Hugging Face's logo

tttx
/
models-p10-limit-data-step1

PEFT
Safetensors
alignment-handbook
trl
sft
Generated from Trainer
Model card Files Community
aadityap commited on
Commit
a0b94ea
·
verified ·
1 Parent(s): ef77660

Upload trainer_state.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. trainer_state.json +270 -0
trainer_state.json ADDED
@@ -0,0 +1,270 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 4.0,
5
+ "eval_steps": 500,
6
+ "global_step": 28,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.14285714285714285,
13
+ "grad_norm": 0.029374502948890973,
14
+ "learning_rate": 1.6666666666666667e-05,
15
+ "loss": 0.2105,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.2857142857142857,
20
+ "grad_norm": 0.02879271003469766,
21
+ "learning_rate": 3.3333333333333335e-05,
22
+ "loss": 0.2059,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.42857142857142855,
27
+ "grad_norm": 0.028417140992448447,
28
+ "learning_rate": 5e-05,
29
+ "loss": 0.2136,
30
+ "step": 3
31
+ },
32
+ {
33
+ "epoch": 0.5714285714285714,
34
+ "grad_norm": 0.029508379259060556,
35
+ "learning_rate": 4.980286753286195e-05,
36
+ "loss": 0.2064,
37
+ "step": 4
38
+ },
39
+ {
40
+ "epoch": 0.7142857142857143,
41
+ "grad_norm": 0.031023789828790938,
42
+ "learning_rate": 4.9214579028215776e-05,
43
+ "loss": 0.2051,
44
+ "step": 5
45
+ },
46
+ {
47
+ "epoch": 0.8571428571428571,
48
+ "grad_norm": 0.03173373208486436,
49
+ "learning_rate": 4.8244412147206284e-05,
50
+ "loss": 0.2009,
51
+ "step": 6
52
+ },
53
+ {
54
+ "epoch": 1.0,
55
+ "grad_norm": 0.02891593307804275,
56
+ "learning_rate": 4.690766700109659e-05,
57
+ "loss": 0.1928,
58
+ "step": 7
59
+ },
60
+ {
61
+ "epoch": 1.0,
62
+ "eval_loss": 0.17100295424461365,
63
+ "eval_runtime": 2.5986,
64
+ "eval_samples_per_second": 0.385,
65
+ "eval_steps_per_second": 0.385,
66
+ "step": 7
67
+ },
68
+ {
69
+ "epoch": 1.1428571428571428,
70
+ "grad_norm": 0.0291759534514554,
71
+ "learning_rate": 4.522542485937369e-05,
72
+ "loss": 0.1816,
73
+ "step": 8
74
+ },
75
+ {
76
+ "epoch": 1.2857142857142856,
77
+ "grad_norm": 0.03503069911331975,
78
+ "learning_rate": 4.3224215685535294e-05,
79
+ "loss": 0.1771,
80
+ "step": 9
81
+ },
82
+ {
83
+ "epoch": 1.4285714285714286,
84
+ "grad_norm": 0.03222295532632566,
85
+ "learning_rate": 4.093559974371725e-05,
86
+ "loss": 0.1714,
87
+ "step": 10
88
+ },
89
+ {
90
+ "epoch": 1.5714285714285714,
91
+ "grad_norm": 0.02924500419270857,
92
+ "learning_rate": 3.8395669874474915e-05,
93
+ "loss": 0.176,
94
+ "step": 11
95
+ },
96
+ {
97
+ "epoch": 1.7142857142857144,
98
+ "grad_norm": 0.025032804888512103,
99
+ "learning_rate": 3.564448228912682e-05,
100
+ "loss": 0.1688,
101
+ "step": 12
102
+ },
103
+ {
104
+ "epoch": 1.8571428571428572,
105
+ "grad_norm": 0.02333745520711948,
106
+ "learning_rate": 3.272542485937369e-05,
107
+ "loss": 0.1641,
108
+ "step": 13
109
+ },
110
+ {
111
+ "epoch": 2.0,
112
+ "grad_norm": 0.02299528178586051,
113
+ "learning_rate": 2.9684532864643122e-05,
114
+ "loss": 0.1516,
115
+ "step": 14
116
+ },
117
+ {
118
+ "epoch": 2.0,
119
+ "eval_loss": 0.14068603515625,
120
+ "eval_runtime": 2.4437,
121
+ "eval_samples_per_second": 0.409,
122
+ "eval_steps_per_second": 0.409,
123
+ "step": 14
124
+ },
125
+ {
126
+ "epoch": 2.142857142857143,
127
+ "grad_norm": 0.02266916678811667,
128
+ "learning_rate": 2.656976298823284e-05,
129
+ "loss": 0.1553,
130
+ "step": 15
131
+ },
132
+ {
133
+ "epoch": 2.2857142857142856,
134
+ "grad_norm": 0.02244871211073816,
135
+ "learning_rate": 2.3430237011767167e-05,
136
+ "loss": 0.1484,
137
+ "step": 16
138
+ },
139
+ {
140
+ "epoch": 2.4285714285714284,
141
+ "grad_norm": 0.024630119269621335,
142
+ "learning_rate": 2.031546713535688e-05,
143
+ "loss": 0.1489,
144
+ "step": 17
145
+ },
146
+ {
147
+ "epoch": 2.571428571428571,
148
+ "grad_norm": 0.02310463707931144,
149
+ "learning_rate": 1.7274575140626318e-05,
150
+ "loss": 0.1433,
151
+ "step": 18
152
+ },
153
+ {
154
+ "epoch": 2.7142857142857144,
155
+ "grad_norm": 0.022565531397988565,
156
+ "learning_rate": 1.4355517710873184e-05,
157
+ "loss": 0.145,
158
+ "step": 19
159
+ },
160
+ {
161
+ "epoch": 2.857142857142857,
162
+ "grad_norm": 0.02303743715249304,
163
+ "learning_rate": 1.1604330125525079e-05,
164
+ "loss": 0.1388,
165
+ "step": 20
166
+ },
167
+ {
168
+ "epoch": 3.0,
169
+ "grad_norm": 0.02213833372976756,
170
+ "learning_rate": 9.064400256282757e-06,
171
+ "loss": 0.1408,
172
+ "step": 21
173
+ },
174
+ {
175
+ "epoch": 3.0,
176
+ "eval_loss": 0.12390902638435364,
177
+ "eval_runtime": 2.5093,
178
+ "eval_samples_per_second": 0.399,
179
+ "eval_steps_per_second": 0.399,
180
+ "step": 21
181
+ },
182
+ {
183
+ "epoch": 3.142857142857143,
184
+ "grad_norm": 0.02206988787856512,
185
+ "learning_rate": 6.775784314464717e-06,
186
+ "loss": 0.1386,
187
+ "step": 22
188
+ },
189
+ {
190
+ "epoch": 3.2857142857142856,
191
+ "grad_norm": 0.02137909507738522,
192
+ "learning_rate": 4.7745751406263165e-06,
193
+ "loss": 0.1363,
194
+ "step": 23
195
+ },
196
+ {
197
+ "epoch": 3.4285714285714284,
198
+ "grad_norm": 0.021550899474365476,
199
+ "learning_rate": 3.092332998903416e-06,
200
+ "loss": 0.1256,
201
+ "step": 24
202
+ },
203
+ {
204
+ "epoch": 3.571428571428571,
205
+ "grad_norm": 0.021209476107635006,
206
+ "learning_rate": 1.7555878527937164e-06,
207
+ "loss": 0.1366,
208
+ "step": 25
209
+ },
210
+ {
211
+ "epoch": 3.7142857142857144,
212
+ "grad_norm": 0.02118940295971805,
213
+ "learning_rate": 7.854209717842231e-07,
214
+ "loss": 0.1325,
215
+ "step": 26
216
+ },
217
+ {
218
+ "epoch": 3.857142857142857,
219
+ "grad_norm": 0.02088779986332402,
220
+ "learning_rate": 1.9713246713805588e-07,
221
+ "loss": 0.1335,
222
+ "step": 27
223
+ },
224
+ {
225
+ "epoch": 4.0,
226
+ "grad_norm": 0.021082645298901994,
227
+ "learning_rate": 0.0,
228
+ "loss": 0.1367,
229
+ "step": 28
230
+ },
231
+ {
232
+ "epoch": 4.0,
233
+ "eval_loss": 0.12003877758979797,
234
+ "eval_runtime": 2.4052,
235
+ "eval_samples_per_second": 0.416,
236
+ "eval_steps_per_second": 0.416,
237
+ "step": 28
238
+ },
239
+ {
240
+ "epoch": 4.0,
241
+ "step": 28,
242
+ "total_flos": 613369774080000.0,
243
+ "train_loss": 0.1637916559619563,
244
+ "train_runtime": 908.7009,
245
+ "train_samples_per_second": 0.88,
246
+ "train_steps_per_second": 0.031
247
+ }
248
+ ],
249
+ "logging_steps": 1,
250
+ "max_steps": 28,
251
+ "num_input_tokens_seen": 0,
252
+ "num_train_epochs": 4,
253
+ "save_steps": 500,
254
+ "stateful_callbacks": {
255
+ "TrainerControl": {
256
+ "args": {
257
+ "should_epoch_stop": false,
258
+ "should_evaluate": false,
259
+ "should_log": false,
260
+ "should_save": true,
261
+ "should_training_stop": true
262
+ },
263
+ "attributes": {}
264
+ }
265
+ },
266
+ "total_flos": 613369774080000.0,
267
+ "train_batch_size": 4,
268
+ "trial_name": null,
269
+ "trial_params": null
270
+ }