "decoder.transformer.h.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
625
+ "decoder.transformer.h.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
626
+ "decoder.transformer.h.5.attn.c_attn.bias": "model-00001-of-00002.safetensors",
627
+ "decoder.transformer.h.5.attn.c_attn.weight": "model-00001-of-00002.safetensors",
628
+ "decoder.transformer.h.5.attn.c_proj.bias": "model-00001-of-00002.safetensors",
629
+ "decoder.transformer.h.5.attn.c_proj.weight": "model-00001-of-00002.safetensors",
630
+ "decoder.transformer.h.5.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
631
+ "decoder.transformer.h.5.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
632
+ "decoder.transformer.h.5.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
633
+ "decoder.transformer.h.5.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
634
+ "decoder.transformer.h.5.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
635
+ "decoder.transformer.h.5.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
636
+ "decoder.transformer.h.5.ln_1.bias": "model-00001-of-00002.safetensors",
637
+ "decoder.transformer.h.5.ln_1.weight": "model-00001-of-00002.safetensors",
638
+ "decoder.transformer.h.5.ln_2.bias": "model-00001-of-00002.safetensors",
639
+ "decoder.transformer.h.5.ln_2.weight": "model-00001-of-00002.safetensors",
640
+ "decoder.transformer.h.5.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
641
+ "decoder.transformer.h.5.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
642
+ "decoder.transformer.h.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
643
+ "decoder.transformer.h.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
644
+ "decoder.transformer.h.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
645
+ "decoder.transformer.h.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
646
+ "decoder.transformer.h.6.attn.c_attn.bias": "model-00001-of-00002.safetensors",
647
+ "decoder.transformer.h.6.attn.c_attn.weight": "model-00001-of-00002.safetensors",
648
+ "decoder.transformer.h.6.attn.c_proj.bias": "model-00001-of-00002.safetensors",
649
+ "decoder.transformer.h.6.attn.c_proj.weight": "model-00001-of-00002.safetensors",
650
+ "decoder.transformer.h.6.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
651
+ "decoder.transformer.h.6.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
652
+ "decoder.transformer.h.6.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
653
+ "decoder.transformer.h.6.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
654
+ "decoder.transformer.h.6.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
655
+ "decoder.transformer.h.6.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
656
+ "decoder.transformer.h.6.ln_1.bias": "model-00001-of-00002.safetensors",
657
+ "decoder.transformer.h.6.ln_1.weight": "model-00001-of-00002.safetensors",
658
+ "decoder.transformer.h.6.ln_2.bias": "model-00001-of-00002.safetensors",
659
+ "decoder.transformer.h.6.ln_2.weight": "model-00001-of-00002.safetensors",
660
+ "decoder.transformer.h.6.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
661
+ "decoder.transformer.h.6.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
662
+ "decoder.transformer.h.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
663
+ "decoder.transformer.h.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
664
+ "decoder.transformer.h.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
665
+ "decoder.transformer.h.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
666
+ "decoder.transformer.h.7.attn.c_attn.bias": "model-00001-of-00002.safetensors",
667
+ "decoder.transformer.h.7.attn.c_attn.weight": "model-00001-of-00002.safetensors",
668
+ "decoder.transformer.h.7.attn.c_proj.bias": "model-00001-of-00002.safetensors",
669
+ "decoder.transformer.h.7.attn.c_proj.weight": "model-00001-of-00002.safetensors",
670
+ "decoder.transformer.h.7.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
671
+ "decoder.transformer.h.7.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
672
+ "decoder.transformer.h.7.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
673
+ "decoder.transformer.h.7.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
674
+ "decoder.transformer.h.7.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
675
+ "decoder.transformer.h.7.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
676
+ "decoder.transformer.h.7.ln_1.bias": "model-00001-of-00002.safetensors",
677
+ "decoder.transformer.h.7.ln_1.weight": "model-00001-of-00002.safetensors",
678
+ "decoder.transformer.h.7.ln_2.bias": "model-00001-of-00002.safetensors",
679
+ "decoder.transformer.h.7.ln_2.weight": "model-00001-of-00002.safetensors",
680
+ "decoder.transformer.h.7.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
681
+ "decoder.transformer.h.7.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
682
+ "decoder.transformer.h.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
683
+ "decoder.transformer.h.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
684
+ "decoder.transformer.h.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
685
+ "decoder.transformer.h.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
686
+ "decoder.transformer.h.8.attn.c_attn.bias": "model-00001-of-00002.safetensors",
687
+ "decoder.transformer.h.8.attn.c_attn.weight": "model-00001-of-00002.safetensors",
688
+ "decoder.transformer.h.8.attn.c_proj.bias": "model-00001-of-00002.safetensors",
689
+ "decoder.transformer.h.8.attn.c_proj.weight": "model-00001-of-00002.safetensors",
690
+ "decoder.transformer.h.8.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
691
+ "decoder.transformer.h.8.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
692
+ "decoder.transformer.h.8.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
693
+ "decoder.transformer.h.8.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
694
+ "decoder.transformer.h.8.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
695
+ "decoder.transformer.h.8.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
696
+ "decoder.transformer.h.8.ln_1.bias": "model-00001-of-00002.safetensors",
697
+ "decoder.transformer.h.8.ln_1.weight": "model-00001-of-00002.safetensors",
698
+ "decoder.transformer.h.8.ln_2.bias": "model-00001-of-00002.safetensors",
699
+ "decoder.transformer.h.8.ln_2.weight": "model-00001-of-00002.safetensors",
700
+ "decoder.transformer.h.8.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
701
+ "decoder.transformer.h.8.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
702
+ "decoder.transformer.h.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
703
+ "decoder.transformer.h.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
704
+ "decoder.transformer.h.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
705
+ "decoder.transformer.h.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
706
+ "decoder.transformer.h.9.attn.c_attn.bias": "model-00001-of-00002.safetensors",
707
+ "decoder.transformer.h.9.attn.c_attn.weight": "model-00001-of-00002.safetensors",
708
+ "decoder.transformer.h.9.attn.c_proj.bias": "model-00001-of-00002.safetensors",
709
+ "decoder.transformer.h.9.attn.c_proj.weight": "model-00001-of-00002.safetensors",
710
+ "decoder.transformer.h.9.crossattention.c_attn.bias": "model-00001-of-00002.safetensors",
711
+ "decoder.transformer.h.9.crossattention.c_attn.weight": "model-00001-of-00002.safetensors",
712
+ "decoder.transformer.h.9.crossattention.c_proj.bias": "model-00001-of-00002.safetensors",
713
+ "decoder.transformer.h.9.crossattention.c_proj.weight": "model-00001-of-00002.safetensors",
714
+ "decoder.transformer.h.9.crossattention.q_attn.bias": "model-00001-of-00002.safetensors",
715
+ "decoder.transformer.h.9.crossattention.q_attn.weight": "model-00001-of-00002.safetensors",
716
+ "decoder.transformer.h.9.ln_1.bias": "model-00001-of-00002.safetensors",
717
+ "decoder.transformer.h.9.ln_1.weight": "model-00001-of-00002.safetensors",
718
+ "decoder.transformer.h.9.ln_2.bias": "model-00001-of-00002.safetensors",
719
+ "decoder.transformer.h.9.ln_2.weight": "model-00001-of-00002.safetensors",
720
+ "decoder.transformer.h.9.ln_cross_attn.bias": "model-00001-of-00002.safetensors",
721
+ "decoder.transformer.h.9.ln_cross_attn.weight": "model-00001-of-00002.safetensors",
722
+ "decoder.transformer.h.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
723
+ "decoder.transformer.h.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
724
+ "decoder.transformer.h.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
725
+ "decoder.transformer.h.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
726
+ "decoder.transformer.ln_f.bias": "model-00002-of-00002.safetensors",
727
+ "decoder.transformer.ln_f.weight": "model-00002-of-00002.safetensors",
728
+ "decoder.transformer.wpe.weight": "model-00001-of-00002.safetensors",
729
+ "decoder.transformer.wte.weight": "model-00001-of-00002.safetensors",
730
+ "encoder.h.0.attn.c_attn.bias": "model-00001-of-00002.safetensors",
731
+ "encoder.h.0.attn.c_attn.weight": "model-00001-of-00002.safetensors",
732
+ "encoder.h.0.attn.c_proj.bias": "model-00001-of-00002.safetensors",
733
+ "encoder.h.0.attn.c_proj.weight": "model-00001-of-00002.safetensors",
734
+ "encoder.h.0.ln_1.bias": "model-00001-of-00002.safetensors",
735
+ "encoder.h.0.ln_1.weight": "model-00001-of-00002.safetensors",
736
+ "encoder.h.0.ln_2.bias": "model-00001-of-00002.safetensors",
737
+ "encoder.h.0.ln_2.weight": "model-00001-of-00002.safetensors",
738
+ "encoder.h.0.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
739
+ "encoder.h.0.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
740
+ "encoder.h.0.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
741
+ "encoder.h.0.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
742
+ "encoder.h.1.attn.c_attn.bias": "model-00001-of-00002.safetensors",
743
+ "encoder.h.1.attn.c_attn.weight": "model-00001-of-00002.safetensors",
744
+ "encoder.h.1.attn.c_proj.bias": "model-00001-of-00002.safetensors",
745
+ "encoder.h.1.attn.c_proj.weight": "model-00001-of-00002.safetensors",
746
+ "encoder.h.1.ln_1.bias": "model-00001-of-00002.safetensors",
747
+ "encoder.h.1.ln_1.weight": "model-00001-of-00002.safetensors",
748
+ "encoder.h.1.ln_2.bias": "model-00001-of-00002.safetensors",
749
+ "encoder.h.1.ln_2.weight": "model-00001-of-00002.safetensors",
750
+ "encoder.h.1.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
751
+ "encoder.h.1.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
752
+ "encoder.h.1.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
753
+ "encoder.h.1.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
754
+ "encoder.h.10.attn.c_attn.bias": "model-00001-of-00002.safetensors",
755
+ "encoder.h.10.attn.c_attn.weight": "model-00001-of-00002.safetensors",
756
+ "encoder.h.10.attn.c_proj.bias": "model-00001-of-00002.safetensors",
757
+ "encoder.h.10.attn.c_proj.weight": "model-00001-of-00002.safetensors",
758
+ "encoder.h.10.ln_1.bias": "model-00001-of-00002.safetensors",
759
+ "encoder.h.10.ln_1.weight": "model-00001-of-00002.safetensors",
760
+ "encoder.h.10.ln_2.bias": "model-00001-of-00002.safetensors",
761
+ "encoder.h.10.ln_2.weight": "model-00001-of-00002.safetensors",
762
+ "encoder.h.10.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
763
+ "encoder.h.10.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
764
+ "encoder.h.10.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
765
+ "encoder.h.10.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
766
+ "encoder.h.11.attn.c_attn.bias": "model-00001-of-00002.safetensors",
767
+ "encoder.h.11.attn.c_attn.weight": "model-00001-of-00002.safetensors",
768
+ "encoder.h.11.attn.c_proj.bias": "model-00001-of-00002.safetensors",
769
+ "encoder.h.11.attn.c_proj.weight": "model-00001-of-00002.safetensors",
770
+ "encoder.h.11.ln_1.bias": "model-00001-of-00002.safetensors",
771
+ "encoder.h.11.ln_1.weight": "model-00001-of-00002.safetensors",
772
+ "encoder.h.11.ln_2.bias": "model-00001-of-00002.safetensors",
773
+ "encoder.h.11.ln_2.weight": "model-00001-of-00002.safetensors",
774
+ "encoder.h.11.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
775
+ "encoder.h.11.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
776
+ "encoder.h.11.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
777
+ "encoder.h.11.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
778
+ "encoder.h.12.attn.c_attn.bias": "model-00001-of-00002.safetensors",
779
+ "encoder.h.12.attn.c_attn.weight": "model-00001-of-00002.safetensors",
780
+ "encoder.h.12.attn.c_proj.bias": "model-00001-of-00002.safetensors",
781
+ "encoder.h.12.attn.c_proj.weight": "model-00001-of-00002.safetensors",
782
+ "encoder.h.12.ln_1.bias": "model-00001-of-00002.safetensors",
783
+ "encoder.h.12.ln_1.weight": "model-00001-of-00002.safetensors",
784
+ "encoder.h.12.ln_2.bias": "model-00001-of-00002.safetensors",
785
+ "encoder.h.12.ln_2.weight": "model-00001-of-00002.safetensors",
786
+ "encoder.h.12.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
787
+ "encoder.h.12.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
788
+ "encoder.h.12.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
789
+ "encoder.h.12.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
790
+ "encoder.h.13.attn.c_attn.bias": "model-00001-of-00002.safetensors",
791
+ "encoder.h.13.attn.c_attn.weight": "model-00001-of-00002.safetensors",
792
+ "encoder.h.13.attn.c_proj.bias": "model-00001-of-00002.safetensors",
793
+ "encoder.h.13.attn.c_proj.weight": "model-00001-of-00002.safetensors",
794
+ "encoder.h.13.ln_1.bias": "model-00001-of-00002.safetensors",
795
+ "encoder.h.13.ln_1.weight": "model-00001-of-00002.safetensors",
796
+ "encoder.h.13.ln_2.bias": "model-00001-of-00002.safetensors",
797
+ "encoder.h.13.ln_2.weight": "model-00001-of-00002.safetensors",
798
+ "encoder.h.13.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
799
+ "encoder.h.13.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
800
+ "encoder.h.13.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
801
+ "encoder.h.13.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
802
+ "encoder.h.14.attn.c_attn.bias": "model-00001-of-00002.safetensors",
803
+ "encoder.h.14.attn.c_attn.weight": "model-00001-of-00002.safetensors",
804
+ "encoder.h.14.attn.c_proj.bias": "model-00001-of-00002.safetensors",
805
+ "encoder.h.14.attn.c_proj.weight": "model-00001-of-00002.safetensors",
806
+ "encoder.h.14.ln_1.bias": "model-00001-of-00002.safetensors",
807
+ "encoder.h.14.ln_1.weight": "model-00001-of-00002.safetensors",
808
+ "encoder.h.14.ln_2.bias": "model-00001-of-00002.safetensors",
809
+ "encoder.h.14.ln_2.weight": "model-00001-of-00002.safetensors",
810
+ "encoder.h.14.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
811
+ "encoder.h.14.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
812
+ "encoder.h.14.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
813
+ "encoder.h.14.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
814
+ "encoder.h.15.attn.c_attn.bias": "model-00001-of-00002.safetensors",
815
+ "encoder.h.15.attn.c_attn.weight": "model-00001-of-00002.safetensors",
816
+ "encoder.h.15.attn.c_proj.bias": "model-00001-of-00002.safetensors",
817
+ "encoder.h.15.attn.c_proj.weight": "model-00001-of-00002.safetensors",
818
+ "encoder.h.15.ln_1.bias": "model-00001-of-00002.safetensors",
819
+ "encoder.h.15.ln_1.weight": "model-00001-of-00002.safetensors",
820
+ "encoder.h.15.ln_2.bias": "model-00001-of-00002.safetensors",
821
+ "encoder.h.15.ln_2.weight": "model-00001-of-00002.safetensors",
822
+ "encoder.h.15.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
823
+ "encoder.h.15.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
824
+ "encoder.h.15.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
825
+ "encoder.h.15.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
826
+ "encoder.h.16.attn.c_attn.bias": "model-00001-of-00002.safetensors",
827
+ "encoder.h.16.attn.c_attn.weight": "model-00001-of-00002.safetensors",
828
+ "encoder.h.16.attn.c_proj.bias": "model-00001-of-00002.safetensors",
829
+ "encoder.h.16.attn.c_proj.weight": "model-00001-of-00002.safetensors",
830
+ "encoder.h.16.ln_1.bias": "model-00001-of-00002.safetensors",
831
+ "encoder.h.16.ln_1.weight": "model-00001-of-00002.safetensors",
832
+ "encoder.h.16.ln_2.bias": "model-00001-of-00002.safetensors",
833
+ "encoder.h.16.ln_2.weight": "model-00001-of-00002.safetensors",
834
+ "encoder.h.16.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
835
+ "encoder.h.16.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
836
+ "encoder.h.16.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
837
+ "encoder.h.16.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
838
+ "encoder.h.17.attn.c_attn.bias": "model-00001-of-00002.safetensors",
839
+ "encoder.h.17.attn.c_attn.weight": "model-00001-of-00002.safetensors",
840
+ "encoder.h.17.attn.c_proj.bias": "model-00001-of-00002.safetensors",
841
+ "encoder.h.17.attn.c_proj.weight": "model-00001-of-00002.safetensors",
842
+ "encoder.h.17.ln_1.bias": "model-00001-of-00002.safetensors",
843
+ "encoder.h.17.ln_1.weight": "model-00001-of-00002.safetensors",
844
+ "encoder.h.17.ln_2.bias": "model-00001-of-00002.safetensors",
845
+ "encoder.h.17.ln_2.weight": "model-00001-of-00002.safetensors",
846
+ "encoder.h.17.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
847
+ "encoder.h.17.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
848
+ "encoder.h.17.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
849
+ "encoder.h.17.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
850
+ "encoder.h.18.attn.c_attn.bias": "model-00001-of-00002.safetensors",
851
+ "encoder.h.18.attn.c_attn.weight": "model-00001-of-00002.safetensors",
852
+ "encoder.h.18.attn.c_proj.bias": "model-00001-of-00002.safetensors",
853
+ "encoder.h.18.attn.c_proj.weight": "model-00001-of-00002.safetensors",
854
+ "encoder.h.18.ln_1.bias": "model-00001-of-00002.safetensors",
855
+ "encoder.h.18.ln_1.weight": "model-00001-of-00002.safetensors",
856
+ "encoder.h.18.ln_2.bias": "model-00001-of-00002.safetensors",
857
+ "encoder.h.18.ln_2.weight": "model-00001-of-00002.safetensors",
858
+ "encoder.h.18.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
859
+ "encoder.h.18.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
860
+ "encoder.h.18.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
861
+ "encoder.h.18.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
862
+ "encoder.h.19.attn.c_attn.bias": "model-00001-of-00002.safetensors",
863
+ "encoder.h.19.attn.c_attn.weight": "model-00001-of-00002.safetensors",
864
+ "encoder.h.19.attn.c_proj.bias": "model-00001-of-00002.safetensors",
865
+ "encoder.h.19.attn.c_proj.weight": "model-00001-of-00002.safetensors",
866
+ "encoder.h.19.ln_1.bias": "model-00001-of-00002.safetensors",
867
+ "encoder.h.19.ln_1.weight": "model-00001-of-00002.safetensors",
868
+ "encoder.h.19.ln_2.bias": "model-00001-of-00002.safetensors",
869
+ "encoder.h.19.ln_2.weight": "model-00001-of-00002.safetensors",
870
+ "encoder.h.19.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
871
+ "encoder.h.19.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
872
+ "encoder.h.19.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
873
+ "encoder.h.19.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
874
+ "encoder.h.2.attn.c_attn.bias": "model-00001-of-00002.safetensors",
875
+ "encoder.h.2.attn.c_attn.weight": "model-00001-of-00002.safetensors",
876
+ "encoder.h.2.attn.c_proj.bias": "model-00001-of-00002.safetensors",
877
+ "encoder.h.2.attn.c_proj.weight": "model-00001-of-00002.safetensors",
878
+ "encoder.h.2.ln_1.bias": "model-00001-of-00002.safetensors",
879
+ "encoder.h.2.ln_1.weight": "model-00001-of-00002.safetensors",
880
+ "encoder.h.2.ln_2.bias": "model-00001-of-00002.safetensors",
881
+ "encoder.h.2.ln_2.weight": "model-00001-of-00002.safetensors",
882
+ "encoder.h.2.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
883
+ "encoder.h.2.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
884
+ "encoder.h.2.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
885
+ "encoder.h.2.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
886
+ "encoder.h.20.attn.c_attn.bias": "model-00001-of-00002.safetensors",
887
+ "encoder.h.20.attn.c_attn.weight": "model-00001-of-00002.safetensors",
888
+ "encoder.h.20.attn.c_proj.bias": "model-00001-of-00002.safetensors",
889
+ "encoder.h.20.attn.c_proj.weight": "model-00001-of-00002.safetensors",
890
+ "encoder.h.20.ln_1.bias": "model-00001-of-00002.safetensors",
891
+ "encoder.h.20.ln_1.weight": "model-00001-of-00002.safetensors",
892
+ "encoder.h.20.ln_2.bias": "model-00001-of-00002.safetensors",
893
+ "encoder.h.20.ln_2.weight": "model-00001-of-00002.safetensors",
894
+ "encoder.h.20.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
895
+ "encoder.h.20.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
896
+ "encoder.h.20.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
897
+ "encoder.h.20.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
898
+ "encoder.h.21.attn.c_attn.bias": "model-00001-of-00002.safetensors",
899
+ "encoder.h.21.attn.c_attn.weight": "model-00001-of-00002.safetensors",
900
+ "encoder.h.21.attn.c_proj.bias": "model-00001-of-00002.safetensors",
901
+ "encoder.h.21.attn.c_proj.weight": "model-00001-of-00002.safetensors",
902
+ "encoder.h.21.ln_1.bias": "model-00001-of-00002.safetensors",
903
+ "encoder.h.21.ln_1.weight": "model-00001-of-00002.safetensors",
904
+ "encoder.h.21.ln_2.bias": "model-00001-of-00002.safetensors",
905
+ "encoder.h.21.ln_2.weight": "model-00001-of-00002.safetensors",
906
+ "encoder.h.21.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
907
+ "encoder.h.21.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
908
+ "encoder.h.21.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
909
+ "encoder.h.21.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
910
+ "encoder.h.22.attn.c_attn.bias": "model-00001-of-00002.safetensors",
911
+ "encoder.h.22.attn.c_attn.weight": "model-00001-of-00002.safetensors",
912
+ "encoder.h.22.attn.c_proj.bias": "model-00001-of-00002.safetensors",
913
+ "encoder.h.22.attn.c_proj.weight": "model-00001-of-00002.safetensors",
914
+ "encoder.h.22.ln_1.bias": "model-00001-of-00002.safetensors",
915
+ "encoder.h.22.ln_1.weight": "model-00001-of-00002.safetensors",
916
+ "encoder.h.22.ln_2.bias": "model-00001-of-00002.safetensors",
917
+ "encoder.h.22.ln_2.weight": "model-00001-of-00002.safetensors",
918
+ "encoder.h.22.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
919
+ "encoder.h.22.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
920
+ "encoder.h.22.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
921
+ "encoder.h.22.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
922
+ "encoder.h.23.attn.c_attn.bias": "model-00001-of-00002.safetensors",
923
+ "encoder.h.23.attn.c_attn.weight": "model-00001-of-00002.safetensors",
924
+ "encoder.h.23.attn.c_proj.bias": "model-00001-of-00002.safetensors",
925
+ "encoder.h.23.attn.c_proj.weight": "model-00001-of-00002.safetensors",
926
+ "encoder.h.23.ln_1.bias": "model-00001-of-00002.safetensors",
927
+ "encoder.h.23.ln_1.weight": "model-00001-of-00002.safetensors",
928
+ "encoder.h.23.ln_2.bias": "model-00001-of-00002.safetensors",
929
+ "encoder.h.23.ln_2.weight": "model-00001-of-00002.safetensors",
930
+ "encoder.h.23.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
931
+ "encoder.h.23.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
932
+ "encoder.h.23.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
933
+ "encoder.h.23.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
934
+ "encoder.h.24.attn.c_attn.bias": "model-00001-of-00002.safetensors",
935
+ "encoder.h.24.attn.c_attn.weight": "model-00001-of-00002.safetensors",
936
+ "encoder.h.24.attn.c_proj.bias": "model-00001-of-00002.safetensors",
937
+ "encoder.h.24.attn.c_proj.weight": "model-00001-of-00002.safetensors",
938
+ "encoder.h.24.ln_1.bias": "model-00001-of-00002.safetensors",
939
+ "encoder.h.24.ln_1.weight": "model-00001-of-00002.safetensors",
940
+ "encoder.h.24.ln_2.bias": "model-00001-of-00002.safetensors",
941
+ "encoder.h.24.ln_2.weight": "model-00001-of-00002.safetensors",
942
+ "encoder.h.24.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
943
+ "encoder.h.24.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
944
+ "encoder.h.24.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
945
+ "encoder.h.24.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
946
+ "encoder.h.25.attn.c_attn.bias": "model-00001-of-00002.safetensors",
947
+ "encoder.h.25.attn.c_attn.weight": "model-00001-of-00002.safetensors",
948
+ "encoder.h.25.attn.c_proj.bias": "model-00001-of-00002.safetensors",
949
+ "encoder.h.25.attn.c_proj.weight": "model-00001-of-00002.safetensors",
950
+ "encoder.h.25.ln_1.bias": "model-00001-of-00002.safetensors",
951
+ "encoder.h.25.ln_1.weight": "model-00001-of-00002.safetensors",
952
+ "encoder.h.25.ln_2.bias": "model-00001-of-00002.safetensors",
953
+ "encoder.h.25.ln_2.weight": "model-00001-of-00002.safetensors",
954
+ "encoder.h.25.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
955
+ "encoder.h.25.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
956
+ "encoder.h.25.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
957
+ "encoder.h.25.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
958
+ "encoder.h.26.attn.c_attn.bias": "model-00001-of-00002.safetensors",
959
+ "encoder.h.26.attn.c_attn.weight": "model-00001-of-00002.safetensors",
960
+ "encoder.h.26.attn.c_proj.bias": "model-00001-of-00002.safetensors",
961
+ "encoder.h.26.attn.c_proj.weight": "model-00001-of-00002.safetensors",
962
+ "encoder.h.26.ln_1.bias": "model-00001-of-00002.safetensors",
963
+ "encoder.h.26.ln_1.weight": "model-00001-of-00002.safetensors",
964
+ "encoder.h.26.ln_2.bias": "model-00001-of-00002.safetensors",
965
+ "encoder.h.26.ln_2.weight": "model-00001-of-00002.safetensors",
966
+ "encoder.h.26.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
967
+ "encoder.h.26.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
968
+ "encoder.h.26.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
969
+ "encoder.h.26.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
970
+ "encoder.h.27.attn.c_attn.bias": "model-00001-of-00002.safetensors",
971
+ "encoder.h.27.attn.c_attn.weight": "model-00001-of-00002.safetensors",
972
+ "encoder.h.27.attn.c_proj.bias": "model-00001-of-00002.safetensors",
973
+ "encoder.h.27.attn.c_proj.weight": "model-00001-of-00002.safetensors",
974
+ "encoder.h.27.ln_1.bias": "model-00001-of-00002.safetensors",
975
+ "encoder.h.27.ln_1.weight": "model-00001-of-00002.safetensors",
976
+ "encoder.h.27.ln_2.bias": "model-00001-of-00002.safetensors",
977
+ "encoder.h.27.ln_2.weight": "model-00001-of-00002.safetensors",
978
+ "encoder.h.27.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
979
+ "encoder.h.27.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
980
+ "encoder.h.27.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
981
+ "encoder.h.27.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
982
+ "encoder.h.28.attn.c_attn.bias": "model-00001-of-00002.safetensors",
983
+ "encoder.h.28.attn.c_attn.weight": "model-00001-of-00002.safetensors",
984
+ "encoder.h.28.attn.c_proj.bias": "model-00001-of-00002.safetensors",
985
+ "encoder.h.28.attn.c_proj.weight": "model-00001-of-00002.safetensors",
986
+ "encoder.h.28.ln_1.bias": "model-00001-of-00002.safetensors",
987
+ "encoder.h.28.ln_1.weight": "model-00001-of-00002.safetensors",
988
+ "encoder.h.28.ln_2.bias": "model-00001-of-00002.safetensors",
989
+ "encoder.h.28.ln_2.weight": "model-00001-of-00002.safetensors",
990
+ "encoder.h.28.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
991
+ "encoder.h.28.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
992
+ "encoder.h.28.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
993
+ "encoder.h.28.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
994
+ "encoder.h.29.attn.c_attn.bias": "model-00001-of-00002.safetensors",
995
+ "encoder.h.29.attn.c_attn.weight": "model-00001-of-00002.safetensors",
996
+ "encoder.h.29.attn.c_proj.bias": "model-00001-of-00002.safetensors",
997
+ "encoder.h.29.attn.c_proj.weight": "model-00001-of-00002.safetensors",
998
+ "encoder.h.29.ln_1.bias": "model-00001-of-00002.safetensors",
999
+ "encoder.h.29.ln_1.weight": "model-00001-of-00002.safetensors",
1000
+ "encoder.h.29.ln_2.bias": "model-00001-of-00002.safetensors",
1001
+ "encoder.h.29.ln_2.weight": "model-00001-of-00002.safetensors",
1002
+ "encoder.h.29.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1003
+ "encoder.h.29.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1004
+ "encoder.h.29.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1005
+ "encoder.h.29.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1006
+ "encoder.h.3.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1007
+ "encoder.h.3.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1008
+ "encoder.h.3.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1009
+ "encoder.h.3.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1010
+ "encoder.h.3.ln_1.bias": "model-00001-of-00002.safetensors",
1011
+ "encoder.h.3.ln_1.weight": "model-00001-of-00002.safetensors",
1012
+ "encoder.h.3.ln_2.bias": "model-00001-of-00002.safetensors",
1013
+ "encoder.h.3.ln_2.weight": "model-00001-of-00002.safetensors",
1014
+ "encoder.h.3.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1015
+ "encoder.h.3.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1016
+ "encoder.h.3.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1017
+ "encoder.h.3.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1018
+ "encoder.h.30.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1019
+ "encoder.h.30.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1020
+ "encoder.h.30.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1021
+ "encoder.h.30.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1022
+ "encoder.h.30.ln_1.bias": "model-00001-of-00002.safetensors",
1023
+ "encoder.h.30.ln_1.weight": "model-00001-of-00002.safetensors",
1024
+ "encoder.h.30.ln_2.bias": "model-00001-of-00002.safetensors",
1025
+ "encoder.h.30.ln_2.weight": "model-00001-of-00002.safetensors",
1026
+ "encoder.h.30.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1027
+ "encoder.h.30.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1028
+ "encoder.h.30.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1029
+ "encoder.h.30.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1030
+ "encoder.h.31.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1031
+ "encoder.h.31.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1032
+ "encoder.h.31.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1033
+ "encoder.h.31.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1034
+ "encoder.h.31.ln_1.bias": "model-00001-of-00002.safetensors",
1035
+ "encoder.h.31.ln_1.weight": "model-00001-of-00002.safetensors",
1036
+ "encoder.h.31.ln_2.bias": "model-00001-of-00002.safetensors",
1037
+ "encoder.h.31.ln_2.weight": "model-00001-of-00002.safetensors",
1038
+ "encoder.h.31.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1039
+ "encoder.h.31.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1040
+ "encoder.h.31.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1041
+ "encoder.h.31.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1042
+ "encoder.h.32.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1043
+ "encoder.h.32.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1044
+ "encoder.h.32.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1045
+ "encoder.h.32.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1046
+ "encoder.h.32.ln_1.bias": "model-00001-of-00002.safetensors",
1047
+ "encoder.h.32.ln_1.weight": "model-00001-of-00002.safetensors",
1048
+ "encoder.h.32.ln_2.bias": "model-00001-of-00002.safetensors",
1049
+ "encoder.h.32.ln_2.weight": "model-00001-of-00002.safetensors",
1050
+ "encoder.h.32.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1051
+ "encoder.h.32.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1052
+ "encoder.h.32.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1053
+ "encoder.h.32.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1054
+ "encoder.h.33.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1055
+ "encoder.h.33.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1056
+ "encoder.h.33.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1057
+ "encoder.h.33.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1058
+ "encoder.h.33.ln_1.bias": "model-00001-of-00002.safetensors",
1059
+ "encoder.h.33.ln_1.weight": "model-00001-of-00002.safetensors",
1060
+ "encoder.h.33.ln_2.bias": "model-00001-of-00002.safetensors",
1061
+ "encoder.h.33.ln_2.weight": "model-00001-of-00002.safetensors",
1062
+ "encoder.h.33.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1063
+ "encoder.h.33.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1064
+ "encoder.h.33.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1065
+ "encoder.h.33.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1066
+ "encoder.h.34.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1067
+ "encoder.h.34.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1068
+ "encoder.h.34.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1069
+ "encoder.h.34.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1070
+ "encoder.h.34.ln_1.bias": "model-00001-of-00002.safetensors",
1071
+ "encoder.h.34.ln_1.weight": "model-00001-of-00002.safetensors",
1072
+ "encoder.h.34.ln_2.bias": "model-00001-of-00002.safetensors",
1073
+ "encoder.h.34.ln_2.weight": "model-00001-of-00002.safetensors",
1074
+ "encoder.h.34.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1075
+ "encoder.h.34.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1076
+ "encoder.h.34.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1077
+ "encoder.h.34.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1078
+ "encoder.h.35.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1079
+ "encoder.h.35.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1080
+ "encoder.h.35.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1081
+ "encoder.h.35.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1082
+ "encoder.h.35.ln_1.bias": "model-00001-of-00002.safetensors",
1083
+ "encoder.h.35.ln_1.weight": "model-00001-of-00002.safetensors",
1084
+ "encoder.h.35.ln_2.bias": "model-00001-of-00002.safetensors",
1085
+ "encoder.h.35.ln_2.weight": "model-00001-of-00002.safetensors",
1086
+ "encoder.h.35.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1087
+ "encoder.h.35.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1088
+ "encoder.h.35.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1089
+ "encoder.h.35.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1090
+ "encoder.h.4.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1091
+ "encoder.h.4.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1092
+ "encoder.h.4.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1093
+ "encoder.h.4.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1094
+ "encoder.h.4.ln_1.bias": "model-00001-of-00002.safetensors",
1095
+ "encoder.h.4.ln_1.weight": "model-00001-of-00002.safetensors",
1096
+ "encoder.h.4.ln_2.bias": "model-00001-of-00002.safetensors",
1097
+ "encoder.h.4.ln_2.weight": "model-00001-of-00002.safetensors",
1098
+ "encoder.h.4.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1099
+ "encoder.h.4.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1100
+ "encoder.h.4.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1101
+ "encoder.h.4.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1102
+ "encoder.h.5.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1103
+ "encoder.h.5.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1104
+ "encoder.h.5.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1105
+ "encoder.h.5.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1106
+ "encoder.h.5.ln_1.bias": "model-00001-of-00002.safetensors",
1107
+ "encoder.h.5.ln_1.weight": "model-00001-of-00002.safetensors",
1108
+ "encoder.h.5.ln_2.bias": "model-00001-of-00002.safetensors",
1109
+ "encoder.h.5.ln_2.weight": "model-00001-of-00002.safetensors",
1110
+ "encoder.h.5.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1111
+ "encoder.h.5.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1112
+ "encoder.h.5.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1113
+ "encoder.h.5.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1114
+ "encoder.h.6.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1115
+ "encoder.h.6.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1116
+ "encoder.h.6.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1117
+ "encoder.h.6.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1118
+ "encoder.h.6.ln_1.bias": "model-00001-of-00002.safetensors",
1119
+ "encoder.h.6.ln_1.weight": "model-00001-of-00002.safetensors",
1120
+ "encoder.h.6.ln_2.bias": "model-00001-of-00002.safetensors",
1121
+ "encoder.h.6.ln_2.weight": "model-00001-of-00002.safetensors",
1122
+ "encoder.h.6.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1123
+ "encoder.h.6.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1124
+ "encoder.h.6.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1125
+ "encoder.h.6.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1126
+ "encoder.h.7.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1127
+ "encoder.h.7.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1128
+ "encoder.h.7.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1129
+ "encoder.h.7.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1130
+ "encoder.h.7.ln_1.bias": "model-00001-of-00002.safetensors",
1131
+ "encoder.h.7.ln_1.weight": "model-00001-of-00002.safetensors",
1132
+ "encoder.h.7.ln_2.bias": "model-00001-of-00002.safetensors",
1133
+ "encoder.h.7.ln_2.weight": "model-00001-of-00002.safetensors",
1134
+ "encoder.h.7.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1135
+ "encoder.h.7.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1136
+ "encoder.h.7.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1137
+ "encoder.h.7.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1138
+ "encoder.h.8.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1139
+ "encoder.h.8.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1140
+ "encoder.h.8.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1141
+ "encoder.h.8.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1142
+ "encoder.h.8.ln_1.bias": "model-00001-of-00002.safetensors",
1143
+ "encoder.h.8.ln_1.weight": "model-00001-of-00002.safetensors",
1144
+ "encoder.h.8.ln_2.bias": "model-00001-of-00002.safetensors",
1145
+ "encoder.h.8.ln_2.weight": "model-00001-of-00002.safetensors",
1146
+ "encoder.h.8.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1147
+ "encoder.h.8.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1148
+ "encoder.h.8.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1149
+ "encoder.h.8.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1150
+ "encoder.h.9.attn.c_attn.bias": "model-00001-of-00002.safetensors",
1151
+ "encoder.h.9.attn.c_attn.weight": "model-00001-of-00002.safetensors",
1152
+ "encoder.h.9.attn.c_proj.bias": "model-00001-of-00002.safetensors",
1153
+ "encoder.h.9.attn.c_proj.weight": "model-00001-of-00002.safetensors",
1154
+ "encoder.h.9.ln_1.bias": "model-00001-of-00002.safetensors",
1155
+ "encoder.h.9.ln_1.weight": "model-00001-of-00002.safetensors",
1156
+ "encoder.h.9.ln_2.bias": "model-00001-of-00002.safetensors",
1157
+ "encoder.h.9.ln_2.weight": "model-00001-of-00002.safetensors",
1158
+ "encoder.h.9.mlp.c_fc.bias": "model-00001-of-00002.safetensors",
1159
+ "encoder.h.9.mlp.c_fc.weight": "model-00001-of-00002.safetensors",
1160
+ "encoder.h.9.mlp.c_proj.bias": "model-00001-of-00002.safetensors",
1161
+ "encoder.h.9.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
1162
+ "encoder.ln_f.bias": "model-00001-of-00002.safetensors",
1163
+ "encoder.ln_f.weight": "model-00001-of-00002.safetensors",
1164
+ "encoder.wpe.weight": "model-00001-of-00002.safetensors",
1165
+ "encoder.wte.weight": "model-00001-of-00002.safetensors"
1166
+ }
1167
+ }
runs/Mar13_12-57-59_gpu09gmz.benchmarkcenter.megware.com/events.out.tfevents.1741870786.gpu09gmz.benchmarkcenter.megware.com.12123.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ead00ec38e49b399b3678a80ec57f2711fd44307f1773afd37406d91c23b8c7
3
+ size 11132
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
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "pad_token": "<|endoftext|>",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c9f77bb1fa4897a8893da2f76f6d81861e92f44830a0f58fa3b196f004ead28b
3
+ size 5496
vocab.json ADDED
The diff for this file is too large to render. See raw diff