Hugging Face's logo

tyzhu
/
lmind_hotpot_train5000_eval5000_v1_recite_qa_gpt2-xl

Text Generation
Transformers
PyTorch
gpt2
Generated from Trainer
Eval Results
text-generation-inference
Model card Files Community
1
tyzhu commited on
Commit
49a9c83
·
verified ·
1 Parent(s): 9a4ead6

Training in progress, epoch 5, checkpoint

Browse files
Files changed (14) hide show
  1. checkpoint-7260/added_tokens.json +3 -0
  2. checkpoint-7260/config.json +42 -0
  3. checkpoint-7260/generation_config.json +8 -0
  4. checkpoint-7260/merges.txt +0 -0
  5. checkpoint-7260/optimizer.pt +3 -0
  6. checkpoint-7260/pytorch_model.bin +3 -0
  7. checkpoint-7260/rng_state.pth +3 -0
  8. checkpoint-7260/scheduler.pt +3 -0
  9. checkpoint-7260/special_tokens_map.json +6 -0
  10. checkpoint-7260/tokenizer.json +0 -0
  11. checkpoint-7260/tokenizer_config.json +20 -0
  12. checkpoint-7260/trainer_state.json +258 -0
  13. checkpoint-7260/training_args.bin +3 -0
  14. checkpoint-7260/vocab.json +0 -0
checkpoint-7260/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 50256
3
+ }
checkpoint-7260/config.json ADDED
@@ -0,0 +1,42 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2-xl",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "do_sample": true,
10
+ "embd_pdrop": 0.1,
11
+ "eos_token_id": 50256,
12
+ "initializer_range": 0.02,
13
+ "layer_norm_epsilon": 1e-05,
14
+ "max_length": 50,
15
+ "model_type": "gpt2",
16
+ "n_ctx": 1024,
17
+ "n_embd": 1600,
18
+ "n_head": 25,
19
+ "n_inner": null,
20
+ "n_layer": 48,
21
+ "n_positions": 1024,
22
+ "output_past": true,
23
+ "reorder_and_upcast_attn": false,
24
+ "resid_pdrop": 0.1,
25
+ "scale_attn_by_inverse_layer_idx": false,
26
+ "scale_attn_weights": true,
27
+ "summary_activation": null,
28
+ "summary_first_dropout": 0.1,
29
+ "summary_proj_to_labels": true,
30
+ "summary_type": "cls_index",
31
+ "summary_use_proj": true,
32
+ "task_specific_params": {
33
+ "text-generation": {
34
+ "do_sample": true,
35
+ "max_length": 50
36
+ }
37
+ },
38
+ "torch_dtype": "float32",
39
+ "transformers_version": "4.34.0",
40
+ "use_cache": true,
41
+ "vocab_size": 50257
42
+ }
checkpoint-7260/generation_config.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "do_sample": true,
5
+ "eos_token_id": 50256,
6
+ "max_length": 50,
7
+ "transformers_version": "4.34.0"
8
+ }
checkpoint-7260/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7260/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edd96d5737340f65c37b3f2f898b791128113a4a2c2eb0d6aad0226aa17a8db0
3
+ size 12461385454
checkpoint-7260/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b7ef896fe9f308b43c8c06af2ebea8526f671228f8515b053b45280d92d66de8
3
+ size 6230637102
checkpoint-7260/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fa64fbb8199a3122f1230cb1bfc429f81af167f14c810f8d93c0424c19cbcd9
3
+ size 14244
checkpoint-7260/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4894419dbbc4dde3d545220aec20642d670c4770e4981e0a3e015bff3d265190
3
+ size 1064
checkpoint-7260/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
+ }
checkpoint-7260/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-7260/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": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ }
12
+ },
13
+ "additional_special_tokens": [],
14
+ "bos_token": "<|endoftext|>",
15
+ "clean_up_tokenization_spaces": true,
16
+ "eos_token": "<|endoftext|>",
17
+ "model_max_length": 1024,
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
checkpoint-7260/trainer_state.json ADDED
@@ -0,0 +1,258 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 5.0,
5
+ "eval_steps": 1452,
6
+ "global_step": 7260,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.2,
13
+ "learning_rate": 3e-05,
14
+ "loss": 2.2721,
15
+ "step": 291
16
+ },
17
+ {
18
+ "epoch": 0.4,
19
+ "learning_rate": 3e-05,
20
+ "loss": 2.0539,
21
+ "step": 582
22
+ },
23
+ {
24
+ "epoch": 0.6,
25
+ "learning_rate": 3e-05,
26
+ "loss": 2.0079,
27
+ "step": 873
28
+ },
29
+ {
30
+ "epoch": 0.8,
31
+ "learning_rate": 3e-05,
32
+ "loss": 1.9686,
33
+ "step": 1164
34
+ },
35
+ {
36
+ "epoch": 1.0,
37
+ "eval_accuracy": 0.668149394347241,
38
+ "eval_loss": 1.5662202835083008,
39
+ "eval_runtime": 137.7572,
40
+ "eval_samples_per_second": 36.296,
41
+ "eval_steps_per_second": 2.272,
42
+ "step": 1452
43
+ },
44
+ {
45
+ "epoch": 1.0,
46
+ "eval_exact_match": 9.04,
47
+ "eval_f1": 13.087046780017364,
48
+ "eval_qa_bleu": 5.297766450359973,
49
+ "eval_qa_exact_match": 0.085,
50
+ "eval_recite_bleu": 14.716037077738477,
51
+ "eval_recite_exact_match": 0.0,
52
+ "step": 1452
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "learning_rate": 3e-05,
57
+ "loss": 1.9076,
58
+ "step": 1455
59
+ },
60
+ {
61
+ "epoch": 1.2,
62
+ "learning_rate": 3e-05,
63
+ "loss": 1.5211,
64
+ "step": 1746
65
+ },
66
+ {
67
+ "epoch": 1.4,
68
+ "learning_rate": 3e-05,
69
+ "loss": 1.5039,
70
+ "step": 2037
71
+ },
72
+ {
73
+ "epoch": 1.6,
74
+ "learning_rate": 3e-05,
75
+ "loss": 1.4834,
76
+ "step": 2328
77
+ },
78
+ {
79
+ "epoch": 1.8,
80
+ "learning_rate": 3e-05,
81
+ "loss": 1.4604,
82
+ "step": 2619
83
+ },
84
+ {
85
+ "epoch": 2.0,
86
+ "eval_accuracy": 0.6914514131897712,
87
+ "eval_loss": 1.206884741783142,
88
+ "eval_runtime": 139.7305,
89
+ "eval_samples_per_second": 35.783,
90
+ "eval_steps_per_second": 2.24,
91
+ "step": 2904
92
+ },
93
+ {
94
+ "epoch": 2.0,
95
+ "eval_exact_match": 11.64,
96
+ "eval_f1": 16.917593370997537,
97
+ "eval_qa_bleu": 7.627957176996819,
98
+ "eval_qa_exact_match": 0.1106,
99
+ "eval_recite_bleu": 18.356743056908567,
100
+ "eval_recite_exact_match": 0.0,
101
+ "step": 2904
102
+ },
103
+ {
104
+ "epoch": 2.0,
105
+ "learning_rate": 3e-05,
106
+ "loss": 1.4368,
107
+ "step": 2910
108
+ },
109
+ {
110
+ "epoch": 2.2,
111
+ "learning_rate": 3e-05,
112
+ "loss": 1.0705,
113
+ "step": 3201
114
+ },
115
+ {
116
+ "epoch": 2.4,
117
+ "learning_rate": 3e-05,
118
+ "loss": 1.0685,
119
+ "step": 3492
120
+ },
121
+ {
122
+ "epoch": 2.61,
123
+ "learning_rate": 3e-05,
124
+ "loss": 1.0554,
125
+ "step": 3783
126
+ },
127
+ {
128
+ "epoch": 2.81,
129
+ "learning_rate": 3e-05,
130
+ "loss": 1.0522,
131
+ "step": 4074
132
+ },
133
+ {
134
+ "epoch": 3.0,
135
+ "eval_accuracy": 0.7148896366083446,
136
+ "eval_loss": 0.9045500159263611,
137
+ "eval_runtime": 141.6455,
138
+ "eval_samples_per_second": 35.299,
139
+ "eval_steps_per_second": 2.21,
140
+ "step": 4356
141
+ },
142
+ {
143
+ "epoch": 3.0,
144
+ "eval_exact_match": 12.84,
145
+ "eval_f1": 18.555356874635468,
146
+ "eval_qa_bleu": 8.395879921702353,
147
+ "eval_qa_exact_match": 0.1188,
148
+ "eval_recite_bleu": 20.824221134285214,
149
+ "eval_recite_exact_match": 0.0,
150
+ "step": 4356
151
+ },
152
+ {
153
+ "epoch": 3.01,
154
+ "learning_rate": 3e-05,
155
+ "loss": 1.0375,
156
+ "step": 4365
157
+ },
158
+ {
159
+ "epoch": 3.21,
160
+ "learning_rate": 3e-05,
161
+ "loss": 0.7373,
162
+ "step": 4656
163
+ },
164
+ {
165
+ "epoch": 3.41,
166
+ "learning_rate": 3e-05,
167
+ "loss": 0.7497,
168
+ "step": 4947
169
+ },
170
+ {
171
+ "epoch": 3.61,
172
+ "learning_rate": 3e-05,
173
+ "loss": 0.7479,
174
+ "step": 5238
175
+ },
176
+ {
177
+ "epoch": 3.81,
178
+ "learning_rate": 3e-05,
179
+ "loss": 0.7455,
180
+ "step": 5529
181
+ },
182
+ {
183
+ "epoch": 4.0,
184
+ "eval_accuracy": 0.7340508748317631,
185
+ "eval_loss": 0.6843858957290649,
186
+ "eval_runtime": 143.5169,
187
+ "eval_samples_per_second": 34.839,
188
+ "eval_steps_per_second": 2.181,
189
+ "step": 5808
190
+ },
191
+ {
192
+ "epoch": 4.0,
193
+ "eval_exact_match": 13.44,
194
+ "eval_f1": 19.418847592695993,
195
+ "eval_qa_bleu": 8.731752840499793,
196
+ "eval_qa_exact_match": 0.1256,
197
+ "eval_recite_bleu": 24.11683390270685,
198
+ "eval_recite_exact_match": 0.0002,
199
+ "step": 5808
200
+ },
201
+ {
202
+ "epoch": 4.01,
203
+ "learning_rate": 3e-05,
204
+ "loss": 0.7226,
205
+ "step": 5820
206
+ },
207
+ {
208
+ "epoch": 4.21,
209
+ "learning_rate": 3e-05,
210
+ "loss": 0.5117,
211
+ "step": 6111
212
+ },
213
+ {
214
+ "epoch": 4.41,
215
+ "learning_rate": 3e-05,
216
+ "loss": 0.5208,
217
+ "step": 6402
218
+ },
219
+ {
220
+ "epoch": 4.61,
221
+ "learning_rate": 3e-05,
222
+ "loss": 0.5271,
223
+ "step": 6693
224
+ },
225
+ {
226
+ "epoch": 4.81,
227
+ "learning_rate": 3e-05,
228
+ "loss": 0.5307,
229
+ "step": 6984
230
+ },
231
+ {
232
+ "epoch": 5.0,
233
+ "eval_accuracy": 0.7474745625841185,
234
+ "eval_loss": 0.5420497059822083,
235
+ "eval_runtime": 144.0475,
236
+ "eval_samples_per_second": 34.711,
237
+ "eval_steps_per_second": 2.173,
238
+ "step": 7260
239
+ },
240
+ {
241
+ "epoch": 5.0,
242
+ "eval_exact_match": 15.08,
243
+ "eval_f1": 21.4491845879152,
244
+ "eval_qa_bleu": 10.89471390420015,
245
+ "eval_qa_exact_match": 0.1434,
246
+ "eval_recite_bleu": 27.62075386163555,
247
+ "eval_recite_exact_match": 0.0014,
248
+ "step": 7260
249
+ }
250
+ ],
251
+ "logging_steps": 291,
252
+ "max_steps": 29040,
253
+ "num_train_epochs": 20,
254
+ "save_steps": 500,
255
+ "total_flos": 3.317149357867008e+17,
256
+ "trial_name": null,
257
+ "trial_params": null
258
+ }
checkpoint-7260/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d993d1b3c496d84dac3774165ea3fdf800c76280a435c891943344990df3011
3
+ size 4728
checkpoint-7260/vocab.json ADDED
The diff for this file is too large to render. See raw diff