Hugging Face's logo

tyzhu
/
lmind_hotpot_train8000_eval7405_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
3d3516c
·
verified ·
1 Parent(s): 936cc71

Training in progress, epoch 6, checkpoint

Browse files
Files changed (14) hide show
  1. checkpoint-13074/added_tokens.json +3 -0
  2. checkpoint-13074/config.json +42 -0
  3. checkpoint-13074/generation_config.json +8 -0
  4. checkpoint-13074/merges.txt +0 -0
  5. checkpoint-13074/optimizer.pt +3 -0
  6. checkpoint-13074/pytorch_model.bin +3 -0
  7. checkpoint-13074/rng_state.pth +3 -0
  8. checkpoint-13074/scheduler.pt +3 -0
  9. checkpoint-13074/special_tokens_map.json +6 -0
  10. checkpoint-13074/tokenizer.json +0 -0
  11. checkpoint-13074/tokenizer_config.json +20 -0
  12. checkpoint-13074/trainer_state.json +307 -0
  13. checkpoint-13074/training_args.bin +3 -0
  14. checkpoint-13074/vocab.json +0 -0
checkpoint-13074/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<|endoftext|>": 50256
3
+ }
checkpoint-13074/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-13074/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-13074/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-13074/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ae55ccd020d560fa3ce288cb106795049eafbf4be95907816f8c155f7dd3ab5
3
+ size 12461385454
checkpoint-13074/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:24a0c1a40b223b4b2dde0461bc90a82a3aaabf7205cd9f898c5dde219c886606
3
+ size 6230637102
checkpoint-13074/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e7543b9514a8bdb90c1e121d545bd4ca346cc54f5aca2e616560b4c22c18572d
3
+ size 14244
checkpoint-13074/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4c124c17936f5319daf483f0399d8c037a87a48d6617da4f428dba073a55452e
3
+ size 1064
checkpoint-13074/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-13074/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-13074/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-13074/trainer_state.json ADDED
@@ -0,0 +1,307 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 6.0,
5
+ "eval_steps": 2179,
6
+ "global_step": 13074,
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": 2.2464,
15
+ "step": 436
16
+ },
17
+ {
18
+ "epoch": 0.4,
19
+ "learning_rate": 3e-05,
20
+ "loss": 2.0544,
21
+ "step": 872
22
+ },
23
+ {
24
+ "epoch": 0.6,
25
+ "learning_rate": 3e-05,
26
+ "loss": 1.9968,
27
+ "step": 1308
28
+ },
29
+ {
30
+ "epoch": 0.8,
31
+ "learning_rate": 3e-05,
32
+ "loss": 1.9537,
33
+ "step": 1744
34
+ },
35
+ {
36
+ "epoch": 1.0,
37
+ "eval_accuracy": 0.6693020521036767,
38
+ "eval_loss": 1.5623117685317993,
39
+ "eval_runtime": 220.2436,
40
+ "eval_samples_per_second": 33.622,
41
+ "eval_steps_per_second": 2.102,
42
+ "step": 2179
43
+ },
44
+ {
45
+ "epoch": 1.0,
46
+ "eval_exact_match": 10.24983119513842,
47
+ "eval_f1": 15.257097040945366,
48
+ "eval_qa_bleu": 5.8382901471491575,
49
+ "eval_qa_exact_match": 0.09736664415935178,
50
+ "eval_recite_bleu": 17.0930417063702,
51
+ "eval_recite_exact_match": 0.0,
52
+ "step": 2179
53
+ },
54
+ {
55
+ "epoch": 1.0,
56
+ "learning_rate": 3e-05,
57
+ "loss": 1.9029,
58
+ "step": 2180
59
+ },
60
+ {
61
+ "epoch": 1.2,
62
+ "learning_rate": 3e-05,
63
+ "loss": 1.5215,
64
+ "step": 2616
65
+ },
66
+ {
67
+ "epoch": 1.4,
68
+ "learning_rate": 3e-05,
69
+ "loss": 1.5077,
70
+ "step": 3052
71
+ },
72
+ {
73
+ "epoch": 1.6,
74
+ "learning_rate": 3e-05,
75
+ "loss": 1.5016,
76
+ "step": 3488
77
+ },
78
+ {
79
+ "epoch": 1.8,
80
+ "learning_rate": 3e-05,
81
+ "loss": 1.4734,
82
+ "step": 3924
83
+ },
84
+ {
85
+ "epoch": 2.0,
86
+ "eval_accuracy": 0.6923618412861704,
87
+ "eval_loss": 1.2099343538284302,
88
+ "eval_runtime": 222.1542,
89
+ "eval_samples_per_second": 33.333,
90
+ "eval_steps_per_second": 2.084,
91
+ "step": 4358
92
+ },
93
+ {
94
+ "epoch": 2.0,
95
+ "eval_exact_match": 10.61444969615125,
96
+ "eval_f1": 15.935672519396675,
97
+ "eval_qa_bleu": 6.524750946372503,
98
+ "eval_qa_exact_match": 0.10209318028359217,
99
+ "eval_recite_bleu": 18.618913529891735,
100
+ "eval_recite_exact_match": 0.0,
101
+ "step": 4358
102
+ },
103
+ {
104
+ "epoch": 2.0,
105
+ "learning_rate": 3e-05,
106
+ "loss": 1.4425,
107
+ "step": 4360
108
+ },
109
+ {
110
+ "epoch": 2.2,
111
+ "learning_rate": 3e-05,
112
+ "loss": 1.0827,
113
+ "step": 4796
114
+ },
115
+ {
116
+ "epoch": 2.4,
117
+ "learning_rate": 3e-05,
118
+ "loss": 1.0876,
119
+ "step": 5232
120
+ },
121
+ {
122
+ "epoch": 2.6,
123
+ "learning_rate": 3e-05,
124
+ "loss": 1.0887,
125
+ "step": 5668
126
+ },
127
+ {
128
+ "epoch": 2.8,
129
+ "learning_rate": 3e-05,
130
+ "loss": 1.0665,
131
+ "step": 6104
132
+ },
133
+ {
134
+ "epoch": 3.0,
135
+ "eval_accuracy": 0.7147465927772421,
136
+ "eval_loss": 0.9177509546279907,
137
+ "eval_runtime": 226.6441,
138
+ "eval_samples_per_second": 32.672,
139
+ "eval_steps_per_second": 2.043,
140
+ "step": 6537
141
+ },
142
+ {
143
+ "epoch": 3.0,
144
+ "eval_exact_match": 12.397029034436192,
145
+ "eval_f1": 17.721572805797255,
146
+ "eval_qa_bleu": 8.139751517788653,
147
+ "eval_qa_exact_match": 0.11978392977717758,
148
+ "eval_recite_bleu": 21.769859861022653,
149
+ "eval_recite_exact_match": 0.0,
150
+ "step": 6537
151
+ },
152
+ {
153
+ "epoch": 3.0,
154
+ "learning_rate": 3e-05,
155
+ "loss": 0.7772,
156
+ "step": 6540
157
+ },
158
+ {
159
+ "epoch": 3.2,
160
+ "learning_rate": 3e-05,
161
+ "loss": 0.7688,
162
+ "step": 6976
163
+ },
164
+ {
165
+ "epoch": 3.4,
166
+ "learning_rate": 3e-05,
167
+ "loss": 0.7598,
168
+ "step": 7412
169
+ },
170
+ {
171
+ "epoch": 3.6,
172
+ "learning_rate": 3e-05,
173
+ "loss": 0.77,
174
+ "step": 7848
175
+ },
176
+ {
177
+ "epoch": 3.8,
178
+ "learning_rate": 3e-05,
179
+ "loss": 0.7684,
180
+ "step": 8284
181
+ },
182
+ {
183
+ "epoch": 4.0,
184
+ "eval_accuracy": 0.7331474586575765,
185
+ "eval_loss": 0.6987847089767456,
186
+ "eval_runtime": 221.7249,
187
+ "eval_samples_per_second": 33.397,
188
+ "eval_steps_per_second": 2.088,
189
+ "step": 8716
190
+ },
191
+ {
192
+ "epoch": 4.0,
193
+ "eval_exact_match": 13.409858203916272,
194
+ "eval_f1": 19.326947132558665,
195
+ "eval_qa_bleu": 8.79837393539171,
196
+ "eval_qa_exact_match": 0.12869682646860228,
197
+ "eval_recite_bleu": 23.848732661358156,
198
+ "eval_recite_exact_match": 0.0,
199
+ "step": 8716
200
+ },
201
+ {
202
+ "epoch": 4.0,
203
+ "learning_rate": 3e-05,
204
+ "loss": 0.7559,
205
+ "step": 8720
206
+ },
207
+ {
208
+ "epoch": 4.2,
209
+ "learning_rate": 3e-05,
210
+ "loss": 0.5354,
211
+ "step": 9156
212
+ },
213
+ {
214
+ "epoch": 4.4,
215
+ "learning_rate": 3e-05,
216
+ "loss": 0.5511,
217
+ "step": 9592
218
+ },
219
+ {
220
+ "epoch": 4.6,
221
+ "learning_rate": 3e-05,
222
+ "loss": 0.5496,
223
+ "step": 10028
224
+ },
225
+ {
226
+ "epoch": 4.8,
227
+ "learning_rate": 3e-05,
228
+ "loss": 0.548,
229
+ "step": 10464
230
+ },
231
+ {
232
+ "epoch": 5.0,
233
+ "eval_accuracy": 0.7465929589970037,
234
+ "eval_loss": 0.5567022562026978,
235
+ "eval_runtime": 226.4909,
236
+ "eval_samples_per_second": 32.694,
237
+ "eval_steps_per_second": 2.044,
238
+ "step": 10895
239
+ },
240
+ {
241
+ "epoch": 5.0,
242
+ "eval_exact_match": 14.220121539500338,
243
+ "eval_f1": 20.736856637999697,
244
+ "eval_qa_bleu": 9.542509177303282,
245
+ "eval_qa_exact_match": 0.13652937204591492,
246
+ "eval_recite_bleu": 27.315073242003514,
247
+ "eval_recite_exact_match": 0.0009453072248480756,
248
+ "step": 10895
249
+ },
250
+ {
251
+ "epoch": 5.0,
252
+ "learning_rate": 3e-05,
253
+ "loss": 0.5518,
254
+ "step": 10900
255
+ },
256
+ {
257
+ "epoch": 5.2,
258
+ "learning_rate": 3e-05,
259
+ "loss": 0.3837,
260
+ "step": 11336
261
+ },
262
+ {
263
+ "epoch": 5.4,
264
+ "learning_rate": 3e-05,
265
+ "loss": 0.3963,
266
+ "step": 11772
267
+ },
268
+ {
269
+ "epoch": 5.6,
270
+ "learning_rate": 3e-05,
271
+ "loss": 0.404,
272
+ "step": 12208
273
+ },
274
+ {
275
+ "epoch": 5.8,
276
+ "learning_rate": 3e-05,
277
+ "loss": 0.4039,
278
+ "step": 12644
279
+ },
280
+ {
281
+ "epoch": 6.0,
282
+ "eval_accuracy": 0.7551407100982113,
283
+ "eval_loss": 0.4728190004825592,
284
+ "eval_runtime": 223.8973,
285
+ "eval_samples_per_second": 33.073,
286
+ "eval_steps_per_second": 2.068,
287
+ "step": 13074
288
+ },
289
+ {
290
+ "epoch": 6.0,
291
+ "eval_exact_match": 16.12424037812289,
292
+ "eval_f1": 23.0965386152669,
293
+ "eval_qa_bleu": 10.706681466541065,
294
+ "eval_qa_exact_match": 0.1524645509790682,
295
+ "eval_recite_bleu": 31.825156559426375,
296
+ "eval_recite_exact_match": 0.0036461850101282916,
297
+ "step": 13074
298
+ }
299
+ ],
300
+ "logging_steps": 436,
301
+ "max_steps": 43580,
302
+ "num_train_epochs": 20,
303
+ "save_steps": 500,
304
+ "total_flos": 6.047189989653504e+17,
305
+ "trial_name": null,
306
+ "trial_params": null
307
+ }
checkpoint-13074/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e80417f2e51a1b94f31dc1cc7a0a8119a42bea950f9110af66e643c4cb34fee
3
+ size 4728
checkpoint-13074/vocab.json ADDED
The diff for this file is too large to render. See raw diff