craa commited on
Commit
4177a38
·
verified ·
1 Parent(s): e819276

Training in progress, step 10000, checkpoint

Browse files
checkpoint-10000/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPT2LMHeadModel"
5
+ ],
6
+ "attn_pdrop": 0.1,
7
+ "bos_token_id": 50256,
8
+ "embd_pdrop": 0.1,
9
+ "eos_token_id": 50256,
10
+ "initializer_range": 0.02,
11
+ "layer_norm_epsilon": 1e-05,
12
+ "model_type": "gpt2",
13
+ "n_embd": 768,
14
+ "n_head": 12,
15
+ "n_inner": null,
16
+ "n_layer": 12,
17
+ "n_positions": 1024,
18
+ "reorder_and_upcast_attn": false,
19
+ "resid_pdrop": 0.1,
20
+ "scale_attn_by_inverse_layer_idx": false,
21
+ "scale_attn_weights": true,
22
+ "summary_activation": null,
23
+ "summary_first_dropout": 0.1,
24
+ "summary_proj_to_labels": true,
25
+ "summary_type": "cls_index",
26
+ "summary_use_proj": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.47.0.dev0",
29
+ "use_cache": true,
30
+ "vocab_size": 52000
31
+ }
checkpoint-10000/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": "4.47.0.dev0"
6
+ }
checkpoint-10000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:31a122e54d54e7545738387569fe7f2c22ef807cefa0a90068ce2902ac9c31da
3
+ size 503128704
checkpoint-10000/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8adf5c376e8b1f678ecc2613ea530135b298357761a4285b75f8ec172fabb13a
3
+ size 1006351290
checkpoint-10000/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:53a47b99e43a46630effbfb45e7223c8a493d5b48b00aebac2470d03eafb7364
3
+ size 14244
checkpoint-10000/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baaddc0dc177f093cb7927eae851400a36fde6f45129bcaf2df443bdde5269f4
3
+ size 1064
checkpoint-10000/special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {}
checkpoint-10000/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-10000/tokenizer_config.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-10000/trainer_state.json ADDED
@@ -0,0 +1,1523 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 3.7489497661590576,
3
+ "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M__1208/checkpoint-10000",
4
+ "epoch": 1.0781671159029649,
5
+ "eval_steps": 1000,
6
+ "global_step": 10000,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.005390835579514825,
13
+ "grad_norm": 1.3872060775756836,
14
+ "learning_rate": 0.0003,
15
+ "loss": 8.7938,
16
+ "step": 50
17
+ },
18
+ {
19
+ "epoch": 0.01078167115902965,
20
+ "grad_norm": 3.891347885131836,
21
+ "learning_rate": 0.0006,
22
+ "loss": 6.992,
23
+ "step": 100
24
+ },
25
+ {
26
+ "epoch": 0.016172506738544475,
27
+ "grad_norm": 1.848716378211975,
28
+ "learning_rate": 0.0005996762007555315,
29
+ "loss": 6.5045,
30
+ "step": 150
31
+ },
32
+ {
33
+ "epoch": 0.0215633423180593,
34
+ "grad_norm": 1.3053127527236938,
35
+ "learning_rate": 0.000599352401511063,
36
+ "loss": 6.2433,
37
+ "step": 200
38
+ },
39
+ {
40
+ "epoch": 0.026954177897574125,
41
+ "grad_norm": 1.2040534019470215,
42
+ "learning_rate": 0.0005990286022665946,
43
+ "loss": 6.0719,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.03234501347708895,
48
+ "grad_norm": 1.105197548866272,
49
+ "learning_rate": 0.0005987048030221263,
50
+ "loss": 5.9606,
51
+ "step": 300
52
+ },
53
+ {
54
+ "epoch": 0.03773584905660377,
55
+ "grad_norm": 1.502025842666626,
56
+ "learning_rate": 0.0005983810037776578,
57
+ "loss": 5.8818,
58
+ "step": 350
59
+ },
60
+ {
61
+ "epoch": 0.0431266846361186,
62
+ "grad_norm": 2.493607759475708,
63
+ "learning_rate": 0.0005980572045331894,
64
+ "loss": 5.8101,
65
+ "step": 400
66
+ },
67
+ {
68
+ "epoch": 0.04851752021563342,
69
+ "grad_norm": 1.3217182159423828,
70
+ "learning_rate": 0.0005977334052887209,
71
+ "loss": 5.7175,
72
+ "step": 450
73
+ },
74
+ {
75
+ "epoch": 0.05390835579514825,
76
+ "grad_norm": 1.215240240097046,
77
+ "learning_rate": 0.0005974096060442526,
78
+ "loss": 5.6639,
79
+ "step": 500
80
+ },
81
+ {
82
+ "epoch": 0.05929919137466307,
83
+ "grad_norm": 1.3821635246276855,
84
+ "learning_rate": 0.0005970858067997841,
85
+ "loss": 5.5744,
86
+ "step": 550
87
+ },
88
+ {
89
+ "epoch": 0.0646900269541779,
90
+ "grad_norm": 1.378275752067566,
91
+ "learning_rate": 0.0005967620075553157,
92
+ "loss": 5.5287,
93
+ "step": 600
94
+ },
95
+ {
96
+ "epoch": 0.07008086253369272,
97
+ "grad_norm": 1.1444066762924194,
98
+ "learning_rate": 0.0005964382083108472,
99
+ "loss": 5.4332,
100
+ "step": 650
101
+ },
102
+ {
103
+ "epoch": 0.07547169811320754,
104
+ "grad_norm": 1.657812476158142,
105
+ "learning_rate": 0.0005961144090663788,
106
+ "loss": 5.4098,
107
+ "step": 700
108
+ },
109
+ {
110
+ "epoch": 0.08086253369272237,
111
+ "grad_norm": 1.3864067792892456,
112
+ "learning_rate": 0.0005957906098219104,
113
+ "loss": 5.3168,
114
+ "step": 750
115
+ },
116
+ {
117
+ "epoch": 0.0862533692722372,
118
+ "grad_norm": 1.0762931108474731,
119
+ "learning_rate": 0.0005954668105774419,
120
+ "loss": 5.2678,
121
+ "step": 800
122
+ },
123
+ {
124
+ "epoch": 0.09164420485175202,
125
+ "grad_norm": 1.2499072551727295,
126
+ "learning_rate": 0.0005951430113329735,
127
+ "loss": 5.2107,
128
+ "step": 850
129
+ },
130
+ {
131
+ "epoch": 0.09703504043126684,
132
+ "grad_norm": 0.9282752275466919,
133
+ "learning_rate": 0.0005948192120885051,
134
+ "loss": 5.1849,
135
+ "step": 900
136
+ },
137
+ {
138
+ "epoch": 0.10242587601078167,
139
+ "grad_norm": 1.071282982826233,
140
+ "learning_rate": 0.0005944954128440366,
141
+ "loss": 5.1275,
142
+ "step": 950
143
+ },
144
+ {
145
+ "epoch": 0.1078167115902965,
146
+ "grad_norm": 1.3316810131072998,
147
+ "learning_rate": 0.0005941716135995682,
148
+ "loss": 5.1151,
149
+ "step": 1000
150
+ },
151
+ {
152
+ "epoch": 0.1078167115902965,
153
+ "eval_accuracy": 0.22672739349001877,
154
+ "eval_loss": 5.027144432067871,
155
+ "eval_runtime": 184.654,
156
+ "eval_samples_per_second": 97.539,
157
+ "eval_steps_per_second": 6.098,
158
+ "step": 1000
159
+ },
160
+ {
161
+ "epoch": 0.11320754716981132,
162
+ "grad_norm": 1.1736186742782593,
163
+ "learning_rate": 0.0005938478143550997,
164
+ "loss": 5.0742,
165
+ "step": 1050
166
+ },
167
+ {
168
+ "epoch": 0.11859838274932614,
169
+ "grad_norm": 1.217553973197937,
170
+ "learning_rate": 0.0005935240151106314,
171
+ "loss": 5.0002,
172
+ "step": 1100
173
+ },
174
+ {
175
+ "epoch": 0.12398921832884097,
176
+ "grad_norm": 1.3116732835769653,
177
+ "learning_rate": 0.0005932002158661629,
178
+ "loss": 4.9959,
179
+ "step": 1150
180
+ },
181
+ {
182
+ "epoch": 0.1293800539083558,
183
+ "grad_norm": 1.10372793674469,
184
+ "learning_rate": 0.0005928764166216945,
185
+ "loss": 4.9613,
186
+ "step": 1200
187
+ },
188
+ {
189
+ "epoch": 0.1347708894878706,
190
+ "grad_norm": 1.0251359939575195,
191
+ "learning_rate": 0.000592552617377226,
192
+ "loss": 4.9081,
193
+ "step": 1250
194
+ },
195
+ {
196
+ "epoch": 0.14016172506738545,
197
+ "grad_norm": 1.1212270259857178,
198
+ "learning_rate": 0.0005922288181327577,
199
+ "loss": 4.9098,
200
+ "step": 1300
201
+ },
202
+ {
203
+ "epoch": 0.14555256064690028,
204
+ "grad_norm": 1.0358482599258423,
205
+ "learning_rate": 0.0005919050188882893,
206
+ "loss": 4.8835,
207
+ "step": 1350
208
+ },
209
+ {
210
+ "epoch": 0.1509433962264151,
211
+ "grad_norm": 1.127200722694397,
212
+ "learning_rate": 0.0005915812196438207,
213
+ "loss": 4.85,
214
+ "step": 1400
215
+ },
216
+ {
217
+ "epoch": 0.15633423180592992,
218
+ "grad_norm": 0.9541453123092651,
219
+ "learning_rate": 0.0005912574203993524,
220
+ "loss": 4.8199,
221
+ "step": 1450
222
+ },
223
+ {
224
+ "epoch": 0.16172506738544473,
225
+ "grad_norm": 0.9905341863632202,
226
+ "learning_rate": 0.0005909336211548839,
227
+ "loss": 4.812,
228
+ "step": 1500
229
+ },
230
+ {
231
+ "epoch": 0.16711590296495957,
232
+ "grad_norm": 0.9883430600166321,
233
+ "learning_rate": 0.0005906098219104155,
234
+ "loss": 4.7776,
235
+ "step": 1550
236
+ },
237
+ {
238
+ "epoch": 0.1725067385444744,
239
+ "grad_norm": 1.14104425907135,
240
+ "learning_rate": 0.000590286022665947,
241
+ "loss": 4.7877,
242
+ "step": 1600
243
+ },
244
+ {
245
+ "epoch": 0.1778975741239892,
246
+ "grad_norm": 0.915554404258728,
247
+ "learning_rate": 0.0005899622234214787,
248
+ "loss": 4.7362,
249
+ "step": 1650
250
+ },
251
+ {
252
+ "epoch": 0.18328840970350405,
253
+ "grad_norm": 0.8896968364715576,
254
+ "learning_rate": 0.0005896384241770102,
255
+ "loss": 4.6944,
256
+ "step": 1700
257
+ },
258
+ {
259
+ "epoch": 0.18867924528301888,
260
+ "grad_norm": 0.7295605540275574,
261
+ "learning_rate": 0.0005893146249325418,
262
+ "loss": 4.679,
263
+ "step": 1750
264
+ },
265
+ {
266
+ "epoch": 0.1940700808625337,
267
+ "grad_norm": 0.8035542368888855,
268
+ "learning_rate": 0.0005889908256880733,
269
+ "loss": 4.669,
270
+ "step": 1800
271
+ },
272
+ {
273
+ "epoch": 0.19946091644204852,
274
+ "grad_norm": 0.9095346927642822,
275
+ "learning_rate": 0.0005886670264436049,
276
+ "loss": 4.6768,
277
+ "step": 1850
278
+ },
279
+ {
280
+ "epoch": 0.20485175202156333,
281
+ "grad_norm": 0.8672581315040588,
282
+ "learning_rate": 0.0005883432271991365,
283
+ "loss": 4.6219,
284
+ "step": 1900
285
+ },
286
+ {
287
+ "epoch": 0.21024258760107817,
288
+ "grad_norm": 0.7349112033843994,
289
+ "learning_rate": 0.0005880194279546681,
290
+ "loss": 4.6077,
291
+ "step": 1950
292
+ },
293
+ {
294
+ "epoch": 0.215633423180593,
295
+ "grad_norm": 0.9716205596923828,
296
+ "learning_rate": 0.0005876956287101996,
297
+ "loss": 4.5792,
298
+ "step": 2000
299
+ },
300
+ {
301
+ "epoch": 0.215633423180593,
302
+ "eval_accuracy": 0.27099464439050674,
303
+ "eval_loss": 4.504695415496826,
304
+ "eval_runtime": 184.1142,
305
+ "eval_samples_per_second": 97.825,
306
+ "eval_steps_per_second": 6.116,
307
+ "step": 2000
308
+ },
309
+ {
310
+ "epoch": 0.2210242587601078,
311
+ "grad_norm": 1.203096628189087,
312
+ "learning_rate": 0.0005873718294657312,
313
+ "loss": 4.5695,
314
+ "step": 2050
315
+ },
316
+ {
317
+ "epoch": 0.22641509433962265,
318
+ "grad_norm": 0.9952515959739685,
319
+ "learning_rate": 0.0005870480302212628,
320
+ "loss": 4.5412,
321
+ "step": 2100
322
+ },
323
+ {
324
+ "epoch": 0.23180592991913745,
325
+ "grad_norm": 0.8814882040023804,
326
+ "learning_rate": 0.0005867242309767943,
327
+ "loss": 4.5392,
328
+ "step": 2150
329
+ },
330
+ {
331
+ "epoch": 0.2371967654986523,
332
+ "grad_norm": 1.0191090106964111,
333
+ "learning_rate": 0.0005864004317323259,
334
+ "loss": 4.5151,
335
+ "step": 2200
336
+ },
337
+ {
338
+ "epoch": 0.24258760107816713,
339
+ "grad_norm": 0.8580870628356934,
340
+ "learning_rate": 0.0005860766324878575,
341
+ "loss": 4.5014,
342
+ "step": 2250
343
+ },
344
+ {
345
+ "epoch": 0.24797843665768193,
346
+ "grad_norm": 0.8280946016311646,
347
+ "learning_rate": 0.000585752833243389,
348
+ "loss": 4.4755,
349
+ "step": 2300
350
+ },
351
+ {
352
+ "epoch": 0.25336927223719674,
353
+ "grad_norm": 0.9512032866477966,
354
+ "learning_rate": 0.0005854290339989206,
355
+ "loss": 4.4624,
356
+ "step": 2350
357
+ },
358
+ {
359
+ "epoch": 0.2587601078167116,
360
+ "grad_norm": 1.000420093536377,
361
+ "learning_rate": 0.0005851052347544521,
362
+ "loss": 4.4523,
363
+ "step": 2400
364
+ },
365
+ {
366
+ "epoch": 0.2641509433962264,
367
+ "grad_norm": 0.9130716919898987,
368
+ "learning_rate": 0.0005847814355099838,
369
+ "loss": 4.4414,
370
+ "step": 2450
371
+ },
372
+ {
373
+ "epoch": 0.2695417789757412,
374
+ "grad_norm": 0.8131120800971985,
375
+ "learning_rate": 0.0005844576362655154,
376
+ "loss": 4.4556,
377
+ "step": 2500
378
+ },
379
+ {
380
+ "epoch": 0.2749326145552561,
381
+ "grad_norm": 0.9587961435317993,
382
+ "learning_rate": 0.0005841338370210469,
383
+ "loss": 4.4083,
384
+ "step": 2550
385
+ },
386
+ {
387
+ "epoch": 0.2803234501347709,
388
+ "grad_norm": 0.846928060054779,
389
+ "learning_rate": 0.0005838100377765785,
390
+ "loss": 4.3825,
391
+ "step": 2600
392
+ },
393
+ {
394
+ "epoch": 0.2857142857142857,
395
+ "grad_norm": 0.9020141363143921,
396
+ "learning_rate": 0.0005834862385321101,
397
+ "loss": 4.3869,
398
+ "step": 2650
399
+ },
400
+ {
401
+ "epoch": 0.29110512129380056,
402
+ "grad_norm": 1.0460323095321655,
403
+ "learning_rate": 0.0005831624392876417,
404
+ "loss": 4.381,
405
+ "step": 2700
406
+ },
407
+ {
408
+ "epoch": 0.29649595687331537,
409
+ "grad_norm": 0.8714868426322937,
410
+ "learning_rate": 0.0005828386400431731,
411
+ "loss": 4.3618,
412
+ "step": 2750
413
+ },
414
+ {
415
+ "epoch": 0.3018867924528302,
416
+ "grad_norm": 0.7531731724739075,
417
+ "learning_rate": 0.0005825148407987048,
418
+ "loss": 4.361,
419
+ "step": 2800
420
+ },
421
+ {
422
+ "epoch": 0.30727762803234504,
423
+ "grad_norm": 0.7624960541725159,
424
+ "learning_rate": 0.0005821910415542363,
425
+ "loss": 4.3642,
426
+ "step": 2850
427
+ },
428
+ {
429
+ "epoch": 0.31266846361185985,
430
+ "grad_norm": 0.8426908850669861,
431
+ "learning_rate": 0.0005818672423097679,
432
+ "loss": 4.3367,
433
+ "step": 2900
434
+ },
435
+ {
436
+ "epoch": 0.31805929919137466,
437
+ "grad_norm": 0.9197669625282288,
438
+ "learning_rate": 0.0005815434430652994,
439
+ "loss": 4.3496,
440
+ "step": 2950
441
+ },
442
+ {
443
+ "epoch": 0.32345013477088946,
444
+ "grad_norm": 0.8955463171005249,
445
+ "learning_rate": 0.0005812196438208311,
446
+ "loss": 4.2988,
447
+ "step": 3000
448
+ },
449
+ {
450
+ "epoch": 0.32345013477088946,
451
+ "eval_accuracy": 0.2984499253065723,
452
+ "eval_loss": 4.238968372344971,
453
+ "eval_runtime": 185.178,
454
+ "eval_samples_per_second": 97.263,
455
+ "eval_steps_per_second": 6.081,
456
+ "step": 3000
457
+ },
458
+ {
459
+ "epoch": 0.3288409703504043,
460
+ "grad_norm": 0.7810043096542358,
461
+ "learning_rate": 0.0005808958445763626,
462
+ "loss": 4.3082,
463
+ "step": 3050
464
+ },
465
+ {
466
+ "epoch": 0.33423180592991913,
467
+ "grad_norm": 0.7354643940925598,
468
+ "learning_rate": 0.0005805720453318942,
469
+ "loss": 4.2858,
470
+ "step": 3100
471
+ },
472
+ {
473
+ "epoch": 0.33962264150943394,
474
+ "grad_norm": 0.7374374270439148,
475
+ "learning_rate": 0.0005802482460874257,
476
+ "loss": 4.2772,
477
+ "step": 3150
478
+ },
479
+ {
480
+ "epoch": 0.3450134770889488,
481
+ "grad_norm": 0.7399438619613647,
482
+ "learning_rate": 0.0005799244468429573,
483
+ "loss": 4.2993,
484
+ "step": 3200
485
+ },
486
+ {
487
+ "epoch": 0.3504043126684636,
488
+ "grad_norm": 0.7688404321670532,
489
+ "learning_rate": 0.0005796006475984889,
490
+ "loss": 4.2841,
491
+ "step": 3250
492
+ },
493
+ {
494
+ "epoch": 0.3557951482479784,
495
+ "grad_norm": 0.7221906781196594,
496
+ "learning_rate": 0.0005792768483540205,
497
+ "loss": 4.2667,
498
+ "step": 3300
499
+ },
500
+ {
501
+ "epoch": 0.3611859838274933,
502
+ "grad_norm": 0.7419180274009705,
503
+ "learning_rate": 0.000578953049109552,
504
+ "loss": 4.2709,
505
+ "step": 3350
506
+ },
507
+ {
508
+ "epoch": 0.3665768194070081,
509
+ "grad_norm": 0.7906516790390015,
510
+ "learning_rate": 0.0005786292498650836,
511
+ "loss": 4.2394,
512
+ "step": 3400
513
+ },
514
+ {
515
+ "epoch": 0.3719676549865229,
516
+ "grad_norm": 0.7784256339073181,
517
+ "learning_rate": 0.0005783054506206152,
518
+ "loss": 4.2519,
519
+ "step": 3450
520
+ },
521
+ {
522
+ "epoch": 0.37735849056603776,
523
+ "grad_norm": 0.7829201221466064,
524
+ "learning_rate": 0.0005779816513761467,
525
+ "loss": 4.2314,
526
+ "step": 3500
527
+ },
528
+ {
529
+ "epoch": 0.38274932614555257,
530
+ "grad_norm": 0.8310431241989136,
531
+ "learning_rate": 0.0005776578521316782,
532
+ "loss": 4.2097,
533
+ "step": 3550
534
+ },
535
+ {
536
+ "epoch": 0.3881401617250674,
537
+ "grad_norm": 0.8430302739143372,
538
+ "learning_rate": 0.0005773340528872099,
539
+ "loss": 4.2306,
540
+ "step": 3600
541
+ },
542
+ {
543
+ "epoch": 0.3935309973045822,
544
+ "grad_norm": 0.733945906162262,
545
+ "learning_rate": 0.0005770102536427414,
546
+ "loss": 4.2221,
547
+ "step": 3650
548
+ },
549
+ {
550
+ "epoch": 0.39892183288409705,
551
+ "grad_norm": 0.6383342742919922,
552
+ "learning_rate": 0.000576686454398273,
553
+ "loss": 4.2028,
554
+ "step": 3700
555
+ },
556
+ {
557
+ "epoch": 0.40431266846361186,
558
+ "grad_norm": 0.6672252416610718,
559
+ "learning_rate": 0.0005763626551538045,
560
+ "loss": 4.1801,
561
+ "step": 3750
562
+ },
563
+ {
564
+ "epoch": 0.40970350404312667,
565
+ "grad_norm": 0.7545523643493652,
566
+ "learning_rate": 0.0005760388559093362,
567
+ "loss": 4.1881,
568
+ "step": 3800
569
+ },
570
+ {
571
+ "epoch": 0.41509433962264153,
572
+ "grad_norm": 0.7340240478515625,
573
+ "learning_rate": 0.0005757150566648678,
574
+ "loss": 4.171,
575
+ "step": 3850
576
+ },
577
+ {
578
+ "epoch": 0.42048517520215634,
579
+ "grad_norm": 0.7903727889060974,
580
+ "learning_rate": 0.0005753912574203993,
581
+ "loss": 4.1784,
582
+ "step": 3900
583
+ },
584
+ {
585
+ "epoch": 0.42587601078167114,
586
+ "grad_norm": 0.8018081188201904,
587
+ "learning_rate": 0.0005750674581759309,
588
+ "loss": 4.1838,
589
+ "step": 3950
590
+ },
591
+ {
592
+ "epoch": 0.431266846361186,
593
+ "grad_norm": 0.722425103187561,
594
+ "learning_rate": 0.0005747436589314624,
595
+ "loss": 4.1617,
596
+ "step": 4000
597
+ },
598
+ {
599
+ "epoch": 0.431266846361186,
600
+ "eval_accuracy": 0.31149631215800827,
601
+ "eval_loss": 4.093442440032959,
602
+ "eval_runtime": 185.3481,
603
+ "eval_samples_per_second": 97.174,
604
+ "eval_steps_per_second": 6.075,
605
+ "step": 4000
606
+ },
607
+ {
608
+ "epoch": 0.4366576819407008,
609
+ "grad_norm": 0.9414203763008118,
610
+ "learning_rate": 0.0005744198596869941,
611
+ "loss": 4.15,
612
+ "step": 4050
613
+ },
614
+ {
615
+ "epoch": 0.4420485175202156,
616
+ "grad_norm": 0.7414451241493225,
617
+ "learning_rate": 0.0005740960604425255,
618
+ "loss": 4.1367,
619
+ "step": 4100
620
+ },
621
+ {
622
+ "epoch": 0.4474393530997305,
623
+ "grad_norm": 0.693946361541748,
624
+ "learning_rate": 0.0005737722611980572,
625
+ "loss": 4.1503,
626
+ "step": 4150
627
+ },
628
+ {
629
+ "epoch": 0.4528301886792453,
630
+ "grad_norm": 0.8669963479042053,
631
+ "learning_rate": 0.0005734484619535887,
632
+ "loss": 4.1342,
633
+ "step": 4200
634
+ },
635
+ {
636
+ "epoch": 0.4582210242587601,
637
+ "grad_norm": 0.717522144317627,
638
+ "learning_rate": 0.0005731246627091203,
639
+ "loss": 4.1361,
640
+ "step": 4250
641
+ },
642
+ {
643
+ "epoch": 0.4636118598382749,
644
+ "grad_norm": 0.5836500525474548,
645
+ "learning_rate": 0.0005728008634646518,
646
+ "loss": 4.1184,
647
+ "step": 4300
648
+ },
649
+ {
650
+ "epoch": 0.46900269541778977,
651
+ "grad_norm": 0.6450395584106445,
652
+ "learning_rate": 0.0005724770642201835,
653
+ "loss": 4.1149,
654
+ "step": 4350
655
+ },
656
+ {
657
+ "epoch": 0.4743935309973046,
658
+ "grad_norm": 0.6113871335983276,
659
+ "learning_rate": 0.000572153264975715,
660
+ "loss": 4.1102,
661
+ "step": 4400
662
+ },
663
+ {
664
+ "epoch": 0.4797843665768194,
665
+ "grad_norm": 0.7142659425735474,
666
+ "learning_rate": 0.0005718294657312466,
667
+ "loss": 4.1382,
668
+ "step": 4450
669
+ },
670
+ {
671
+ "epoch": 0.48517520215633425,
672
+ "grad_norm": 0.6929596066474915,
673
+ "learning_rate": 0.0005715056664867781,
674
+ "loss": 4.0957,
675
+ "step": 4500
676
+ },
677
+ {
678
+ "epoch": 0.49056603773584906,
679
+ "grad_norm": 0.5346873998641968,
680
+ "learning_rate": 0.0005711818672423097,
681
+ "loss": 4.1046,
682
+ "step": 4550
683
+ },
684
+ {
685
+ "epoch": 0.49595687331536387,
686
+ "grad_norm": 0.702147901058197,
687
+ "learning_rate": 0.0005708580679978413,
688
+ "loss": 4.1,
689
+ "step": 4600
690
+ },
691
+ {
692
+ "epoch": 0.5013477088948787,
693
+ "grad_norm": 0.5867359042167664,
694
+ "learning_rate": 0.0005705342687533729,
695
+ "loss": 4.0953,
696
+ "step": 4650
697
+ },
698
+ {
699
+ "epoch": 0.5067385444743935,
700
+ "grad_norm": 0.6331294178962708,
701
+ "learning_rate": 0.0005702104695089044,
702
+ "loss": 4.0979,
703
+ "step": 4700
704
+ },
705
+ {
706
+ "epoch": 0.5121293800539084,
707
+ "grad_norm": 0.6199395656585693,
708
+ "learning_rate": 0.000569886670264436,
709
+ "loss": 4.0987,
710
+ "step": 4750
711
+ },
712
+ {
713
+ "epoch": 0.5175202156334232,
714
+ "grad_norm": 0.6783130764961243,
715
+ "learning_rate": 0.0005695628710199675,
716
+ "loss": 4.0776,
717
+ "step": 4800
718
+ },
719
+ {
720
+ "epoch": 0.522911051212938,
721
+ "grad_norm": 0.6518664956092834,
722
+ "learning_rate": 0.0005692390717754991,
723
+ "loss": 4.1048,
724
+ "step": 4850
725
+ },
726
+ {
727
+ "epoch": 0.5283018867924528,
728
+ "grad_norm": 0.6022597551345825,
729
+ "learning_rate": 0.0005689152725310306,
730
+ "loss": 4.0778,
731
+ "step": 4900
732
+ },
733
+ {
734
+ "epoch": 0.5336927223719676,
735
+ "grad_norm": 0.6618348360061646,
736
+ "learning_rate": 0.0005685914732865623,
737
+ "loss": 4.0617,
738
+ "step": 4950
739
+ },
740
+ {
741
+ "epoch": 0.5390835579514824,
742
+ "grad_norm": 0.5986215472221375,
743
+ "learning_rate": 0.0005682676740420939,
744
+ "loss": 4.0655,
745
+ "step": 5000
746
+ },
747
+ {
748
+ "epoch": 0.5390835579514824,
749
+ "eval_accuracy": 0.32159864036122304,
750
+ "eval_loss": 3.9903128147125244,
751
+ "eval_runtime": 185.2705,
752
+ "eval_samples_per_second": 97.215,
753
+ "eval_steps_per_second": 6.078,
754
+ "step": 5000
755
+ },
756
+ {
757
+ "epoch": 0.5444743935309974,
758
+ "grad_norm": 0.5965196490287781,
759
+ "learning_rate": 0.0005679438747976254,
760
+ "loss": 4.0629,
761
+ "step": 5050
762
+ },
763
+ {
764
+ "epoch": 0.5498652291105122,
765
+ "grad_norm": 0.618198812007904,
766
+ "learning_rate": 0.000567620075553157,
767
+ "loss": 4.048,
768
+ "step": 5100
769
+ },
770
+ {
771
+ "epoch": 0.555256064690027,
772
+ "grad_norm": 0.6078091263771057,
773
+ "learning_rate": 0.0005672962763086886,
774
+ "loss": 4.0601,
775
+ "step": 5150
776
+ },
777
+ {
778
+ "epoch": 0.5606469002695418,
779
+ "grad_norm": 0.5713509917259216,
780
+ "learning_rate": 0.0005669724770642202,
781
+ "loss": 4.054,
782
+ "step": 5200
783
+ },
784
+ {
785
+ "epoch": 0.5660377358490566,
786
+ "grad_norm": 0.7706556916236877,
787
+ "learning_rate": 0.0005666486778197517,
788
+ "loss": 4.0499,
789
+ "step": 5250
790
+ },
791
+ {
792
+ "epoch": 0.5714285714285714,
793
+ "grad_norm": 0.6719933748245239,
794
+ "learning_rate": 0.0005663248785752833,
795
+ "loss": 4.0337,
796
+ "step": 5300
797
+ },
798
+ {
799
+ "epoch": 0.5768194070080862,
800
+ "grad_norm": 0.6685216426849365,
801
+ "learning_rate": 0.0005660010793308148,
802
+ "loss": 4.0071,
803
+ "step": 5350
804
+ },
805
+ {
806
+ "epoch": 0.5822102425876011,
807
+ "grad_norm": 0.6304830312728882,
808
+ "learning_rate": 0.0005656772800863465,
809
+ "loss": 4.0476,
810
+ "step": 5400
811
+ },
812
+ {
813
+ "epoch": 0.5876010781671159,
814
+ "grad_norm": 0.6301653385162354,
815
+ "learning_rate": 0.0005653534808418779,
816
+ "loss": 4.0282,
817
+ "step": 5450
818
+ },
819
+ {
820
+ "epoch": 0.5929919137466307,
821
+ "grad_norm": 0.6767484545707703,
822
+ "learning_rate": 0.0005650296815974096,
823
+ "loss": 4.0349,
824
+ "step": 5500
825
+ },
826
+ {
827
+ "epoch": 0.5983827493261455,
828
+ "grad_norm": 0.6710530519485474,
829
+ "learning_rate": 0.0005647058823529411,
830
+ "loss": 4.0179,
831
+ "step": 5550
832
+ },
833
+ {
834
+ "epoch": 0.6037735849056604,
835
+ "grad_norm": 0.5611121654510498,
836
+ "learning_rate": 0.0005643820831084727,
837
+ "loss": 4.0202,
838
+ "step": 5600
839
+ },
840
+ {
841
+ "epoch": 0.6091644204851752,
842
+ "grad_norm": 0.6187557578086853,
843
+ "learning_rate": 0.0005640582838640042,
844
+ "loss": 3.9935,
845
+ "step": 5650
846
+ },
847
+ {
848
+ "epoch": 0.6145552560646901,
849
+ "grad_norm": 0.6690448522567749,
850
+ "learning_rate": 0.0005637344846195358,
851
+ "loss": 4.0082,
852
+ "step": 5700
853
+ },
854
+ {
855
+ "epoch": 0.6199460916442049,
856
+ "grad_norm": 0.6771255135536194,
857
+ "learning_rate": 0.0005634106853750674,
858
+ "loss": 4.0026,
859
+ "step": 5750
860
+ },
861
+ {
862
+ "epoch": 0.6253369272237197,
863
+ "grad_norm": 0.5823601484298706,
864
+ "learning_rate": 0.000563086886130599,
865
+ "loss": 4.014,
866
+ "step": 5800
867
+ },
868
+ {
869
+ "epoch": 0.6307277628032345,
870
+ "grad_norm": 0.6157869100570679,
871
+ "learning_rate": 0.0005627630868861305,
872
+ "loss": 3.9952,
873
+ "step": 5850
874
+ },
875
+ {
876
+ "epoch": 0.6361185983827493,
877
+ "grad_norm": 0.6894795894622803,
878
+ "learning_rate": 0.0005624392876416621,
879
+ "loss": 3.9947,
880
+ "step": 5900
881
+ },
882
+ {
883
+ "epoch": 0.6415094339622641,
884
+ "grad_norm": 0.6236327290534973,
885
+ "learning_rate": 0.0005621154883971937,
886
+ "loss": 3.996,
887
+ "step": 5950
888
+ },
889
+ {
890
+ "epoch": 0.6469002695417789,
891
+ "grad_norm": 0.6344878077507019,
892
+ "learning_rate": 0.0005617916891527253,
893
+ "loss": 3.9934,
894
+ "step": 6000
895
+ },
896
+ {
897
+ "epoch": 0.6469002695417789,
898
+ "eval_accuracy": 0.3280596843351111,
899
+ "eval_loss": 3.924257278442383,
900
+ "eval_runtime": 185.4361,
901
+ "eval_samples_per_second": 97.128,
902
+ "eval_steps_per_second": 6.072,
903
+ "step": 6000
904
+ },
905
+ {
906
+ "epoch": 0.6522911051212938,
907
+ "grad_norm": 0.5802189111709595,
908
+ "learning_rate": 0.0005614743658931463,
909
+ "loss": 3.9855,
910
+ "step": 6050
911
+ },
912
+ {
913
+ "epoch": 0.6576819407008087,
914
+ "grad_norm": 0.6164999604225159,
915
+ "learning_rate": 0.0005611505666486777,
916
+ "loss": 4.0042,
917
+ "step": 6100
918
+ },
919
+ {
920
+ "epoch": 0.6630727762803235,
921
+ "grad_norm": 0.6946608424186707,
922
+ "learning_rate": 0.0005608267674042094,
923
+ "loss": 3.989,
924
+ "step": 6150
925
+ },
926
+ {
927
+ "epoch": 0.6684636118598383,
928
+ "grad_norm": 0.6158466935157776,
929
+ "learning_rate": 0.0005605029681597409,
930
+ "loss": 3.9871,
931
+ "step": 6200
932
+ },
933
+ {
934
+ "epoch": 0.6738544474393531,
935
+ "grad_norm": 0.6288596391677856,
936
+ "learning_rate": 0.0005601791689152725,
937
+ "loss": 3.9773,
938
+ "step": 6250
939
+ },
940
+ {
941
+ "epoch": 0.6792452830188679,
942
+ "grad_norm": 0.7042039036750793,
943
+ "learning_rate": 0.000559855369670804,
944
+ "loss": 3.9791,
945
+ "step": 6300
946
+ },
947
+ {
948
+ "epoch": 0.6846361185983828,
949
+ "grad_norm": 0.5582059025764465,
950
+ "learning_rate": 0.0005595315704263357,
951
+ "loss": 3.9639,
952
+ "step": 6350
953
+ },
954
+ {
955
+ "epoch": 0.6900269541778976,
956
+ "grad_norm": 0.6060189604759216,
957
+ "learning_rate": 0.0005592077711818672,
958
+ "loss": 3.9775,
959
+ "step": 6400
960
+ },
961
+ {
962
+ "epoch": 0.6954177897574124,
963
+ "grad_norm": 0.7476151585578918,
964
+ "learning_rate": 0.0005588839719373988,
965
+ "loss": 3.9536,
966
+ "step": 6450
967
+ },
968
+ {
969
+ "epoch": 0.7008086253369272,
970
+ "grad_norm": 0.6221930384635925,
971
+ "learning_rate": 0.0005585601726929303,
972
+ "loss": 3.9583,
973
+ "step": 6500
974
+ },
975
+ {
976
+ "epoch": 0.706199460916442,
977
+ "grad_norm": 0.6130439639091492,
978
+ "learning_rate": 0.0005582363734484619,
979
+ "loss": 3.9417,
980
+ "step": 6550
981
+ },
982
+ {
983
+ "epoch": 0.7115902964959568,
984
+ "grad_norm": 0.5701190233230591,
985
+ "learning_rate": 0.0005579125742039935,
986
+ "loss": 3.9488,
987
+ "step": 6600
988
+ },
989
+ {
990
+ "epoch": 0.7169811320754716,
991
+ "grad_norm": 0.5955173969268799,
992
+ "learning_rate": 0.0005575887749595251,
993
+ "loss": 3.9511,
994
+ "step": 6650
995
+ },
996
+ {
997
+ "epoch": 0.7223719676549866,
998
+ "grad_norm": 0.6578053832054138,
999
+ "learning_rate": 0.0005572649757150566,
1000
+ "loss": 3.9526,
1001
+ "step": 6700
1002
+ },
1003
+ {
1004
+ "epoch": 0.7277628032345014,
1005
+ "grad_norm": 0.7318617105484009,
1006
+ "learning_rate": 0.0005569411764705882,
1007
+ "loss": 3.9511,
1008
+ "step": 6750
1009
+ },
1010
+ {
1011
+ "epoch": 0.7331536388140162,
1012
+ "grad_norm": 0.6341660618782043,
1013
+ "learning_rate": 0.0005566173772261198,
1014
+ "loss": 3.918,
1015
+ "step": 6800
1016
+ },
1017
+ {
1018
+ "epoch": 0.738544474393531,
1019
+ "grad_norm": 0.5913158059120178,
1020
+ "learning_rate": 0.0005562935779816513,
1021
+ "loss": 3.9336,
1022
+ "step": 6850
1023
+ },
1024
+ {
1025
+ "epoch": 0.7439353099730458,
1026
+ "grad_norm": 0.6217501759529114,
1027
+ "learning_rate": 0.0005559697787371828,
1028
+ "loss": 3.9485,
1029
+ "step": 6900
1030
+ },
1031
+ {
1032
+ "epoch": 0.7493261455525606,
1033
+ "grad_norm": 0.629990816116333,
1034
+ "learning_rate": 0.0005556459794927145,
1035
+ "loss": 3.9304,
1036
+ "step": 6950
1037
+ },
1038
+ {
1039
+ "epoch": 0.7547169811320755,
1040
+ "grad_norm": 0.598209798336029,
1041
+ "learning_rate": 0.000555322180248246,
1042
+ "loss": 3.9432,
1043
+ "step": 7000
1044
+ },
1045
+ {
1046
+ "epoch": 0.7547169811320755,
1047
+ "eval_accuracy": 0.33339541034990466,
1048
+ "eval_loss": 3.867119312286377,
1049
+ "eval_runtime": 185.4718,
1050
+ "eval_samples_per_second": 97.109,
1051
+ "eval_steps_per_second": 6.071,
1052
+ "step": 7000
1053
+ },
1054
+ {
1055
+ "epoch": 0.7601078167115903,
1056
+ "grad_norm": 0.7586105465888977,
1057
+ "learning_rate": 0.0005549983810037776,
1058
+ "loss": 3.911,
1059
+ "step": 7050
1060
+ },
1061
+ {
1062
+ "epoch": 0.7654986522911051,
1063
+ "grad_norm": 0.6012836694717407,
1064
+ "learning_rate": 0.0005546745817593091,
1065
+ "loss": 3.9126,
1066
+ "step": 7100
1067
+ },
1068
+ {
1069
+ "epoch": 0.77088948787062,
1070
+ "grad_norm": 0.6788042187690735,
1071
+ "learning_rate": 0.0005543507825148408,
1072
+ "loss": 3.9149,
1073
+ "step": 7150
1074
+ },
1075
+ {
1076
+ "epoch": 0.7762803234501348,
1077
+ "grad_norm": 0.60019451379776,
1078
+ "learning_rate": 0.0005540269832703723,
1079
+ "loss": 3.8948,
1080
+ "step": 7200
1081
+ },
1082
+ {
1083
+ "epoch": 0.7816711590296496,
1084
+ "grad_norm": 0.5468128323554993,
1085
+ "learning_rate": 0.0005537031840259039,
1086
+ "loss": 3.9228,
1087
+ "step": 7250
1088
+ },
1089
+ {
1090
+ "epoch": 0.7870619946091644,
1091
+ "grad_norm": 0.560189425945282,
1092
+ "learning_rate": 0.0005533793847814354,
1093
+ "loss": 3.8974,
1094
+ "step": 7300
1095
+ },
1096
+ {
1097
+ "epoch": 0.7924528301886793,
1098
+ "grad_norm": 0.6238011717796326,
1099
+ "learning_rate": 0.000553055585536967,
1100
+ "loss": 3.9169,
1101
+ "step": 7350
1102
+ },
1103
+ {
1104
+ "epoch": 0.7978436657681941,
1105
+ "grad_norm": 0.6309143304824829,
1106
+ "learning_rate": 0.0005527317862924987,
1107
+ "loss": 3.9167,
1108
+ "step": 7400
1109
+ },
1110
+ {
1111
+ "epoch": 0.8032345013477089,
1112
+ "grad_norm": 0.5579202771186829,
1113
+ "learning_rate": 0.0005524079870480301,
1114
+ "loss": 3.9102,
1115
+ "step": 7450
1116
+ },
1117
+ {
1118
+ "epoch": 0.8086253369272237,
1119
+ "grad_norm": 0.637144923210144,
1120
+ "learning_rate": 0.0005520841878035618,
1121
+ "loss": 3.9305,
1122
+ "step": 7500
1123
+ },
1124
+ {
1125
+ "epoch": 0.8140161725067385,
1126
+ "grad_norm": 0.560736358165741,
1127
+ "learning_rate": 0.0005517603885590933,
1128
+ "loss": 3.9012,
1129
+ "step": 7550
1130
+ },
1131
+ {
1132
+ "epoch": 0.8194070080862533,
1133
+ "grad_norm": 0.5624358057975769,
1134
+ "learning_rate": 0.0005514365893146249,
1135
+ "loss": 3.8973,
1136
+ "step": 7600
1137
+ },
1138
+ {
1139
+ "epoch": 0.8247978436657682,
1140
+ "grad_norm": 0.5344393253326416,
1141
+ "learning_rate": 0.0005511127900701564,
1142
+ "loss": 3.8894,
1143
+ "step": 7650
1144
+ },
1145
+ {
1146
+ "epoch": 0.8301886792452831,
1147
+ "grad_norm": 0.5487794280052185,
1148
+ "learning_rate": 0.000550788990825688,
1149
+ "loss": 3.9173,
1150
+ "step": 7700
1151
+ },
1152
+ {
1153
+ "epoch": 0.8355795148247979,
1154
+ "grad_norm": 0.5416852831840515,
1155
+ "learning_rate": 0.0005504651915812196,
1156
+ "loss": 3.9131,
1157
+ "step": 7750
1158
+ },
1159
+ {
1160
+ "epoch": 0.8409703504043127,
1161
+ "grad_norm": 0.6219443678855896,
1162
+ "learning_rate": 0.0005501413923367512,
1163
+ "loss": 3.8783,
1164
+ "step": 7800
1165
+ },
1166
+ {
1167
+ "epoch": 0.8463611859838275,
1168
+ "grad_norm": 0.5484825968742371,
1169
+ "learning_rate": 0.0005498175930922827,
1170
+ "loss": 3.8889,
1171
+ "step": 7850
1172
+ },
1173
+ {
1174
+ "epoch": 0.8517520215633423,
1175
+ "grad_norm": 0.5506182909011841,
1176
+ "learning_rate": 0.0005494937938478143,
1177
+ "loss": 3.8945,
1178
+ "step": 7900
1179
+ },
1180
+ {
1181
+ "epoch": 0.8571428571428571,
1182
+ "grad_norm": 0.5658506155014038,
1183
+ "learning_rate": 0.0005491699946033459,
1184
+ "loss": 3.883,
1185
+ "step": 7950
1186
+ },
1187
+ {
1188
+ "epoch": 0.862533692722372,
1189
+ "grad_norm": 0.5695346593856812,
1190
+ "learning_rate": 0.0005488461953588775,
1191
+ "loss": 3.8869,
1192
+ "step": 8000
1193
+ },
1194
+ {
1195
+ "epoch": 0.862533692722372,
1196
+ "eval_accuracy": 0.33750498852571176,
1197
+ "eval_loss": 3.818067789077759,
1198
+ "eval_runtime": 185.1439,
1199
+ "eval_samples_per_second": 97.281,
1200
+ "eval_steps_per_second": 6.082,
1201
+ "step": 8000
1202
+ },
1203
+ {
1204
+ "epoch": 0.8679245283018868,
1205
+ "grad_norm": 0.5362111926078796,
1206
+ "learning_rate": 0.000548522396114409,
1207
+ "loss": 3.8853,
1208
+ "step": 8050
1209
+ },
1210
+ {
1211
+ "epoch": 0.8733153638814016,
1212
+ "grad_norm": 0.6255955696105957,
1213
+ "learning_rate": 0.0005482050728548299,
1214
+ "loss": 3.8955,
1215
+ "step": 8100
1216
+ },
1217
+ {
1218
+ "epoch": 0.8787061994609164,
1219
+ "grad_norm": 0.4989166557788849,
1220
+ "learning_rate": 0.0005478812736103615,
1221
+ "loss": 3.8747,
1222
+ "step": 8150
1223
+ },
1224
+ {
1225
+ "epoch": 0.8840970350404312,
1226
+ "grad_norm": 0.5338106751441956,
1227
+ "learning_rate": 0.0005475574743658931,
1228
+ "loss": 3.8772,
1229
+ "step": 8200
1230
+ },
1231
+ {
1232
+ "epoch": 0.889487870619946,
1233
+ "grad_norm": 0.6587180495262146,
1234
+ "learning_rate": 0.0005472336751214246,
1235
+ "loss": 3.8723,
1236
+ "step": 8250
1237
+ },
1238
+ {
1239
+ "epoch": 0.894878706199461,
1240
+ "grad_norm": 0.6544439196586609,
1241
+ "learning_rate": 0.0005469098758769562,
1242
+ "loss": 3.856,
1243
+ "step": 8300
1244
+ },
1245
+ {
1246
+ "epoch": 0.9002695417789758,
1247
+ "grad_norm": 0.5896086692810059,
1248
+ "learning_rate": 0.0005465860766324878,
1249
+ "loss": 3.8439,
1250
+ "step": 8350
1251
+ },
1252
+ {
1253
+ "epoch": 0.9056603773584906,
1254
+ "grad_norm": 0.579795241355896,
1255
+ "learning_rate": 0.0005462622773880194,
1256
+ "loss": 3.892,
1257
+ "step": 8400
1258
+ },
1259
+ {
1260
+ "epoch": 0.9110512129380054,
1261
+ "grad_norm": 0.6916389465332031,
1262
+ "learning_rate": 0.000545938478143551,
1263
+ "loss": 3.8569,
1264
+ "step": 8450
1265
+ },
1266
+ {
1267
+ "epoch": 0.9164420485175202,
1268
+ "grad_norm": 0.5650646686553955,
1269
+ "learning_rate": 0.0005456146788990825,
1270
+ "loss": 3.8475,
1271
+ "step": 8500
1272
+ },
1273
+ {
1274
+ "epoch": 0.921832884097035,
1275
+ "grad_norm": 0.5492226481437683,
1276
+ "learning_rate": 0.000545290879654614,
1277
+ "loss": 3.86,
1278
+ "step": 8550
1279
+ },
1280
+ {
1281
+ "epoch": 0.9272237196765498,
1282
+ "grad_norm": 0.6170901656150818,
1283
+ "learning_rate": 0.0005449670804101457,
1284
+ "loss": 3.86,
1285
+ "step": 8600
1286
+ },
1287
+ {
1288
+ "epoch": 0.9326145552560647,
1289
+ "grad_norm": 0.605499804019928,
1290
+ "learning_rate": 0.0005446432811656773,
1291
+ "loss": 3.8616,
1292
+ "step": 8650
1293
+ },
1294
+ {
1295
+ "epoch": 0.9380053908355795,
1296
+ "grad_norm": 0.5699788331985474,
1297
+ "learning_rate": 0.0005443194819212088,
1298
+ "loss": 3.8398,
1299
+ "step": 8700
1300
+ },
1301
+ {
1302
+ "epoch": 0.9433962264150944,
1303
+ "grad_norm": 0.6459490656852722,
1304
+ "learning_rate": 0.0005439956826767404,
1305
+ "loss": 3.8561,
1306
+ "step": 8750
1307
+ },
1308
+ {
1309
+ "epoch": 0.9487870619946092,
1310
+ "grad_norm": 0.5878280401229858,
1311
+ "learning_rate": 0.000543671883432272,
1312
+ "loss": 3.8437,
1313
+ "step": 8800
1314
+ },
1315
+ {
1316
+ "epoch": 0.954177897574124,
1317
+ "grad_norm": 0.5819820165634155,
1318
+ "learning_rate": 0.0005433480841878035,
1319
+ "loss": 3.8392,
1320
+ "step": 8850
1321
+ },
1322
+ {
1323
+ "epoch": 0.9595687331536388,
1324
+ "grad_norm": 0.6361739635467529,
1325
+ "learning_rate": 0.000543024284943335,
1326
+ "loss": 3.8312,
1327
+ "step": 8900
1328
+ },
1329
+ {
1330
+ "epoch": 0.9649595687331537,
1331
+ "grad_norm": 0.6311874389648438,
1332
+ "learning_rate": 0.0005427004856988667,
1333
+ "loss": 3.8481,
1334
+ "step": 8950
1335
+ },
1336
+ {
1337
+ "epoch": 0.9703504043126685,
1338
+ "grad_norm": 0.6704027056694031,
1339
+ "learning_rate": 0.0005423766864543982,
1340
+ "loss": 3.848,
1341
+ "step": 9000
1342
+ },
1343
+ {
1344
+ "epoch": 0.9703504043126685,
1345
+ "eval_accuracy": 0.3414659295509887,
1346
+ "eval_loss": 3.779182195663452,
1347
+ "eval_runtime": 184.1503,
1348
+ "eval_samples_per_second": 97.806,
1349
+ "eval_steps_per_second": 6.115,
1350
+ "step": 9000
1351
+ },
1352
+ {
1353
+ "epoch": 0.9757412398921833,
1354
+ "grad_norm": 0.5311286449432373,
1355
+ "learning_rate": 0.0005420528872099298,
1356
+ "loss": 3.8298,
1357
+ "step": 9050
1358
+ },
1359
+ {
1360
+ "epoch": 0.9811320754716981,
1361
+ "grad_norm": 0.5362455248832703,
1362
+ "learning_rate": 0.0005417290879654613,
1363
+ "loss": 3.8289,
1364
+ "step": 9100
1365
+ },
1366
+ {
1367
+ "epoch": 0.9865229110512129,
1368
+ "grad_norm": 0.6195308566093445,
1369
+ "learning_rate": 0.000541405288720993,
1370
+ "loss": 3.8465,
1371
+ "step": 9150
1372
+ },
1373
+ {
1374
+ "epoch": 0.9919137466307277,
1375
+ "grad_norm": 0.5365903973579407,
1376
+ "learning_rate": 0.0005410814894765245,
1377
+ "loss": 3.8248,
1378
+ "step": 9200
1379
+ },
1380
+ {
1381
+ "epoch": 0.9973045822102425,
1382
+ "grad_norm": 0.652845025062561,
1383
+ "learning_rate": 0.0005407576902320561,
1384
+ "loss": 3.8518,
1385
+ "step": 9250
1386
+ },
1387
+ {
1388
+ "epoch": 1.0026954177897573,
1389
+ "grad_norm": 0.60700523853302,
1390
+ "learning_rate": 0.0005404338909875876,
1391
+ "loss": 3.7881,
1392
+ "step": 9300
1393
+ },
1394
+ {
1395
+ "epoch": 1.0080862533692723,
1396
+ "grad_norm": 0.6487429738044739,
1397
+ "learning_rate": 0.0005401100917431192,
1398
+ "loss": 3.7647,
1399
+ "step": 9350
1400
+ },
1401
+ {
1402
+ "epoch": 1.013477088948787,
1403
+ "grad_norm": 0.561632513999939,
1404
+ "learning_rate": 0.0005397862924986508,
1405
+ "loss": 3.7791,
1406
+ "step": 9400
1407
+ },
1408
+ {
1409
+ "epoch": 1.0188679245283019,
1410
+ "grad_norm": 0.5459903478622437,
1411
+ "learning_rate": 0.0005394624932541824,
1412
+ "loss": 3.7855,
1413
+ "step": 9450
1414
+ },
1415
+ {
1416
+ "epoch": 1.0242587601078168,
1417
+ "grad_norm": 0.5852130055427551,
1418
+ "learning_rate": 0.0005391386940097139,
1419
+ "loss": 3.7552,
1420
+ "step": 9500
1421
+ },
1422
+ {
1423
+ "epoch": 1.0296495956873315,
1424
+ "grad_norm": 0.6032885909080505,
1425
+ "learning_rate": 0.0005388148947652455,
1426
+ "loss": 3.7671,
1427
+ "step": 9550
1428
+ },
1429
+ {
1430
+ "epoch": 1.0350404312668464,
1431
+ "grad_norm": 0.6321649551391602,
1432
+ "learning_rate": 0.000538491095520777,
1433
+ "loss": 3.795,
1434
+ "step": 9600
1435
+ },
1436
+ {
1437
+ "epoch": 1.0404312668463611,
1438
+ "grad_norm": 0.615997314453125,
1439
+ "learning_rate": 0.0005381672962763086,
1440
+ "loss": 3.7544,
1441
+ "step": 9650
1442
+ },
1443
+ {
1444
+ "epoch": 1.045822102425876,
1445
+ "grad_norm": 0.6234314441680908,
1446
+ "learning_rate": 0.0005378434970318403,
1447
+ "loss": 3.7772,
1448
+ "step": 9700
1449
+ },
1450
+ {
1451
+ "epoch": 1.0512129380053907,
1452
+ "grad_norm": 0.638446569442749,
1453
+ "learning_rate": 0.0005375196977873718,
1454
+ "loss": 3.753,
1455
+ "step": 9750
1456
+ },
1457
+ {
1458
+ "epoch": 1.0566037735849056,
1459
+ "grad_norm": 0.6064075827598572,
1460
+ "learning_rate": 0.0005371958985429034,
1461
+ "loss": 3.765,
1462
+ "step": 9800
1463
+ },
1464
+ {
1465
+ "epoch": 1.0619946091644206,
1466
+ "grad_norm": 0.572862982749939,
1467
+ "learning_rate": 0.0005368720992984349,
1468
+ "loss": 3.7706,
1469
+ "step": 9850
1470
+ },
1471
+ {
1472
+ "epoch": 1.0673854447439353,
1473
+ "grad_norm": 0.6439206600189209,
1474
+ "learning_rate": 0.0005365483000539665,
1475
+ "loss": 3.7649,
1476
+ "step": 9900
1477
+ },
1478
+ {
1479
+ "epoch": 1.0727762803234502,
1480
+ "grad_norm": 0.5527343153953552,
1481
+ "learning_rate": 0.0005362245008094981,
1482
+ "loss": 3.7804,
1483
+ "step": 9950
1484
+ },
1485
+ {
1486
+ "epoch": 1.0781671159029649,
1487
+ "grad_norm": 0.5652170777320862,
1488
+ "learning_rate": 0.0005359007015650297,
1489
+ "loss": 3.7701,
1490
+ "step": 10000
1491
+ },
1492
+ {
1493
+ "epoch": 1.0781671159029649,
1494
+ "eval_accuracy": 0.3446935722364057,
1495
+ "eval_loss": 3.7489497661590576,
1496
+ "eval_runtime": 184.1084,
1497
+ "eval_samples_per_second": 97.828,
1498
+ "eval_steps_per_second": 6.116,
1499
+ "step": 10000
1500
+ }
1501
+ ],
1502
+ "logging_steps": 50,
1503
+ "max_steps": 92750,
1504
+ "num_input_tokens_seen": 0,
1505
+ "num_train_epochs": 10,
1506
+ "save_steps": 10000,
1507
+ "stateful_callbacks": {
1508
+ "TrainerControl": {
1509
+ "args": {
1510
+ "should_epoch_stop": false,
1511
+ "should_evaluate": false,
1512
+ "should_log": false,
1513
+ "should_save": true,
1514
+ "should_training_stop": false
1515
+ },
1516
+ "attributes": {}
1517
+ }
1518
+ },
1519
+ "total_flos": 8.36069179392e+16,
1520
+ "train_batch_size": 32,
1521
+ "trial_name": null,
1522
+ "trial_params": null
1523
+ }
checkpoint-10000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac28028fd9c53af1f06caf0d4841e7de5acf445d150f28e624c595d75b8f8c13
3
+ size 5304