Garsa3112 commited on
Commit
d9ef164
·
verified ·
1 Parent(s): 1532b6c

Add files using upload-large-folder tool

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. base/100e1/config.json +39 -0
  2. base/100e1/generation_config.json +6 -0
  3. base/100e1/merges.txt +0 -0
  4. base/100e1/special_tokens_map.json +5 -0
  5. base/100e1/tokenizer.json +0 -0
  6. base/100e1/tokenizer_config.json +20 -0
  7. base/100e1/trainer_state.json +285 -0
  8. base/100e1/vocab.json +0 -0
  9. base/100e2/config.json +39 -0
  10. base/100e2/generation_config.json +6 -0
  11. base/100e2/merges.txt +0 -0
  12. base/100e2/special_tokens_map.json +5 -0
  13. base/100e2/tokenizer.json +0 -0
  14. base/100e2/tokenizer_config.json +20 -0
  15. base/100e2/trainer_state.json +537 -0
  16. base/100e2/vocab.json +0 -0
  17. base/100e3/config.json +39 -0
  18. base/100e3/generation_config.json +6 -0
  19. base/100e3/merges.txt +0 -0
  20. base/100e3/special_tokens_map.json +5 -0
  21. base/100e3/tokenizer.json +0 -0
  22. base/100e3/tokenizer_config.json +20 -0
  23. base/100e3/trainer_state.json +796 -0
  24. base/100e3/vocab.json +0 -0
  25. base/10e1/config.json +39 -0
  26. base/10e1/generation_config.json +6 -0
  27. base/10e1/merges.txt +0 -0
  28. base/10e1/special_tokens_map.json +5 -0
  29. base/10e1/tokenizer.json +0 -0
  30. base/10e1/tokenizer_config.json +20 -0
  31. base/10e1/trainer_state.json +54 -0
  32. base/10e1/vocab.json +0 -0
  33. base/10e2/config.json +39 -0
  34. base/10e2/generation_config.json +6 -0
  35. base/10e2/merges.txt +0 -0
  36. base/10e2/special_tokens_map.json +5 -0
  37. base/10e2/tokenizer.json +0 -0
  38. base/10e2/tokenizer_config.json +20 -0
  39. base/10e2/trainer_state.json +306 -0
  40. base/10e2/vocab.json +0 -0
  41. base/10e3/config.json +39 -0
  42. base/10e3/generation_config.json +6 -0
  43. base/10e3/merges.txt +0 -0
  44. base/10e3/special_tokens_map.json +5 -0
  45. base/10e3/tokenizer.json +0 -0
  46. base/10e3/tokenizer_config.json +20 -0
  47. base/10e3/trainer_state.json +565 -0
  48. base/10e3/vocab.json +0 -0
  49. base/110e3/config.json +39 -0
  50. base/110e3/generation_config.json +6 -0
base/100e1/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/100e1/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/100e1/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/100e1/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/100e1/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/100e1/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/100e1/trainer_state.json ADDED
@@ -0,0 +1,285 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9995601495491533,
5
+ "eval_steps": 500,
6
+ "global_step": 18180,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.10996261271167802,
34
+ "grad_norm": 6.73529577255249,
35
+ "learning_rate": 1.9446951760876e-05,
36
+ "loss": 2.4492,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.13745326588959753,
41
+ "grad_norm": 6.306326866149902,
42
+ "learning_rate": 1.9261985794613793e-05,
43
+ "loss": 2.4463,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.16494391906751704,
48
+ "grad_norm": 5.651560306549072,
49
+ "learning_rate": 1.9077019828351584e-05,
50
+ "loss": 2.3291,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.19243457224543656,
55
+ "grad_norm": 4.0863471031188965,
56
+ "learning_rate": 1.8892053862089378e-05,
57
+ "loss": 2.3461,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.21992522542335605,
62
+ "grad_norm": 4.746509075164795,
63
+ "learning_rate": 1.8707087895827168e-05,
64
+ "loss": 2.2806,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.24741587860127556,
69
+ "grad_norm": 4.339619159698486,
70
+ "learning_rate": 1.8522121929564962e-05,
71
+ "loss": 2.2746,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.27490653177919505,
76
+ "grad_norm": 4.018620014190674,
77
+ "learning_rate": 1.8337155963302752e-05,
78
+ "loss": 2.2403,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.3023971849571146,
83
+ "grad_norm": 4.380566120147705,
84
+ "learning_rate": 1.8152189997040546e-05,
85
+ "loss": 2.226,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.3298878381350341,
90
+ "grad_norm": 4.302157402038574,
91
+ "learning_rate": 1.796722403077834e-05,
92
+ "loss": 2.2119,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.3573784913129536,
97
+ "grad_norm": 4.434130668640137,
98
+ "learning_rate": 1.778225806451613e-05,
99
+ "loss": 2.2138,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.3848691444908731,
104
+ "grad_norm": 3.770848035812378,
105
+ "learning_rate": 1.759729209825392e-05,
106
+ "loss": 2.2103,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.4123597976687926,
111
+ "grad_norm": 2.9288535118103027,
112
+ "learning_rate": 1.7412326131991715e-05,
113
+ "loss": 2.1397,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.4398504508467121,
118
+ "grad_norm": 3.2082479000091553,
119
+ "learning_rate": 1.722736016572951e-05,
120
+ "loss": 2.2021,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.46734110402463164,
125
+ "grad_norm": 3.6131439208984375,
126
+ "learning_rate": 1.70423941994673e-05,
127
+ "loss": 2.2106,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.4948317572025511,
132
+ "grad_norm": 2.992415189743042,
133
+ "learning_rate": 1.6857798165137616e-05,
134
+ "loss": 2.1246,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.5223224103804707,
139
+ "grad_norm": 3.4553306102752686,
140
+ "learning_rate": 1.667283219887541e-05,
141
+ "loss": 2.0973,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.5498130635583901,
146
+ "grad_norm": 2.82578182220459,
147
+ "learning_rate": 1.64878662326132e-05,
148
+ "loss": 2.1257,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.5773037167363096,
153
+ "grad_norm": 2.674267292022705,
154
+ "learning_rate": 1.630290026635099e-05,
155
+ "loss": 2.13,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.6047943699142292,
160
+ "grad_norm": 3.27205491065979,
161
+ "learning_rate": 1.6117934300088784e-05,
162
+ "loss": 2.1855,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 0.6322850230921486,
167
+ "grad_norm": 3.042849540710449,
168
+ "learning_rate": 1.5932968333826578e-05,
169
+ "loss": 2.1532,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 0.6597756762700682,
174
+ "grad_norm": 3.618957042694092,
175
+ "learning_rate": 1.5748372299496894e-05,
176
+ "loss": 2.1202,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 0.6872663294479877,
181
+ "grad_norm": 3.6390328407287598,
182
+ "learning_rate": 1.5563406333234685e-05,
183
+ "loss": 2.0902,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 0.7147569826259071,
188
+ "grad_norm": 3.3778562545776367,
189
+ "learning_rate": 1.537844036697248e-05,
190
+ "loss": 2.0536,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 0.7422476358038267,
195
+ "grad_norm": 2.8149023056030273,
196
+ "learning_rate": 1.519347440071027e-05,
197
+ "loss": 2.1065,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 0.7697382889817462,
202
+ "grad_norm": 2.860196590423584,
203
+ "learning_rate": 1.5008508434448063e-05,
204
+ "loss": 2.0705,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 0.7972289421596657,
209
+ "grad_norm": 3.2587316036224365,
210
+ "learning_rate": 1.4823542468185853e-05,
211
+ "loss": 2.1082,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 0.8247195953375852,
216
+ "grad_norm": 2.4807844161987305,
217
+ "learning_rate": 1.4638576501923647e-05,
218
+ "loss": 2.0576,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 0.8522102485155048,
223
+ "grad_norm": 3.040374755859375,
224
+ "learning_rate": 1.445361053566144e-05,
225
+ "loss": 2.0694,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 0.8797009016934242,
230
+ "grad_norm": 3.2850606441497803,
231
+ "learning_rate": 1.4269014501331757e-05,
232
+ "loss": 2.081,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 0.9071915548713437,
237
+ "grad_norm": 3.131565570831299,
238
+ "learning_rate": 1.4084048535069548e-05,
239
+ "loss": 2.0567,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 0.9346822080492633,
244
+ "grad_norm": 2.221071481704712,
245
+ "learning_rate": 1.389908256880734e-05,
246
+ "loss": 2.0758,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 0.9621728612271827,
251
+ "grad_norm": 2.6275980472564697,
252
+ "learning_rate": 1.3714116602545134e-05,
253
+ "loss": 2.0495,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 0.9896635144051023,
258
+ "grad_norm": 3.2267441749572754,
259
+ "learning_rate": 1.3529150636282926e-05,
260
+ "loss": 2.0287,
261
+ "step": 18000
262
+ }
263
+ ],
264
+ "logging_steps": 500,
265
+ "max_steps": 54564,
266
+ "num_input_tokens_seen": 0,
267
+ "num_train_epochs": 3,
268
+ "save_steps": 1818,
269
+ "stateful_callbacks": {
270
+ "TrainerControl": {
271
+ "args": {
272
+ "should_epoch_stop": false,
273
+ "should_evaluate": false,
274
+ "should_log": false,
275
+ "should_save": true,
276
+ "should_training_stop": false
277
+ },
278
+ "attributes": {}
279
+ }
280
+ },
281
+ "total_flos": 4750289141760000.0,
282
+ "train_batch_size": 4,
283
+ "trial_name": null,
284
+ "trial_params": null
285
+ }
base/100e1/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/100e2/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/100e2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/100e2/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/100e2/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/100e2/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/100e2/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/100e2/trainer_state.json ADDED
@@ -0,0 +1,537 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.9991202990983066,
5
+ "eval_steps": 500,
6
+ "global_step": 36360,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.10996261271167802,
34
+ "grad_norm": 6.73529577255249,
35
+ "learning_rate": 1.9446951760876e-05,
36
+ "loss": 2.4492,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.13745326588959753,
41
+ "grad_norm": 6.306326866149902,
42
+ "learning_rate": 1.9261985794613793e-05,
43
+ "loss": 2.4463,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.16494391906751704,
48
+ "grad_norm": 5.651560306549072,
49
+ "learning_rate": 1.9077019828351584e-05,
50
+ "loss": 2.3291,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.19243457224543656,
55
+ "grad_norm": 4.0863471031188965,
56
+ "learning_rate": 1.8892053862089378e-05,
57
+ "loss": 2.3461,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.21992522542335605,
62
+ "grad_norm": 4.746509075164795,
63
+ "learning_rate": 1.8707087895827168e-05,
64
+ "loss": 2.2806,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.24741587860127556,
69
+ "grad_norm": 4.339619159698486,
70
+ "learning_rate": 1.8522121929564962e-05,
71
+ "loss": 2.2746,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.27490653177919505,
76
+ "grad_norm": 4.018620014190674,
77
+ "learning_rate": 1.8337155963302752e-05,
78
+ "loss": 2.2403,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.3023971849571146,
83
+ "grad_norm": 4.380566120147705,
84
+ "learning_rate": 1.8152189997040546e-05,
85
+ "loss": 2.226,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.3298878381350341,
90
+ "grad_norm": 4.302157402038574,
91
+ "learning_rate": 1.796722403077834e-05,
92
+ "loss": 2.2119,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.3573784913129536,
97
+ "grad_norm": 4.434130668640137,
98
+ "learning_rate": 1.778225806451613e-05,
99
+ "loss": 2.2138,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.3848691444908731,
104
+ "grad_norm": 3.770848035812378,
105
+ "learning_rate": 1.759729209825392e-05,
106
+ "loss": 2.2103,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.4123597976687926,
111
+ "grad_norm": 2.9288535118103027,
112
+ "learning_rate": 1.7412326131991715e-05,
113
+ "loss": 2.1397,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.4398504508467121,
118
+ "grad_norm": 3.2082479000091553,
119
+ "learning_rate": 1.722736016572951e-05,
120
+ "loss": 2.2021,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.46734110402463164,
125
+ "grad_norm": 3.6131439208984375,
126
+ "learning_rate": 1.70423941994673e-05,
127
+ "loss": 2.2106,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.4948317572025511,
132
+ "grad_norm": 2.992415189743042,
133
+ "learning_rate": 1.6857798165137616e-05,
134
+ "loss": 2.1246,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.5223224103804707,
139
+ "grad_norm": 3.4553306102752686,
140
+ "learning_rate": 1.667283219887541e-05,
141
+ "loss": 2.0973,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.5498130635583901,
146
+ "grad_norm": 2.82578182220459,
147
+ "learning_rate": 1.64878662326132e-05,
148
+ "loss": 2.1257,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.5773037167363096,
153
+ "grad_norm": 2.674267292022705,
154
+ "learning_rate": 1.630290026635099e-05,
155
+ "loss": 2.13,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.6047943699142292,
160
+ "grad_norm": 3.27205491065979,
161
+ "learning_rate": 1.6117934300088784e-05,
162
+ "loss": 2.1855,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 0.6322850230921486,
167
+ "grad_norm": 3.042849540710449,
168
+ "learning_rate": 1.5932968333826578e-05,
169
+ "loss": 2.1532,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 0.6597756762700682,
174
+ "grad_norm": 3.618957042694092,
175
+ "learning_rate": 1.5748372299496894e-05,
176
+ "loss": 2.1202,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 0.6872663294479877,
181
+ "grad_norm": 3.6390328407287598,
182
+ "learning_rate": 1.5563406333234685e-05,
183
+ "loss": 2.0902,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 0.7147569826259071,
188
+ "grad_norm": 3.3778562545776367,
189
+ "learning_rate": 1.537844036697248e-05,
190
+ "loss": 2.0536,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 0.7422476358038267,
195
+ "grad_norm": 2.8149023056030273,
196
+ "learning_rate": 1.519347440071027e-05,
197
+ "loss": 2.1065,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 0.7697382889817462,
202
+ "grad_norm": 2.860196590423584,
203
+ "learning_rate": 1.5008508434448063e-05,
204
+ "loss": 2.0705,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 0.7972289421596657,
209
+ "grad_norm": 3.2587316036224365,
210
+ "learning_rate": 1.4823542468185853e-05,
211
+ "loss": 2.1082,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 0.8247195953375852,
216
+ "grad_norm": 2.4807844161987305,
217
+ "learning_rate": 1.4638576501923647e-05,
218
+ "loss": 2.0576,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 0.8522102485155048,
223
+ "grad_norm": 3.040374755859375,
224
+ "learning_rate": 1.445361053566144e-05,
225
+ "loss": 2.0694,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 0.8797009016934242,
230
+ "grad_norm": 3.2850606441497803,
231
+ "learning_rate": 1.4269014501331757e-05,
232
+ "loss": 2.081,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 0.9071915548713437,
237
+ "grad_norm": 3.131565570831299,
238
+ "learning_rate": 1.4084048535069548e-05,
239
+ "loss": 2.0567,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 0.9346822080492633,
244
+ "grad_norm": 2.221071481704712,
245
+ "learning_rate": 1.389908256880734e-05,
246
+ "loss": 2.0758,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 0.9621728612271827,
251
+ "grad_norm": 2.6275980472564697,
252
+ "learning_rate": 1.3714116602545134e-05,
253
+ "loss": 2.0495,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 0.9896635144051023,
258
+ "grad_norm": 3.2267441749572754,
259
+ "learning_rate": 1.3529150636282926e-05,
260
+ "loss": 2.0287,
261
+ "step": 18000
262
+ },
263
+ {
264
+ "epoch": 1.0171541675830218,
265
+ "grad_norm": 3.0593459606170654,
266
+ "learning_rate": 1.334455460195324e-05,
267
+ "loss": 2.0028,
268
+ "step": 18500
269
+ },
270
+ {
271
+ "epoch": 1.0446448207609413,
272
+ "grad_norm": 2.9170665740966797,
273
+ "learning_rate": 1.3159588635691035e-05,
274
+ "loss": 1.9644,
275
+ "step": 19000
276
+ },
277
+ {
278
+ "epoch": 1.0721354739388609,
279
+ "grad_norm": 3.075544834136963,
280
+ "learning_rate": 1.2974622669428827e-05,
281
+ "loss": 1.9723,
282
+ "step": 19500
283
+ },
284
+ {
285
+ "epoch": 1.0996261271167802,
286
+ "grad_norm": 2.9603986740112305,
287
+ "learning_rate": 1.2789656703166617e-05,
288
+ "loss": 1.997,
289
+ "step": 20000
290
+ },
291
+ {
292
+ "epoch": 1.1271167802946997,
293
+ "grad_norm": 2.951988458633423,
294
+ "learning_rate": 1.2605060668836935e-05,
295
+ "loss": 1.9418,
296
+ "step": 20500
297
+ },
298
+ {
299
+ "epoch": 1.1546074334726193,
300
+ "grad_norm": 3.427750587463379,
301
+ "learning_rate": 1.2420094702574727e-05,
302
+ "loss": 1.9553,
303
+ "step": 21000
304
+ },
305
+ {
306
+ "epoch": 1.1820980866505388,
307
+ "grad_norm": 2.6535024642944336,
308
+ "learning_rate": 1.223512873631252e-05,
309
+ "loss": 2.0072,
310
+ "step": 21500
311
+ },
312
+ {
313
+ "epoch": 1.2095887398284584,
314
+ "grad_norm": 2.9090585708618164,
315
+ "learning_rate": 1.2050162770050312e-05,
316
+ "loss": 1.9705,
317
+ "step": 22000
318
+ },
319
+ {
320
+ "epoch": 1.237079393006378,
321
+ "grad_norm": 3.2485570907592773,
322
+ "learning_rate": 1.1865196803788104e-05,
323
+ "loss": 1.9543,
324
+ "step": 22500
325
+ },
326
+ {
327
+ "epoch": 1.2645700461842972,
328
+ "grad_norm": 3.0418505668640137,
329
+ "learning_rate": 1.168060076945842e-05,
330
+ "loss": 2.0092,
331
+ "step": 23000
332
+ },
333
+ {
334
+ "epoch": 1.2920606993622168,
335
+ "grad_norm": 3.1117002964019775,
336
+ "learning_rate": 1.1495634803196214e-05,
337
+ "loss": 1.9702,
338
+ "step": 23500
339
+ },
340
+ {
341
+ "epoch": 1.3195513525401363,
342
+ "grad_norm": 2.928476095199585,
343
+ "learning_rate": 1.1310668836934005e-05,
344
+ "loss": 1.9813,
345
+ "step": 24000
346
+ },
347
+ {
348
+ "epoch": 1.3470420057180559,
349
+ "grad_norm": 3.026412010192871,
350
+ "learning_rate": 1.1125702870671797e-05,
351
+ "loss": 1.9192,
352
+ "step": 24500
353
+ },
354
+ {
355
+ "epoch": 1.3745326588959754,
356
+ "grad_norm": 3.093477249145508,
357
+ "learning_rate": 1.094073690440959e-05,
358
+ "loss": 1.9387,
359
+ "step": 25000
360
+ },
361
+ {
362
+ "epoch": 1.402023312073895,
363
+ "grad_norm": 2.920074701309204,
364
+ "learning_rate": 1.0756140870079905e-05,
365
+ "loss": 1.9651,
366
+ "step": 25500
367
+ },
368
+ {
369
+ "epoch": 1.4295139652518145,
370
+ "grad_norm": 2.5337624549865723,
371
+ "learning_rate": 1.0571174903817697e-05,
372
+ "loss": 1.9262,
373
+ "step": 26000
374
+ },
375
+ {
376
+ "epoch": 1.4570046184297338,
377
+ "grad_norm": 2.506801128387451,
378
+ "learning_rate": 1.0386208937555491e-05,
379
+ "loss": 1.9542,
380
+ "step": 26500
381
+ },
382
+ {
383
+ "epoch": 1.4844952716076534,
384
+ "grad_norm": 2.7995145320892334,
385
+ "learning_rate": 1.0201242971293283e-05,
386
+ "loss": 1.9609,
387
+ "step": 27000
388
+ },
389
+ {
390
+ "epoch": 1.511985924785573,
391
+ "grad_norm": 4.121292591094971,
392
+ "learning_rate": 1.0016646936963598e-05,
393
+ "loss": 1.9317,
394
+ "step": 27500
395
+ },
396
+ {
397
+ "epoch": 1.5394765779634925,
398
+ "grad_norm": 2.802678346633911,
399
+ "learning_rate": 9.831680970701392e-06,
400
+ "loss": 1.9519,
401
+ "step": 28000
402
+ },
403
+ {
404
+ "epoch": 1.5669672311414118,
405
+ "grad_norm": 2.965937852859497,
406
+ "learning_rate": 9.646715004439184e-06,
407
+ "loss": 1.9401,
408
+ "step": 28500
409
+ },
410
+ {
411
+ "epoch": 1.5944578843193313,
412
+ "grad_norm": 2.364539861679077,
413
+ "learning_rate": 9.461749038176976e-06,
414
+ "loss": 1.9686,
415
+ "step": 29000
416
+ },
417
+ {
418
+ "epoch": 1.6219485374972509,
419
+ "grad_norm": 2.8087072372436523,
420
+ "learning_rate": 9.276783071914768e-06,
421
+ "loss": 1.9046,
422
+ "step": 29500
423
+ },
424
+ {
425
+ "epoch": 1.6494391906751704,
426
+ "grad_norm": 3.0667874813079834,
427
+ "learning_rate": 9.092187037585085e-06,
428
+ "loss": 1.8992,
429
+ "step": 30000
430
+ },
431
+ {
432
+ "epoch": 1.67692984385309,
433
+ "grad_norm": 2.4543378353118896,
434
+ "learning_rate": 8.907221071322877e-06,
435
+ "loss": 1.8936,
436
+ "step": 30500
437
+ },
438
+ {
439
+ "epoch": 1.7044204970310095,
440
+ "grad_norm": 2.695256233215332,
441
+ "learning_rate": 8.72225510506067e-06,
442
+ "loss": 1.9429,
443
+ "step": 31000
444
+ },
445
+ {
446
+ "epoch": 1.731911150208929,
447
+ "grad_norm": 2.377511978149414,
448
+ "learning_rate": 8.537289138798461e-06,
449
+ "loss": 1.9787,
450
+ "step": 31500
451
+ },
452
+ {
453
+ "epoch": 1.7594018033868486,
454
+ "grad_norm": 2.74446964263916,
455
+ "learning_rate": 8.35269310446878e-06,
456
+ "loss": 1.9272,
457
+ "step": 32000
458
+ },
459
+ {
460
+ "epoch": 1.7868924565647681,
461
+ "grad_norm": 3.174065589904785,
462
+ "learning_rate": 8.16772713820657e-06,
463
+ "loss": 1.9744,
464
+ "step": 32500
465
+ },
466
+ {
467
+ "epoch": 1.8143831097426875,
468
+ "grad_norm": 3.6638684272766113,
469
+ "learning_rate": 7.982761171944364e-06,
470
+ "loss": 1.9441,
471
+ "step": 33000
472
+ },
473
+ {
474
+ "epoch": 1.841873762920607,
475
+ "grad_norm": 2.563633441925049,
476
+ "learning_rate": 7.797795205682154e-06,
477
+ "loss": 1.9422,
478
+ "step": 33500
479
+ },
480
+ {
481
+ "epoch": 1.8693644160985265,
482
+ "grad_norm": 2.817347764968872,
483
+ "learning_rate": 7.612829239419948e-06,
484
+ "loss": 1.9556,
485
+ "step": 34000
486
+ },
487
+ {
488
+ "epoch": 1.8968550692764459,
489
+ "grad_norm": 2.59664249420166,
490
+ "learning_rate": 7.428233205090264e-06,
491
+ "loss": 1.9754,
492
+ "step": 34500
493
+ },
494
+ {
495
+ "epoch": 1.9243457224543654,
496
+ "grad_norm": 2.615318775177002,
497
+ "learning_rate": 7.2432672388280566e-06,
498
+ "loss": 1.9349,
499
+ "step": 35000
500
+ },
501
+ {
502
+ "epoch": 1.951836375632285,
503
+ "grad_norm": 2.5411360263824463,
504
+ "learning_rate": 7.058301272565849e-06,
505
+ "loss": 1.9612,
506
+ "step": 35500
507
+ },
508
+ {
509
+ "epoch": 1.9793270288102045,
510
+ "grad_norm": 2.809213161468506,
511
+ "learning_rate": 6.873335306303641e-06,
512
+ "loss": 1.9433,
513
+ "step": 36000
514
+ }
515
+ ],
516
+ "logging_steps": 500,
517
+ "max_steps": 54564,
518
+ "num_input_tokens_seen": 0,
519
+ "num_train_epochs": 3,
520
+ "save_steps": 1818,
521
+ "stateful_callbacks": {
522
+ "TrainerControl": {
523
+ "args": {
524
+ "should_epoch_stop": false,
525
+ "should_evaluate": false,
526
+ "should_log": false,
527
+ "should_save": true,
528
+ "should_training_stop": false
529
+ },
530
+ "attributes": {}
531
+ }
532
+ },
533
+ "total_flos": 9500447637504000.0,
534
+ "train_batch_size": 4,
535
+ "trial_name": null,
536
+ "trial_params": null
537
+ }
base/100e2/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/100e3/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/100e3/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/100e3/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/100e3/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/100e3/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/100e3/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/100e3/trainer_state.json ADDED
@@ -0,0 +1,796 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.99868044864746,
5
+ "eval_steps": 500,
6
+ "global_step": 54540,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.10996261271167802,
34
+ "grad_norm": 6.73529577255249,
35
+ "learning_rate": 1.9446951760876e-05,
36
+ "loss": 2.4492,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.13745326588959753,
41
+ "grad_norm": 6.306326866149902,
42
+ "learning_rate": 1.9261985794613793e-05,
43
+ "loss": 2.4463,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.16494391906751704,
48
+ "grad_norm": 5.651560306549072,
49
+ "learning_rate": 1.9077019828351584e-05,
50
+ "loss": 2.3291,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.19243457224543656,
55
+ "grad_norm": 4.0863471031188965,
56
+ "learning_rate": 1.8892053862089378e-05,
57
+ "loss": 2.3461,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.21992522542335605,
62
+ "grad_norm": 4.746509075164795,
63
+ "learning_rate": 1.8707087895827168e-05,
64
+ "loss": 2.2806,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.24741587860127556,
69
+ "grad_norm": 4.339619159698486,
70
+ "learning_rate": 1.8522121929564962e-05,
71
+ "loss": 2.2746,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.27490653177919505,
76
+ "grad_norm": 4.018620014190674,
77
+ "learning_rate": 1.8337155963302752e-05,
78
+ "loss": 2.2403,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.3023971849571146,
83
+ "grad_norm": 4.380566120147705,
84
+ "learning_rate": 1.8152189997040546e-05,
85
+ "loss": 2.226,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.3298878381350341,
90
+ "grad_norm": 4.302157402038574,
91
+ "learning_rate": 1.796722403077834e-05,
92
+ "loss": 2.2119,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.3573784913129536,
97
+ "grad_norm": 4.434130668640137,
98
+ "learning_rate": 1.778225806451613e-05,
99
+ "loss": 2.2138,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.3848691444908731,
104
+ "grad_norm": 3.770848035812378,
105
+ "learning_rate": 1.759729209825392e-05,
106
+ "loss": 2.2103,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.4123597976687926,
111
+ "grad_norm": 2.9288535118103027,
112
+ "learning_rate": 1.7412326131991715e-05,
113
+ "loss": 2.1397,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.4398504508467121,
118
+ "grad_norm": 3.2082479000091553,
119
+ "learning_rate": 1.722736016572951e-05,
120
+ "loss": 2.2021,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.46734110402463164,
125
+ "grad_norm": 3.6131439208984375,
126
+ "learning_rate": 1.70423941994673e-05,
127
+ "loss": 2.2106,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.4948317572025511,
132
+ "grad_norm": 2.992415189743042,
133
+ "learning_rate": 1.6857798165137616e-05,
134
+ "loss": 2.1246,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.5223224103804707,
139
+ "grad_norm": 3.4553306102752686,
140
+ "learning_rate": 1.667283219887541e-05,
141
+ "loss": 2.0973,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.5498130635583901,
146
+ "grad_norm": 2.82578182220459,
147
+ "learning_rate": 1.64878662326132e-05,
148
+ "loss": 2.1257,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.5773037167363096,
153
+ "grad_norm": 2.674267292022705,
154
+ "learning_rate": 1.630290026635099e-05,
155
+ "loss": 2.13,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.6047943699142292,
160
+ "grad_norm": 3.27205491065979,
161
+ "learning_rate": 1.6117934300088784e-05,
162
+ "loss": 2.1855,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 0.6322850230921486,
167
+ "grad_norm": 3.042849540710449,
168
+ "learning_rate": 1.5932968333826578e-05,
169
+ "loss": 2.1532,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 0.6597756762700682,
174
+ "grad_norm": 3.618957042694092,
175
+ "learning_rate": 1.5748372299496894e-05,
176
+ "loss": 2.1202,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 0.6872663294479877,
181
+ "grad_norm": 3.6390328407287598,
182
+ "learning_rate": 1.5563406333234685e-05,
183
+ "loss": 2.0902,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 0.7147569826259071,
188
+ "grad_norm": 3.3778562545776367,
189
+ "learning_rate": 1.537844036697248e-05,
190
+ "loss": 2.0536,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 0.7422476358038267,
195
+ "grad_norm": 2.8149023056030273,
196
+ "learning_rate": 1.519347440071027e-05,
197
+ "loss": 2.1065,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 0.7697382889817462,
202
+ "grad_norm": 2.860196590423584,
203
+ "learning_rate": 1.5008508434448063e-05,
204
+ "loss": 2.0705,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 0.7972289421596657,
209
+ "grad_norm": 3.2587316036224365,
210
+ "learning_rate": 1.4823542468185853e-05,
211
+ "loss": 2.1082,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 0.8247195953375852,
216
+ "grad_norm": 2.4807844161987305,
217
+ "learning_rate": 1.4638576501923647e-05,
218
+ "loss": 2.0576,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 0.8522102485155048,
223
+ "grad_norm": 3.040374755859375,
224
+ "learning_rate": 1.445361053566144e-05,
225
+ "loss": 2.0694,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 0.8797009016934242,
230
+ "grad_norm": 3.2850606441497803,
231
+ "learning_rate": 1.4269014501331757e-05,
232
+ "loss": 2.081,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 0.9071915548713437,
237
+ "grad_norm": 3.131565570831299,
238
+ "learning_rate": 1.4084048535069548e-05,
239
+ "loss": 2.0567,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 0.9346822080492633,
244
+ "grad_norm": 2.221071481704712,
245
+ "learning_rate": 1.389908256880734e-05,
246
+ "loss": 2.0758,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 0.9621728612271827,
251
+ "grad_norm": 2.6275980472564697,
252
+ "learning_rate": 1.3714116602545134e-05,
253
+ "loss": 2.0495,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 0.9896635144051023,
258
+ "grad_norm": 3.2267441749572754,
259
+ "learning_rate": 1.3529150636282926e-05,
260
+ "loss": 2.0287,
261
+ "step": 18000
262
+ },
263
+ {
264
+ "epoch": 1.0171541675830218,
265
+ "grad_norm": 3.0593459606170654,
266
+ "learning_rate": 1.334455460195324e-05,
267
+ "loss": 2.0028,
268
+ "step": 18500
269
+ },
270
+ {
271
+ "epoch": 1.0446448207609413,
272
+ "grad_norm": 2.9170665740966797,
273
+ "learning_rate": 1.3159588635691035e-05,
274
+ "loss": 1.9644,
275
+ "step": 19000
276
+ },
277
+ {
278
+ "epoch": 1.0721354739388609,
279
+ "grad_norm": 3.075544834136963,
280
+ "learning_rate": 1.2974622669428827e-05,
281
+ "loss": 1.9723,
282
+ "step": 19500
283
+ },
284
+ {
285
+ "epoch": 1.0996261271167802,
286
+ "grad_norm": 2.9603986740112305,
287
+ "learning_rate": 1.2789656703166617e-05,
288
+ "loss": 1.997,
289
+ "step": 20000
290
+ },
291
+ {
292
+ "epoch": 1.1271167802946997,
293
+ "grad_norm": 2.951988458633423,
294
+ "learning_rate": 1.2605060668836935e-05,
295
+ "loss": 1.9418,
296
+ "step": 20500
297
+ },
298
+ {
299
+ "epoch": 1.1546074334726193,
300
+ "grad_norm": 3.427750587463379,
301
+ "learning_rate": 1.2420094702574727e-05,
302
+ "loss": 1.9553,
303
+ "step": 21000
304
+ },
305
+ {
306
+ "epoch": 1.1820980866505388,
307
+ "grad_norm": 2.6535024642944336,
308
+ "learning_rate": 1.223512873631252e-05,
309
+ "loss": 2.0072,
310
+ "step": 21500
311
+ },
312
+ {
313
+ "epoch": 1.2095887398284584,
314
+ "grad_norm": 2.9090585708618164,
315
+ "learning_rate": 1.2050162770050312e-05,
316
+ "loss": 1.9705,
317
+ "step": 22000
318
+ },
319
+ {
320
+ "epoch": 1.237079393006378,
321
+ "grad_norm": 3.2485570907592773,
322
+ "learning_rate": 1.1865196803788104e-05,
323
+ "loss": 1.9543,
324
+ "step": 22500
325
+ },
326
+ {
327
+ "epoch": 1.2645700461842972,
328
+ "grad_norm": 3.0418505668640137,
329
+ "learning_rate": 1.168060076945842e-05,
330
+ "loss": 2.0092,
331
+ "step": 23000
332
+ },
333
+ {
334
+ "epoch": 1.2920606993622168,
335
+ "grad_norm": 3.1117002964019775,
336
+ "learning_rate": 1.1495634803196214e-05,
337
+ "loss": 1.9702,
338
+ "step": 23500
339
+ },
340
+ {
341
+ "epoch": 1.3195513525401363,
342
+ "grad_norm": 2.928476095199585,
343
+ "learning_rate": 1.1310668836934005e-05,
344
+ "loss": 1.9813,
345
+ "step": 24000
346
+ },
347
+ {
348
+ "epoch": 1.3470420057180559,
349
+ "grad_norm": 3.026412010192871,
350
+ "learning_rate": 1.1125702870671797e-05,
351
+ "loss": 1.9192,
352
+ "step": 24500
353
+ },
354
+ {
355
+ "epoch": 1.3745326588959754,
356
+ "grad_norm": 3.093477249145508,
357
+ "learning_rate": 1.094073690440959e-05,
358
+ "loss": 1.9387,
359
+ "step": 25000
360
+ },
361
+ {
362
+ "epoch": 1.402023312073895,
363
+ "grad_norm": 2.920074701309204,
364
+ "learning_rate": 1.0756140870079905e-05,
365
+ "loss": 1.9651,
366
+ "step": 25500
367
+ },
368
+ {
369
+ "epoch": 1.4295139652518145,
370
+ "grad_norm": 2.5337624549865723,
371
+ "learning_rate": 1.0571174903817697e-05,
372
+ "loss": 1.9262,
373
+ "step": 26000
374
+ },
375
+ {
376
+ "epoch": 1.4570046184297338,
377
+ "grad_norm": 2.506801128387451,
378
+ "learning_rate": 1.0386208937555491e-05,
379
+ "loss": 1.9542,
380
+ "step": 26500
381
+ },
382
+ {
383
+ "epoch": 1.4844952716076534,
384
+ "grad_norm": 2.7995145320892334,
385
+ "learning_rate": 1.0201242971293283e-05,
386
+ "loss": 1.9609,
387
+ "step": 27000
388
+ },
389
+ {
390
+ "epoch": 1.511985924785573,
391
+ "grad_norm": 4.121292591094971,
392
+ "learning_rate": 1.0016646936963598e-05,
393
+ "loss": 1.9317,
394
+ "step": 27500
395
+ },
396
+ {
397
+ "epoch": 1.5394765779634925,
398
+ "grad_norm": 2.802678346633911,
399
+ "learning_rate": 9.831680970701392e-06,
400
+ "loss": 1.9519,
401
+ "step": 28000
402
+ },
403
+ {
404
+ "epoch": 1.5669672311414118,
405
+ "grad_norm": 2.965937852859497,
406
+ "learning_rate": 9.646715004439184e-06,
407
+ "loss": 1.9401,
408
+ "step": 28500
409
+ },
410
+ {
411
+ "epoch": 1.5944578843193313,
412
+ "grad_norm": 2.364539861679077,
413
+ "learning_rate": 9.461749038176976e-06,
414
+ "loss": 1.9686,
415
+ "step": 29000
416
+ },
417
+ {
418
+ "epoch": 1.6219485374972509,
419
+ "grad_norm": 2.8087072372436523,
420
+ "learning_rate": 9.276783071914768e-06,
421
+ "loss": 1.9046,
422
+ "step": 29500
423
+ },
424
+ {
425
+ "epoch": 1.6494391906751704,
426
+ "grad_norm": 3.0667874813079834,
427
+ "learning_rate": 9.092187037585085e-06,
428
+ "loss": 1.8992,
429
+ "step": 30000
430
+ },
431
+ {
432
+ "epoch": 1.67692984385309,
433
+ "grad_norm": 2.4543378353118896,
434
+ "learning_rate": 8.907221071322877e-06,
435
+ "loss": 1.8936,
436
+ "step": 30500
437
+ },
438
+ {
439
+ "epoch": 1.7044204970310095,
440
+ "grad_norm": 2.695256233215332,
441
+ "learning_rate": 8.72225510506067e-06,
442
+ "loss": 1.9429,
443
+ "step": 31000
444
+ },
445
+ {
446
+ "epoch": 1.731911150208929,
447
+ "grad_norm": 2.377511978149414,
448
+ "learning_rate": 8.537289138798461e-06,
449
+ "loss": 1.9787,
450
+ "step": 31500
451
+ },
452
+ {
453
+ "epoch": 1.7594018033868486,
454
+ "grad_norm": 2.74446964263916,
455
+ "learning_rate": 8.35269310446878e-06,
456
+ "loss": 1.9272,
457
+ "step": 32000
458
+ },
459
+ {
460
+ "epoch": 1.7868924565647681,
461
+ "grad_norm": 3.174065589904785,
462
+ "learning_rate": 8.16772713820657e-06,
463
+ "loss": 1.9744,
464
+ "step": 32500
465
+ },
466
+ {
467
+ "epoch": 1.8143831097426875,
468
+ "grad_norm": 3.6638684272766113,
469
+ "learning_rate": 7.982761171944364e-06,
470
+ "loss": 1.9441,
471
+ "step": 33000
472
+ },
473
+ {
474
+ "epoch": 1.841873762920607,
475
+ "grad_norm": 2.563633441925049,
476
+ "learning_rate": 7.797795205682154e-06,
477
+ "loss": 1.9422,
478
+ "step": 33500
479
+ },
480
+ {
481
+ "epoch": 1.8693644160985265,
482
+ "grad_norm": 2.817347764968872,
483
+ "learning_rate": 7.612829239419948e-06,
484
+ "loss": 1.9556,
485
+ "step": 34000
486
+ },
487
+ {
488
+ "epoch": 1.8968550692764459,
489
+ "grad_norm": 2.59664249420166,
490
+ "learning_rate": 7.428233205090264e-06,
491
+ "loss": 1.9754,
492
+ "step": 34500
493
+ },
494
+ {
495
+ "epoch": 1.9243457224543654,
496
+ "grad_norm": 2.615318775177002,
497
+ "learning_rate": 7.2432672388280566e-06,
498
+ "loss": 1.9349,
499
+ "step": 35000
500
+ },
501
+ {
502
+ "epoch": 1.951836375632285,
503
+ "grad_norm": 2.5411360263824463,
504
+ "learning_rate": 7.058301272565849e-06,
505
+ "loss": 1.9612,
506
+ "step": 35500
507
+ },
508
+ {
509
+ "epoch": 1.9793270288102045,
510
+ "grad_norm": 2.809213161468506,
511
+ "learning_rate": 6.873335306303641e-06,
512
+ "loss": 1.9433,
513
+ "step": 36000
514
+ },
515
+ {
516
+ "epoch": 2.006817681988124,
517
+ "grad_norm": 2.415309190750122,
518
+ "learning_rate": 6.688739271973957e-06,
519
+ "loss": 1.9222,
520
+ "step": 36500
521
+ },
522
+ {
523
+ "epoch": 2.0343083351660436,
524
+ "grad_norm": 2.3479654788970947,
525
+ "learning_rate": 6.5037733057117494e-06,
526
+ "loss": 1.8997,
527
+ "step": 37000
528
+ },
529
+ {
530
+ "epoch": 2.061798988343963,
531
+ "grad_norm": 2.566889762878418,
532
+ "learning_rate": 6.3188073394495416e-06,
533
+ "loss": 1.8604,
534
+ "step": 37500
535
+ },
536
+ {
537
+ "epoch": 2.0892896415218827,
538
+ "grad_norm": 2.9693121910095215,
539
+ "learning_rate": 6.133841373187334e-06,
540
+ "loss": 1.9029,
541
+ "step": 38000
542
+ },
543
+ {
544
+ "epoch": 2.1167802946998022,
545
+ "grad_norm": 3.270578145980835,
546
+ "learning_rate": 5.949245338857651e-06,
547
+ "loss": 1.8985,
548
+ "step": 38500
549
+ },
550
+ {
551
+ "epoch": 2.1442709478777218,
552
+ "grad_norm": 3.2845070362091064,
553
+ "learning_rate": 5.764279372595442e-06,
554
+ "loss": 1.8781,
555
+ "step": 39000
556
+ },
557
+ {
558
+ "epoch": 2.171761601055641,
559
+ "grad_norm": 3.1421778202056885,
560
+ "learning_rate": 5.579313406333235e-06,
561
+ "loss": 1.8677,
562
+ "step": 39500
563
+ },
564
+ {
565
+ "epoch": 2.1992522542335604,
566
+ "grad_norm": 2.8013062477111816,
567
+ "learning_rate": 5.394347440071027e-06,
568
+ "loss": 1.9078,
569
+ "step": 40000
570
+ },
571
+ {
572
+ "epoch": 2.22674290741148,
573
+ "grad_norm": 2.775585651397705,
574
+ "learning_rate": 5.209751405741345e-06,
575
+ "loss": 1.9292,
576
+ "step": 40500
577
+ },
578
+ {
579
+ "epoch": 2.2542335605893995,
580
+ "grad_norm": 2.672166585922241,
581
+ "learning_rate": 5.024785439479136e-06,
582
+ "loss": 1.8568,
583
+ "step": 41000
584
+ },
585
+ {
586
+ "epoch": 2.281724213767319,
587
+ "grad_norm": 2.4413459300994873,
588
+ "learning_rate": 4.839819473216928e-06,
589
+ "loss": 1.8813,
590
+ "step": 41500
591
+ },
592
+ {
593
+ "epoch": 2.3092148669452386,
594
+ "grad_norm": 3.139625072479248,
595
+ "learning_rate": 4.654853506954721e-06,
596
+ "loss": 1.8763,
597
+ "step": 42000
598
+ },
599
+ {
600
+ "epoch": 2.336705520123158,
601
+ "grad_norm": 2.938086986541748,
602
+ "learning_rate": 4.469887540692513e-06,
603
+ "loss": 1.9047,
604
+ "step": 42500
605
+ },
606
+ {
607
+ "epoch": 2.3641961733010777,
608
+ "grad_norm": 3.119866371154785,
609
+ "learning_rate": 4.28529150636283e-06,
610
+ "loss": 1.8985,
611
+ "step": 43000
612
+ },
613
+ {
614
+ "epoch": 2.391686826478997,
615
+ "grad_norm": 2.9585859775543213,
616
+ "learning_rate": 4.100325540100622e-06,
617
+ "loss": 1.8584,
618
+ "step": 43500
619
+ },
620
+ {
621
+ "epoch": 2.4191774796569168,
622
+ "grad_norm": 3.028029441833496,
623
+ "learning_rate": 3.915359573838414e-06,
624
+ "loss": 1.8514,
625
+ "step": 44000
626
+ },
627
+ {
628
+ "epoch": 2.4466681328348363,
629
+ "grad_norm": 2.885305643081665,
630
+ "learning_rate": 3.7303936075762065e-06,
631
+ "loss": 1.9058,
632
+ "step": 44500
633
+ },
634
+ {
635
+ "epoch": 2.474158786012756,
636
+ "grad_norm": 2.529195547103882,
637
+ "learning_rate": 3.5454276413139987e-06,
638
+ "loss": 1.899,
639
+ "step": 45000
640
+ },
641
+ {
642
+ "epoch": 2.5016494391906754,
643
+ "grad_norm": 2.8629934787750244,
644
+ "learning_rate": 3.360831606984315e-06,
645
+ "loss": 1.8854,
646
+ "step": 45500
647
+ },
648
+ {
649
+ "epoch": 2.5291400923685945,
650
+ "grad_norm": 3.1434662342071533,
651
+ "learning_rate": 3.1758656407221073e-06,
652
+ "loss": 1.8475,
653
+ "step": 46000
654
+ },
655
+ {
656
+ "epoch": 2.556630745546514,
657
+ "grad_norm": 2.7778401374816895,
658
+ "learning_rate": 2.9908996744598994e-06,
659
+ "loss": 1.893,
660
+ "step": 46500
661
+ },
662
+ {
663
+ "epoch": 2.5841213987244336,
664
+ "grad_norm": 3.024404525756836,
665
+ "learning_rate": 2.8059337081976915e-06,
666
+ "loss": 1.8792,
667
+ "step": 47000
668
+ },
669
+ {
670
+ "epoch": 2.611612051902353,
671
+ "grad_norm": 3.283930778503418,
672
+ "learning_rate": 2.6213376738680084e-06,
673
+ "loss": 1.8411,
674
+ "step": 47500
675
+ },
676
+ {
677
+ "epoch": 2.6391027050802727,
678
+ "grad_norm": 2.3540070056915283,
679
+ "learning_rate": 2.436371707605801e-06,
680
+ "loss": 1.8602,
681
+ "step": 48000
682
+ },
683
+ {
684
+ "epoch": 2.666593358258192,
685
+ "grad_norm": 2.4063596725463867,
686
+ "learning_rate": 2.251405741343593e-06,
687
+ "loss": 1.8422,
688
+ "step": 48500
689
+ },
690
+ {
691
+ "epoch": 2.6940840114361118,
692
+ "grad_norm": 2.7958314418792725,
693
+ "learning_rate": 2.0664397750813853e-06,
694
+ "loss": 1.8726,
695
+ "step": 49000
696
+ },
697
+ {
698
+ "epoch": 2.7215746646140313,
699
+ "grad_norm": 2.9718141555786133,
700
+ "learning_rate": 1.8814738088191776e-06,
701
+ "loss": 1.9078,
702
+ "step": 49500
703
+ },
704
+ {
705
+ "epoch": 2.749065317791951,
706
+ "grad_norm": 3.2070326805114746,
707
+ "learning_rate": 1.696877774489494e-06,
708
+ "loss": 1.8734,
709
+ "step": 50000
710
+ },
711
+ {
712
+ "epoch": 2.7765559709698704,
713
+ "grad_norm": 3.239140272140503,
714
+ "learning_rate": 1.5119118082272862e-06,
715
+ "loss": 1.8441,
716
+ "step": 50500
717
+ },
718
+ {
719
+ "epoch": 2.80404662414779,
720
+ "grad_norm": 2.4471523761749268,
721
+ "learning_rate": 1.3269458419650785e-06,
722
+ "loss": 1.8329,
723
+ "step": 51000
724
+ },
725
+ {
726
+ "epoch": 2.831537277325709,
727
+ "grad_norm": 2.8477437496185303,
728
+ "learning_rate": 1.1419798757028707e-06,
729
+ "loss": 1.8734,
730
+ "step": 51500
731
+ },
732
+ {
733
+ "epoch": 2.859027930503629,
734
+ "grad_norm": 2.9819397926330566,
735
+ "learning_rate": 9.573838413731873e-07,
736
+ "loss": 1.883,
737
+ "step": 52000
738
+ },
739
+ {
740
+ "epoch": 2.886518583681548,
741
+ "grad_norm": 2.5117220878601074,
742
+ "learning_rate": 7.724178751109797e-07,
743
+ "loss": 1.8841,
744
+ "step": 52500
745
+ },
746
+ {
747
+ "epoch": 2.9140092368594677,
748
+ "grad_norm": 3.4342823028564453,
749
+ "learning_rate": 5.874519088487718e-07,
750
+ "loss": 1.8441,
751
+ "step": 53000
752
+ },
753
+ {
754
+ "epoch": 2.941499890037387,
755
+ "grad_norm": 3.2193424701690674,
756
+ "learning_rate": 4.024859425865641e-07,
757
+ "loss": 1.8699,
758
+ "step": 53500
759
+ },
760
+ {
761
+ "epoch": 2.9689905432153068,
762
+ "grad_norm": 3.543229579925537,
763
+ "learning_rate": 2.175199763243563e-07,
764
+ "loss": 1.8415,
765
+ "step": 54000
766
+ },
767
+ {
768
+ "epoch": 2.9964811963932263,
769
+ "grad_norm": 3.1084017753601074,
770
+ "learning_rate": 3.292394199467298e-08,
771
+ "loss": 1.8803,
772
+ "step": 54500
773
+ }
774
+ ],
775
+ "logging_steps": 500,
776
+ "max_steps": 54564,
777
+ "num_input_tokens_seen": 0,
778
+ "num_train_epochs": 3,
779
+ "save_steps": 1818,
780
+ "stateful_callbacks": {
781
+ "TrainerControl": {
782
+ "args": {
783
+ "should_epoch_stop": false,
784
+ "should_evaluate": false,
785
+ "should_log": false,
786
+ "should_save": true,
787
+ "should_training_stop": false
788
+ },
789
+ "attributes": {}
790
+ }
791
+ },
792
+ "total_flos": 1.4250606133248e+16,
793
+ "train_batch_size": 4,
794
+ "trial_name": null,
795
+ "trial_params": null
796
+ }
base/100e3/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e1/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/10e1/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/10e1/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/10e1/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/10e1/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e1/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/10e1/trainer_state.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.09995601495491532,
5
+ "eval_steps": 500,
6
+ "global_step": 1818,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ }
32
+ ],
33
+ "logging_steps": 500,
34
+ "max_steps": 54564,
35
+ "num_input_tokens_seen": 0,
36
+ "num_train_epochs": 3,
37
+ "save_steps": 1818,
38
+ "stateful_callbacks": {
39
+ "TrainerControl": {
40
+ "args": {
41
+ "should_epoch_stop": false,
42
+ "should_evaluate": false,
43
+ "should_log": false,
44
+ "should_save": true,
45
+ "should_training_stop": false
46
+ },
47
+ "attributes": {}
48
+ }
49
+ },
50
+ "total_flos": 475028914176000.0,
51
+ "train_batch_size": 4,
52
+ "trial_name": null,
53
+ "trial_params": null
54
+ }
base/10e1/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e2/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/10e2/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/10e2/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/10e2/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/10e2/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e2/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/10e2/trainer_state.json ADDED
@@ -0,0 +1,306 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.0995161645040685,
5
+ "eval_steps": 500,
6
+ "global_step": 19998,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.10996261271167802,
34
+ "grad_norm": 6.73529577255249,
35
+ "learning_rate": 1.9446951760876e-05,
36
+ "loss": 2.4492,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.13745326588959753,
41
+ "grad_norm": 6.306326866149902,
42
+ "learning_rate": 1.9261985794613793e-05,
43
+ "loss": 2.4463,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.16494391906751704,
48
+ "grad_norm": 5.651560306549072,
49
+ "learning_rate": 1.9077019828351584e-05,
50
+ "loss": 2.3291,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.19243457224543656,
55
+ "grad_norm": 4.0863471031188965,
56
+ "learning_rate": 1.8892053862089378e-05,
57
+ "loss": 2.3461,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.21992522542335605,
62
+ "grad_norm": 4.746509075164795,
63
+ "learning_rate": 1.8707087895827168e-05,
64
+ "loss": 2.2806,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.24741587860127556,
69
+ "grad_norm": 4.339619159698486,
70
+ "learning_rate": 1.8522121929564962e-05,
71
+ "loss": 2.2746,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.27490653177919505,
76
+ "grad_norm": 4.018620014190674,
77
+ "learning_rate": 1.8337155963302752e-05,
78
+ "loss": 2.2403,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.3023971849571146,
83
+ "grad_norm": 4.380566120147705,
84
+ "learning_rate": 1.8152189997040546e-05,
85
+ "loss": 2.226,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.3298878381350341,
90
+ "grad_norm": 4.302157402038574,
91
+ "learning_rate": 1.796722403077834e-05,
92
+ "loss": 2.2119,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.3573784913129536,
97
+ "grad_norm": 4.434130668640137,
98
+ "learning_rate": 1.778225806451613e-05,
99
+ "loss": 2.2138,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.3848691444908731,
104
+ "grad_norm": 3.770848035812378,
105
+ "learning_rate": 1.759729209825392e-05,
106
+ "loss": 2.2103,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.4123597976687926,
111
+ "grad_norm": 2.9288535118103027,
112
+ "learning_rate": 1.7412326131991715e-05,
113
+ "loss": 2.1397,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.4398504508467121,
118
+ "grad_norm": 3.2082479000091553,
119
+ "learning_rate": 1.722736016572951e-05,
120
+ "loss": 2.2021,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.46734110402463164,
125
+ "grad_norm": 3.6131439208984375,
126
+ "learning_rate": 1.70423941994673e-05,
127
+ "loss": 2.2106,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.4948317572025511,
132
+ "grad_norm": 2.992415189743042,
133
+ "learning_rate": 1.6857798165137616e-05,
134
+ "loss": 2.1246,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.5223224103804707,
139
+ "grad_norm": 3.4553306102752686,
140
+ "learning_rate": 1.667283219887541e-05,
141
+ "loss": 2.0973,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.5498130635583901,
146
+ "grad_norm": 2.82578182220459,
147
+ "learning_rate": 1.64878662326132e-05,
148
+ "loss": 2.1257,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.5773037167363096,
153
+ "grad_norm": 2.674267292022705,
154
+ "learning_rate": 1.630290026635099e-05,
155
+ "loss": 2.13,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.6047943699142292,
160
+ "grad_norm": 3.27205491065979,
161
+ "learning_rate": 1.6117934300088784e-05,
162
+ "loss": 2.1855,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 0.6322850230921486,
167
+ "grad_norm": 3.042849540710449,
168
+ "learning_rate": 1.5932968333826578e-05,
169
+ "loss": 2.1532,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 0.6597756762700682,
174
+ "grad_norm": 3.618957042694092,
175
+ "learning_rate": 1.5748372299496894e-05,
176
+ "loss": 2.1202,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 0.6872663294479877,
181
+ "grad_norm": 3.6390328407287598,
182
+ "learning_rate": 1.5563406333234685e-05,
183
+ "loss": 2.0902,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 0.7147569826259071,
188
+ "grad_norm": 3.3778562545776367,
189
+ "learning_rate": 1.537844036697248e-05,
190
+ "loss": 2.0536,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 0.7422476358038267,
195
+ "grad_norm": 2.8149023056030273,
196
+ "learning_rate": 1.519347440071027e-05,
197
+ "loss": 2.1065,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 0.7697382889817462,
202
+ "grad_norm": 2.860196590423584,
203
+ "learning_rate": 1.5008508434448063e-05,
204
+ "loss": 2.0705,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 0.7972289421596657,
209
+ "grad_norm": 3.2587316036224365,
210
+ "learning_rate": 1.4823542468185853e-05,
211
+ "loss": 2.1082,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 0.8247195953375852,
216
+ "grad_norm": 2.4807844161987305,
217
+ "learning_rate": 1.4638576501923647e-05,
218
+ "loss": 2.0576,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 0.8522102485155048,
223
+ "grad_norm": 3.040374755859375,
224
+ "learning_rate": 1.445361053566144e-05,
225
+ "loss": 2.0694,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 0.8797009016934242,
230
+ "grad_norm": 3.2850606441497803,
231
+ "learning_rate": 1.4269014501331757e-05,
232
+ "loss": 2.081,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 0.9071915548713437,
237
+ "grad_norm": 3.131565570831299,
238
+ "learning_rate": 1.4084048535069548e-05,
239
+ "loss": 2.0567,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 0.9346822080492633,
244
+ "grad_norm": 2.221071481704712,
245
+ "learning_rate": 1.389908256880734e-05,
246
+ "loss": 2.0758,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 0.9621728612271827,
251
+ "grad_norm": 2.6275980472564697,
252
+ "learning_rate": 1.3714116602545134e-05,
253
+ "loss": 2.0495,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 0.9896635144051023,
258
+ "grad_norm": 3.2267441749572754,
259
+ "learning_rate": 1.3529150636282926e-05,
260
+ "loss": 2.0287,
261
+ "step": 18000
262
+ },
263
+ {
264
+ "epoch": 1.0171541675830218,
265
+ "grad_norm": 3.0593459606170654,
266
+ "learning_rate": 1.334455460195324e-05,
267
+ "loss": 2.0028,
268
+ "step": 18500
269
+ },
270
+ {
271
+ "epoch": 1.0446448207609413,
272
+ "grad_norm": 2.9170665740966797,
273
+ "learning_rate": 1.3159588635691035e-05,
274
+ "loss": 1.9644,
275
+ "step": 19000
276
+ },
277
+ {
278
+ "epoch": 1.0721354739388609,
279
+ "grad_norm": 3.075544834136963,
280
+ "learning_rate": 1.2974622669428827e-05,
281
+ "loss": 1.9723,
282
+ "step": 19500
283
+ }
284
+ ],
285
+ "logging_steps": 500,
286
+ "max_steps": 54564,
287
+ "num_input_tokens_seen": 0,
288
+ "num_train_epochs": 3,
289
+ "save_steps": 1818,
290
+ "stateful_callbacks": {
291
+ "TrainerControl": {
292
+ "args": {
293
+ "should_epoch_stop": false,
294
+ "should_evaluate": false,
295
+ "should_log": false,
296
+ "should_save": true,
297
+ "should_training_stop": false
298
+ },
299
+ "attributes": {}
300
+ }
301
+ },
302
+ "total_flos": 5225187409920000.0,
303
+ "train_batch_size": 4,
304
+ "trial_name": null,
305
+ "trial_params": null
306
+ }
base/10e2/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e3/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/10e3/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }
base/10e3/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
base/10e3/special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<|endoftext|>",
3
+ "eos_token": "<|endoftext|>",
4
+ "unk_token": "<|endoftext|>"
5
+ }
base/10e3/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
base/10e3/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": false,
15
+ "eos_token": "<|endoftext|>",
16
+ "model_max_length": 1024,
17
+ "return_tensors": "pt",
18
+ "tokenizer_class": "GPT2Tokenizer",
19
+ "unk_token": "<|endoftext|>"
20
+ }
base/10e3/trainer_state.json ADDED
@@ -0,0 +1,565 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 2.099076314053222,
5
+ "eval_steps": 500,
6
+ "global_step": 38178,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.027490653177919506,
13
+ "grad_norm": 7.844790935516357,
14
+ "learning_rate": 1.98e-05,
15
+ "loss": 3.3532,
16
+ "step": 500
17
+ },
18
+ {
19
+ "epoch": 0.05498130635583901,
20
+ "grad_norm": 7.924389362335205,
21
+ "learning_rate": 1.9816883693400415e-05,
22
+ "loss": 2.6056,
23
+ "step": 1000
24
+ },
25
+ {
26
+ "epoch": 0.08247195953375852,
27
+ "grad_norm": 7.497577667236328,
28
+ "learning_rate": 1.963191772713821e-05,
29
+ "loss": 2.5447,
30
+ "step": 1500
31
+ },
32
+ {
33
+ "epoch": 0.10996261271167802,
34
+ "grad_norm": 6.73529577255249,
35
+ "learning_rate": 1.9446951760876e-05,
36
+ "loss": 2.4492,
37
+ "step": 2000
38
+ },
39
+ {
40
+ "epoch": 0.13745326588959753,
41
+ "grad_norm": 6.306326866149902,
42
+ "learning_rate": 1.9261985794613793e-05,
43
+ "loss": 2.4463,
44
+ "step": 2500
45
+ },
46
+ {
47
+ "epoch": 0.16494391906751704,
48
+ "grad_norm": 5.651560306549072,
49
+ "learning_rate": 1.9077019828351584e-05,
50
+ "loss": 2.3291,
51
+ "step": 3000
52
+ },
53
+ {
54
+ "epoch": 0.19243457224543656,
55
+ "grad_norm": 4.0863471031188965,
56
+ "learning_rate": 1.8892053862089378e-05,
57
+ "loss": 2.3461,
58
+ "step": 3500
59
+ },
60
+ {
61
+ "epoch": 0.21992522542335605,
62
+ "grad_norm": 4.746509075164795,
63
+ "learning_rate": 1.8707087895827168e-05,
64
+ "loss": 2.2806,
65
+ "step": 4000
66
+ },
67
+ {
68
+ "epoch": 0.24741587860127556,
69
+ "grad_norm": 4.339619159698486,
70
+ "learning_rate": 1.8522121929564962e-05,
71
+ "loss": 2.2746,
72
+ "step": 4500
73
+ },
74
+ {
75
+ "epoch": 0.27490653177919505,
76
+ "grad_norm": 4.018620014190674,
77
+ "learning_rate": 1.8337155963302752e-05,
78
+ "loss": 2.2403,
79
+ "step": 5000
80
+ },
81
+ {
82
+ "epoch": 0.3023971849571146,
83
+ "grad_norm": 4.380566120147705,
84
+ "learning_rate": 1.8152189997040546e-05,
85
+ "loss": 2.226,
86
+ "step": 5500
87
+ },
88
+ {
89
+ "epoch": 0.3298878381350341,
90
+ "grad_norm": 4.302157402038574,
91
+ "learning_rate": 1.796722403077834e-05,
92
+ "loss": 2.2119,
93
+ "step": 6000
94
+ },
95
+ {
96
+ "epoch": 0.3573784913129536,
97
+ "grad_norm": 4.434130668640137,
98
+ "learning_rate": 1.778225806451613e-05,
99
+ "loss": 2.2138,
100
+ "step": 6500
101
+ },
102
+ {
103
+ "epoch": 0.3848691444908731,
104
+ "grad_norm": 3.770848035812378,
105
+ "learning_rate": 1.759729209825392e-05,
106
+ "loss": 2.2103,
107
+ "step": 7000
108
+ },
109
+ {
110
+ "epoch": 0.4123597976687926,
111
+ "grad_norm": 2.9288535118103027,
112
+ "learning_rate": 1.7412326131991715e-05,
113
+ "loss": 2.1397,
114
+ "step": 7500
115
+ },
116
+ {
117
+ "epoch": 0.4398504508467121,
118
+ "grad_norm": 3.2082479000091553,
119
+ "learning_rate": 1.722736016572951e-05,
120
+ "loss": 2.2021,
121
+ "step": 8000
122
+ },
123
+ {
124
+ "epoch": 0.46734110402463164,
125
+ "grad_norm": 3.6131439208984375,
126
+ "learning_rate": 1.70423941994673e-05,
127
+ "loss": 2.2106,
128
+ "step": 8500
129
+ },
130
+ {
131
+ "epoch": 0.4948317572025511,
132
+ "grad_norm": 2.992415189743042,
133
+ "learning_rate": 1.6857798165137616e-05,
134
+ "loss": 2.1246,
135
+ "step": 9000
136
+ },
137
+ {
138
+ "epoch": 0.5223224103804707,
139
+ "grad_norm": 3.4553306102752686,
140
+ "learning_rate": 1.667283219887541e-05,
141
+ "loss": 2.0973,
142
+ "step": 9500
143
+ },
144
+ {
145
+ "epoch": 0.5498130635583901,
146
+ "grad_norm": 2.82578182220459,
147
+ "learning_rate": 1.64878662326132e-05,
148
+ "loss": 2.1257,
149
+ "step": 10000
150
+ },
151
+ {
152
+ "epoch": 0.5773037167363096,
153
+ "grad_norm": 2.674267292022705,
154
+ "learning_rate": 1.630290026635099e-05,
155
+ "loss": 2.13,
156
+ "step": 10500
157
+ },
158
+ {
159
+ "epoch": 0.6047943699142292,
160
+ "grad_norm": 3.27205491065979,
161
+ "learning_rate": 1.6117934300088784e-05,
162
+ "loss": 2.1855,
163
+ "step": 11000
164
+ },
165
+ {
166
+ "epoch": 0.6322850230921486,
167
+ "grad_norm": 3.042849540710449,
168
+ "learning_rate": 1.5932968333826578e-05,
169
+ "loss": 2.1532,
170
+ "step": 11500
171
+ },
172
+ {
173
+ "epoch": 0.6597756762700682,
174
+ "grad_norm": 3.618957042694092,
175
+ "learning_rate": 1.5748372299496894e-05,
176
+ "loss": 2.1202,
177
+ "step": 12000
178
+ },
179
+ {
180
+ "epoch": 0.6872663294479877,
181
+ "grad_norm": 3.6390328407287598,
182
+ "learning_rate": 1.5563406333234685e-05,
183
+ "loss": 2.0902,
184
+ "step": 12500
185
+ },
186
+ {
187
+ "epoch": 0.7147569826259071,
188
+ "grad_norm": 3.3778562545776367,
189
+ "learning_rate": 1.537844036697248e-05,
190
+ "loss": 2.0536,
191
+ "step": 13000
192
+ },
193
+ {
194
+ "epoch": 0.7422476358038267,
195
+ "grad_norm": 2.8149023056030273,
196
+ "learning_rate": 1.519347440071027e-05,
197
+ "loss": 2.1065,
198
+ "step": 13500
199
+ },
200
+ {
201
+ "epoch": 0.7697382889817462,
202
+ "grad_norm": 2.860196590423584,
203
+ "learning_rate": 1.5008508434448063e-05,
204
+ "loss": 2.0705,
205
+ "step": 14000
206
+ },
207
+ {
208
+ "epoch": 0.7972289421596657,
209
+ "grad_norm": 3.2587316036224365,
210
+ "learning_rate": 1.4823542468185853e-05,
211
+ "loss": 2.1082,
212
+ "step": 14500
213
+ },
214
+ {
215
+ "epoch": 0.8247195953375852,
216
+ "grad_norm": 2.4807844161987305,
217
+ "learning_rate": 1.4638576501923647e-05,
218
+ "loss": 2.0576,
219
+ "step": 15000
220
+ },
221
+ {
222
+ "epoch": 0.8522102485155048,
223
+ "grad_norm": 3.040374755859375,
224
+ "learning_rate": 1.445361053566144e-05,
225
+ "loss": 2.0694,
226
+ "step": 15500
227
+ },
228
+ {
229
+ "epoch": 0.8797009016934242,
230
+ "grad_norm": 3.2850606441497803,
231
+ "learning_rate": 1.4269014501331757e-05,
232
+ "loss": 2.081,
233
+ "step": 16000
234
+ },
235
+ {
236
+ "epoch": 0.9071915548713437,
237
+ "grad_norm": 3.131565570831299,
238
+ "learning_rate": 1.4084048535069548e-05,
239
+ "loss": 2.0567,
240
+ "step": 16500
241
+ },
242
+ {
243
+ "epoch": 0.9346822080492633,
244
+ "grad_norm": 2.221071481704712,
245
+ "learning_rate": 1.389908256880734e-05,
246
+ "loss": 2.0758,
247
+ "step": 17000
248
+ },
249
+ {
250
+ "epoch": 0.9621728612271827,
251
+ "grad_norm": 2.6275980472564697,
252
+ "learning_rate": 1.3714116602545134e-05,
253
+ "loss": 2.0495,
254
+ "step": 17500
255
+ },
256
+ {
257
+ "epoch": 0.9896635144051023,
258
+ "grad_norm": 3.2267441749572754,
259
+ "learning_rate": 1.3529150636282926e-05,
260
+ "loss": 2.0287,
261
+ "step": 18000
262
+ },
263
+ {
264
+ "epoch": 1.0171541675830218,
265
+ "grad_norm": 3.0593459606170654,
266
+ "learning_rate": 1.334455460195324e-05,
267
+ "loss": 2.0028,
268
+ "step": 18500
269
+ },
270
+ {
271
+ "epoch": 1.0446448207609413,
272
+ "grad_norm": 2.9170665740966797,
273
+ "learning_rate": 1.3159588635691035e-05,
274
+ "loss": 1.9644,
275
+ "step": 19000
276
+ },
277
+ {
278
+ "epoch": 1.0721354739388609,
279
+ "grad_norm": 3.075544834136963,
280
+ "learning_rate": 1.2974622669428827e-05,
281
+ "loss": 1.9723,
282
+ "step": 19500
283
+ },
284
+ {
285
+ "epoch": 1.0996261271167802,
286
+ "grad_norm": 2.9603986740112305,
287
+ "learning_rate": 1.2789656703166617e-05,
288
+ "loss": 1.997,
289
+ "step": 20000
290
+ },
291
+ {
292
+ "epoch": 1.1271167802946997,
293
+ "grad_norm": 2.951988458633423,
294
+ "learning_rate": 1.2605060668836935e-05,
295
+ "loss": 1.9418,
296
+ "step": 20500
297
+ },
298
+ {
299
+ "epoch": 1.1546074334726193,
300
+ "grad_norm": 3.427750587463379,
301
+ "learning_rate": 1.2420094702574727e-05,
302
+ "loss": 1.9553,
303
+ "step": 21000
304
+ },
305
+ {
306
+ "epoch": 1.1820980866505388,
307
+ "grad_norm": 2.6535024642944336,
308
+ "learning_rate": 1.223512873631252e-05,
309
+ "loss": 2.0072,
310
+ "step": 21500
311
+ },
312
+ {
313
+ "epoch": 1.2095887398284584,
314
+ "grad_norm": 2.9090585708618164,
315
+ "learning_rate": 1.2050162770050312e-05,
316
+ "loss": 1.9705,
317
+ "step": 22000
318
+ },
319
+ {
320
+ "epoch": 1.237079393006378,
321
+ "grad_norm": 3.2485570907592773,
322
+ "learning_rate": 1.1865196803788104e-05,
323
+ "loss": 1.9543,
324
+ "step": 22500
325
+ },
326
+ {
327
+ "epoch": 1.2645700461842972,
328
+ "grad_norm": 3.0418505668640137,
329
+ "learning_rate": 1.168060076945842e-05,
330
+ "loss": 2.0092,
331
+ "step": 23000
332
+ },
333
+ {
334
+ "epoch": 1.2920606993622168,
335
+ "grad_norm": 3.1117002964019775,
336
+ "learning_rate": 1.1495634803196214e-05,
337
+ "loss": 1.9702,
338
+ "step": 23500
339
+ },
340
+ {
341
+ "epoch": 1.3195513525401363,
342
+ "grad_norm": 2.928476095199585,
343
+ "learning_rate": 1.1310668836934005e-05,
344
+ "loss": 1.9813,
345
+ "step": 24000
346
+ },
347
+ {
348
+ "epoch": 1.3470420057180559,
349
+ "grad_norm": 3.026412010192871,
350
+ "learning_rate": 1.1125702870671797e-05,
351
+ "loss": 1.9192,
352
+ "step": 24500
353
+ },
354
+ {
355
+ "epoch": 1.3745326588959754,
356
+ "grad_norm": 3.093477249145508,
357
+ "learning_rate": 1.094073690440959e-05,
358
+ "loss": 1.9387,
359
+ "step": 25000
360
+ },
361
+ {
362
+ "epoch": 1.402023312073895,
363
+ "grad_norm": 2.920074701309204,
364
+ "learning_rate": 1.0756140870079905e-05,
365
+ "loss": 1.9651,
366
+ "step": 25500
367
+ },
368
+ {
369
+ "epoch": 1.4295139652518145,
370
+ "grad_norm": 2.5337624549865723,
371
+ "learning_rate": 1.0571174903817697e-05,
372
+ "loss": 1.9262,
373
+ "step": 26000
374
+ },
375
+ {
376
+ "epoch": 1.4570046184297338,
377
+ "grad_norm": 2.506801128387451,
378
+ "learning_rate": 1.0386208937555491e-05,
379
+ "loss": 1.9542,
380
+ "step": 26500
381
+ },
382
+ {
383
+ "epoch": 1.4844952716076534,
384
+ "grad_norm": 2.7995145320892334,
385
+ "learning_rate": 1.0201242971293283e-05,
386
+ "loss": 1.9609,
387
+ "step": 27000
388
+ },
389
+ {
390
+ "epoch": 1.511985924785573,
391
+ "grad_norm": 4.121292591094971,
392
+ "learning_rate": 1.0016646936963598e-05,
393
+ "loss": 1.9317,
394
+ "step": 27500
395
+ },
396
+ {
397
+ "epoch": 1.5394765779634925,
398
+ "grad_norm": 2.802678346633911,
399
+ "learning_rate": 9.831680970701392e-06,
400
+ "loss": 1.9519,
401
+ "step": 28000
402
+ },
403
+ {
404
+ "epoch": 1.5669672311414118,
405
+ "grad_norm": 2.965937852859497,
406
+ "learning_rate": 9.646715004439184e-06,
407
+ "loss": 1.9401,
408
+ "step": 28500
409
+ },
410
+ {
411
+ "epoch": 1.5944578843193313,
412
+ "grad_norm": 2.364539861679077,
413
+ "learning_rate": 9.461749038176976e-06,
414
+ "loss": 1.9686,
415
+ "step": 29000
416
+ },
417
+ {
418
+ "epoch": 1.6219485374972509,
419
+ "grad_norm": 2.8087072372436523,
420
+ "learning_rate": 9.276783071914768e-06,
421
+ "loss": 1.9046,
422
+ "step": 29500
423
+ },
424
+ {
425
+ "epoch": 1.6494391906751704,
426
+ "grad_norm": 3.0667874813079834,
427
+ "learning_rate": 9.092187037585085e-06,
428
+ "loss": 1.8992,
429
+ "step": 30000
430
+ },
431
+ {
432
+ "epoch": 1.67692984385309,
433
+ "grad_norm": 2.4543378353118896,
434
+ "learning_rate": 8.907221071322877e-06,
435
+ "loss": 1.8936,
436
+ "step": 30500
437
+ },
438
+ {
439
+ "epoch": 1.7044204970310095,
440
+ "grad_norm": 2.695256233215332,
441
+ "learning_rate": 8.72225510506067e-06,
442
+ "loss": 1.9429,
443
+ "step": 31000
444
+ },
445
+ {
446
+ "epoch": 1.731911150208929,
447
+ "grad_norm": 2.377511978149414,
448
+ "learning_rate": 8.537289138798461e-06,
449
+ "loss": 1.9787,
450
+ "step": 31500
451
+ },
452
+ {
453
+ "epoch": 1.7594018033868486,
454
+ "grad_norm": 2.74446964263916,
455
+ "learning_rate": 8.35269310446878e-06,
456
+ "loss": 1.9272,
457
+ "step": 32000
458
+ },
459
+ {
460
+ "epoch": 1.7868924565647681,
461
+ "grad_norm": 3.174065589904785,
462
+ "learning_rate": 8.16772713820657e-06,
463
+ "loss": 1.9744,
464
+ "step": 32500
465
+ },
466
+ {
467
+ "epoch": 1.8143831097426875,
468
+ "grad_norm": 3.6638684272766113,
469
+ "learning_rate": 7.982761171944364e-06,
470
+ "loss": 1.9441,
471
+ "step": 33000
472
+ },
473
+ {
474
+ "epoch": 1.841873762920607,
475
+ "grad_norm": 2.563633441925049,
476
+ "learning_rate": 7.797795205682154e-06,
477
+ "loss": 1.9422,
478
+ "step": 33500
479
+ },
480
+ {
481
+ "epoch": 1.8693644160985265,
482
+ "grad_norm": 2.817347764968872,
483
+ "learning_rate": 7.612829239419948e-06,
484
+ "loss": 1.9556,
485
+ "step": 34000
486
+ },
487
+ {
488
+ "epoch": 1.8968550692764459,
489
+ "grad_norm": 2.59664249420166,
490
+ "learning_rate": 7.428233205090264e-06,
491
+ "loss": 1.9754,
492
+ "step": 34500
493
+ },
494
+ {
495
+ "epoch": 1.9243457224543654,
496
+ "grad_norm": 2.615318775177002,
497
+ "learning_rate": 7.2432672388280566e-06,
498
+ "loss": 1.9349,
499
+ "step": 35000
500
+ },
501
+ {
502
+ "epoch": 1.951836375632285,
503
+ "grad_norm": 2.5411360263824463,
504
+ "learning_rate": 7.058301272565849e-06,
505
+ "loss": 1.9612,
506
+ "step": 35500
507
+ },
508
+ {
509
+ "epoch": 1.9793270288102045,
510
+ "grad_norm": 2.809213161468506,
511
+ "learning_rate": 6.873335306303641e-06,
512
+ "loss": 1.9433,
513
+ "step": 36000
514
+ },
515
+ {
516
+ "epoch": 2.006817681988124,
517
+ "grad_norm": 2.415309190750122,
518
+ "learning_rate": 6.688739271973957e-06,
519
+ "loss": 1.9222,
520
+ "step": 36500
521
+ },
522
+ {
523
+ "epoch": 2.0343083351660436,
524
+ "grad_norm": 2.3479654788970947,
525
+ "learning_rate": 6.5037733057117494e-06,
526
+ "loss": 1.8997,
527
+ "step": 37000
528
+ },
529
+ {
530
+ "epoch": 2.061798988343963,
531
+ "grad_norm": 2.566889762878418,
532
+ "learning_rate": 6.3188073394495416e-06,
533
+ "loss": 1.8604,
534
+ "step": 37500
535
+ },
536
+ {
537
+ "epoch": 2.0892896415218827,
538
+ "grad_norm": 2.9693121910095215,
539
+ "learning_rate": 6.133841373187334e-06,
540
+ "loss": 1.9029,
541
+ "step": 38000
542
+ }
543
+ ],
544
+ "logging_steps": 500,
545
+ "max_steps": 54564,
546
+ "num_input_tokens_seen": 0,
547
+ "num_train_epochs": 3,
548
+ "save_steps": 1818,
549
+ "stateful_callbacks": {
550
+ "TrainerControl": {
551
+ "args": {
552
+ "should_epoch_stop": false,
553
+ "should_evaluate": false,
554
+ "should_log": false,
555
+ "should_save": true,
556
+ "should_training_stop": false
557
+ },
558
+ "attributes": {}
559
+ }
560
+ },
561
+ "total_flos": 9975345905664000.0,
562
+ "train_batch_size": 4,
563
+ "trial_name": null,
564
+ "trial_params": null
565
+ }
base/10e3/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
base/110e3/config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "gpt2",
3
+ "activation_function": "gelu_new",
4
+ "architectures": [
5
+ "GPT2LMHeadModel"
6
+ ],
7
+ "attn_pdrop": 0.1,
8
+ "bos_token_id": 50256,
9
+ "embd_pdrop": 0.1,
10
+ "eos_token_id": 50256,
11
+ "initializer_range": 0.02,
12
+ "layer_norm_epsilon": 1e-05,
13
+ "model_type": "gpt2",
14
+ "n_ctx": 1024,
15
+ "n_embd": 768,
16
+ "n_head": 12,
17
+ "n_inner": null,
18
+ "n_layer": 12,
19
+ "n_positions": 1024,
20
+ "reorder_and_upcast_attn": false,
21
+ "resid_pdrop": 0.1,
22
+ "scale_attn_by_inverse_layer_idx": false,
23
+ "scale_attn_weights": true,
24
+ "summary_activation": null,
25
+ "summary_first_dropout": 0.1,
26
+ "summary_proj_to_labels": true,
27
+ "summary_type": "cls_index",
28
+ "summary_use_proj": true,
29
+ "task_specific_params": {
30
+ "text-generation": {
31
+ "do_sample": true,
32
+ "max_length": 50
33
+ }
34
+ },
35
+ "torch_dtype": "float32",
36
+ "transformers_version": "4.46.3",
37
+ "use_cache": true,
38
+ "vocab_size": 50257
39
+ }
base/110e3/generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 50256,
4
+ "eos_token_id": 50256,
5
+ "transformers_version": "4.46.3"
6
+ }