terry69 commited on
Commit
ecef6e0
·
verified ·
1 Parent(s): bdf88c1

Model save

Browse files
Files changed (4) hide show
  1. README.md +69 -0
  2. all_results.json +9 -0
  3. train_results.json +9 -0
  4. trainer_state.json +1380 -0
README.md ADDED
@@ -0,0 +1,69 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: llama3
3
+ library_name: peft
4
+ tags:
5
+ - trl
6
+ - sft
7
+ - generated_from_trainer
8
+ base_model: meta-llama/Meta-Llama-3-8B
9
+ datasets:
10
+ - generator
11
+ model-index:
12
+ - name: downstream-7b
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # downstream-7b
20
+
21
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 1.0792
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 0.0002
43
+ - train_batch_size: 4
44
+ - eval_batch_size: 1
45
+ - seed: 42
46
+ - distributed_type: multi-GPU
47
+ - num_devices: 4
48
+ - gradient_accumulation_steps: 4
49
+ - total_train_batch_size: 64
50
+ - total_eval_batch_size: 4
51
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
52
+ - lr_scheduler_type: cosine
53
+ - lr_scheduler_warmup_ratio: 0.1
54
+ - num_epochs: 1
55
+
56
+ ### Training results
57
+
58
+ | Training Loss | Epoch | Step | Validation Loss |
59
+ |:-------------:|:------:|:----:|:---------------:|
60
+ | 1.0709 | 0.9992 | 949 | 1.0792 |
61
+
62
+
63
+ ### Framework versions
64
+
65
+ - PEFT 0.11.1
66
+ - Transformers 4.43.4
67
+ - Pytorch 2.3.1+cu121
68
+ - Datasets 2.19.1
69
+ - Tokenizers 0.19.1
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9992103185048697,
3
+ "total_flos": 1959448100732928.0,
4
+ "train_loss": 1.0930153100081064,
5
+ "train_runtime": 22340.3866,
6
+ "train_samples": 103932,
7
+ "train_samples_per_second": 2.72,
8
+ "train_steps_per_second": 0.042
9
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.9992103185048697,
3
+ "total_flos": 1959448100732928.0,
4
+ "train_loss": 1.0930153100081064,
5
+ "train_runtime": 22340.3866,
6
+ "train_samples": 103932,
7
+ "train_samples_per_second": 2.72,
8
+ "train_steps_per_second": 0.042
9
+ }
trainer_state.json ADDED
@@ -0,0 +1,1380 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.9992103185048697,
5
+ "eval_steps": 500,
6
+ "global_step": 949,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.0010529086601737299,
13
+ "grad_norm": 4.026114469450001,
14
+ "learning_rate": 2.105263157894737e-06,
15
+ "loss": 1.3755,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.0052645433008686494,
20
+ "grad_norm": 1.1251945428540173,
21
+ "learning_rate": 1.0526315789473684e-05,
22
+ "loss": 1.3102,
23
+ "step": 5
24
+ },
25
+ {
26
+ "epoch": 0.010529086601737299,
27
+ "grad_norm": 0.563945023031292,
28
+ "learning_rate": 2.105263157894737e-05,
29
+ "loss": 1.2626,
30
+ "step": 10
31
+ },
32
+ {
33
+ "epoch": 0.01579362990260595,
34
+ "grad_norm": 0.4336837922055097,
35
+ "learning_rate": 3.157894736842105e-05,
36
+ "loss": 1.2133,
37
+ "step": 15
38
+ },
39
+ {
40
+ "epoch": 0.021058173203474598,
41
+ "grad_norm": 0.33707427690007363,
42
+ "learning_rate": 4.210526315789474e-05,
43
+ "loss": 1.1636,
44
+ "step": 20
45
+ },
46
+ {
47
+ "epoch": 0.026322716504343247,
48
+ "grad_norm": 0.2412443925117212,
49
+ "learning_rate": 5.2631578947368424e-05,
50
+ "loss": 1.1845,
51
+ "step": 25
52
+ },
53
+ {
54
+ "epoch": 0.0315872598052119,
55
+ "grad_norm": 0.2414606698915264,
56
+ "learning_rate": 6.31578947368421e-05,
57
+ "loss": 1.1401,
58
+ "step": 30
59
+ },
60
+ {
61
+ "epoch": 0.03685180310608055,
62
+ "grad_norm": 0.22628642837844729,
63
+ "learning_rate": 7.368421052631579e-05,
64
+ "loss": 1.1591,
65
+ "step": 35
66
+ },
67
+ {
68
+ "epoch": 0.042116346406949196,
69
+ "grad_norm": 0.2208133931977146,
70
+ "learning_rate": 8.421052631578948e-05,
71
+ "loss": 1.1397,
72
+ "step": 40
73
+ },
74
+ {
75
+ "epoch": 0.04738088970781785,
76
+ "grad_norm": 0.21905054641153135,
77
+ "learning_rate": 9.473684210526316e-05,
78
+ "loss": 1.1295,
79
+ "step": 45
80
+ },
81
+ {
82
+ "epoch": 0.052645433008686494,
83
+ "grad_norm": 0.1951061692287286,
84
+ "learning_rate": 0.00010526315789473685,
85
+ "loss": 1.141,
86
+ "step": 50
87
+ },
88
+ {
89
+ "epoch": 0.05790997630955515,
90
+ "grad_norm": 0.18792729468212402,
91
+ "learning_rate": 0.00011578947368421053,
92
+ "loss": 1.121,
93
+ "step": 55
94
+ },
95
+ {
96
+ "epoch": 0.0631745196104238,
97
+ "grad_norm": 0.189919407852987,
98
+ "learning_rate": 0.0001263157894736842,
99
+ "loss": 1.1347,
100
+ "step": 60
101
+ },
102
+ {
103
+ "epoch": 0.06843906291129244,
104
+ "grad_norm": 0.18793881851552416,
105
+ "learning_rate": 0.0001368421052631579,
106
+ "loss": 1.0961,
107
+ "step": 65
108
+ },
109
+ {
110
+ "epoch": 0.0737036062121611,
111
+ "grad_norm": 0.18263188517758086,
112
+ "learning_rate": 0.00014736842105263158,
113
+ "loss": 1.0937,
114
+ "step": 70
115
+ },
116
+ {
117
+ "epoch": 0.07896814951302975,
118
+ "grad_norm": 0.18520098125405152,
119
+ "learning_rate": 0.00015789473684210527,
120
+ "loss": 1.1419,
121
+ "step": 75
122
+ },
123
+ {
124
+ "epoch": 0.08423269281389839,
125
+ "grad_norm": 0.18675524775004465,
126
+ "learning_rate": 0.00016842105263157895,
127
+ "loss": 1.1094,
128
+ "step": 80
129
+ },
130
+ {
131
+ "epoch": 0.08949723611476705,
132
+ "grad_norm": 0.18469057661828525,
133
+ "learning_rate": 0.00017894736842105264,
134
+ "loss": 1.0952,
135
+ "step": 85
136
+ },
137
+ {
138
+ "epoch": 0.0947617794156357,
139
+ "grad_norm": 0.17860571701450936,
140
+ "learning_rate": 0.00018947368421052632,
141
+ "loss": 1.1035,
142
+ "step": 90
143
+ },
144
+ {
145
+ "epoch": 0.10002632271650434,
146
+ "grad_norm": 0.2032976356381528,
147
+ "learning_rate": 0.0002,
148
+ "loss": 1.1329,
149
+ "step": 95
150
+ },
151
+ {
152
+ "epoch": 0.10529086601737299,
153
+ "grad_norm": 0.18932762375677964,
154
+ "learning_rate": 0.0001999830846194422,
155
+ "loss": 1.0902,
156
+ "step": 100
157
+ },
158
+ {
159
+ "epoch": 0.11055540931824165,
160
+ "grad_norm": 0.17823414518835126,
161
+ "learning_rate": 0.00019993234420037073,
162
+ "loss": 1.0951,
163
+ "step": 105
164
+ },
165
+ {
166
+ "epoch": 0.1158199526191103,
167
+ "grad_norm": 0.19033211796864122,
168
+ "learning_rate": 0.00019984779590865556,
169
+ "loss": 1.11,
170
+ "step": 110
171
+ },
172
+ {
173
+ "epoch": 0.12108449591997894,
174
+ "grad_norm": 0.1781090004150184,
175
+ "learning_rate": 0.0001997294683476273,
176
+ "loss": 1.1216,
177
+ "step": 115
178
+ },
179
+ {
180
+ "epoch": 0.1263490392208476,
181
+ "grad_norm": 0.20142566240295628,
182
+ "learning_rate": 0.0001995774015484005,
183
+ "loss": 1.088,
184
+ "step": 120
185
+ },
186
+ {
187
+ "epoch": 0.13161358252171623,
188
+ "grad_norm": 0.16738672746077932,
189
+ "learning_rate": 0.00019939164695633067,
190
+ "loss": 1.1069,
191
+ "step": 125
192
+ },
193
+ {
194
+ "epoch": 0.13687812582258488,
195
+ "grad_norm": 0.17141306079033702,
196
+ "learning_rate": 0.00019917226741361015,
197
+ "loss": 1.1178,
198
+ "step": 130
199
+ },
200
+ {
201
+ "epoch": 0.14214266912345355,
202
+ "grad_norm": 0.18242919862111662,
203
+ "learning_rate": 0.00019891933713800798,
204
+ "loss": 1.115,
205
+ "step": 135
206
+ },
207
+ {
208
+ "epoch": 0.1474072124243222,
209
+ "grad_norm": 0.18858703761293544,
210
+ "learning_rate": 0.00019863294169776148,
211
+ "loss": 1.092,
212
+ "step": 140
213
+ },
214
+ {
215
+ "epoch": 0.15267175572519084,
216
+ "grad_norm": 0.1851910906506613,
217
+ "learning_rate": 0.00019831317798262786,
218
+ "loss": 1.1015,
219
+ "step": 145
220
+ },
221
+ {
222
+ "epoch": 0.1579362990260595,
223
+ "grad_norm": 0.17061718616532065,
224
+ "learning_rate": 0.00019796015417110577,
225
+ "loss": 1.0834,
226
+ "step": 150
227
+ },
228
+ {
229
+ "epoch": 0.16320084232692814,
230
+ "grad_norm": 0.19083175263550564,
231
+ "learning_rate": 0.0001975739896938375,
232
+ "loss": 1.0915,
233
+ "step": 155
234
+ },
235
+ {
236
+ "epoch": 0.16846538562779678,
237
+ "grad_norm": 0.17041981222039004,
238
+ "learning_rate": 0.00019715481519320496,
239
+ "loss": 1.1045,
240
+ "step": 160
241
+ },
242
+ {
243
+ "epoch": 0.17372992892866543,
244
+ "grad_norm": 0.17539080495334333,
245
+ "learning_rate": 0.00019670277247913205,
246
+ "loss": 1.0822,
247
+ "step": 165
248
+ },
249
+ {
250
+ "epoch": 0.1789944722295341,
251
+ "grad_norm": 0.16847918243353582,
252
+ "learning_rate": 0.00019621801448110952,
253
+ "loss": 1.1113,
254
+ "step": 170
255
+ },
256
+ {
257
+ "epoch": 0.18425901553040275,
258
+ "grad_norm": 0.16577520965121645,
259
+ "learning_rate": 0.00019570070519645767,
260
+ "loss": 1.0726,
261
+ "step": 175
262
+ },
263
+ {
264
+ "epoch": 0.1895235588312714,
265
+ "grad_norm": 0.17216940817918563,
266
+ "learning_rate": 0.00019515101963484485,
267
+ "loss": 1.1214,
268
+ "step": 180
269
+ },
270
+ {
271
+ "epoch": 0.19478810213214004,
272
+ "grad_norm": 0.16717603329959776,
273
+ "learning_rate": 0.00019456914375908023,
274
+ "loss": 1.0749,
275
+ "step": 185
276
+ },
277
+ {
278
+ "epoch": 0.20005264543300869,
279
+ "grad_norm": 0.16743311436795275,
280
+ "learning_rate": 0.0001939552744222014,
281
+ "loss": 1.0856,
282
+ "step": 190
283
+ },
284
+ {
285
+ "epoch": 0.20531718873387733,
286
+ "grad_norm": 0.16628473925396028,
287
+ "learning_rate": 0.00019330961930087725,
288
+ "loss": 1.1088,
289
+ "step": 195
290
+ },
291
+ {
292
+ "epoch": 0.21058173203474598,
293
+ "grad_norm": 0.1735673710306468,
294
+ "learning_rate": 0.00019263239682514952,
295
+ "loss": 1.094,
296
+ "step": 200
297
+ },
298
+ {
299
+ "epoch": 0.21584627533561462,
300
+ "grad_norm": 0.16463208491106188,
301
+ "learning_rate": 0.00019192383610453618,
302
+ "loss": 1.1191,
303
+ "step": 205
304
+ },
305
+ {
306
+ "epoch": 0.2211108186364833,
307
+ "grad_norm": 0.1697201646634803,
308
+ "learning_rate": 0.00019118417685052194,
309
+ "loss": 1.1188,
310
+ "step": 210
311
+ },
312
+ {
313
+ "epoch": 0.22637536193735194,
314
+ "grad_norm": 0.15930348674889006,
315
+ "learning_rate": 0.00019041366929546219,
316
+ "loss": 1.1132,
317
+ "step": 215
318
+ },
319
+ {
320
+ "epoch": 0.2316399052382206,
321
+ "grad_norm": 0.16154587528605638,
322
+ "learning_rate": 0.0001896125741079272,
323
+ "loss": 1.1029,
324
+ "step": 220
325
+ },
326
+ {
327
+ "epoch": 0.23690444853908924,
328
+ "grad_norm": 0.1689593754891321,
329
+ "learning_rate": 0.00018878116230451613,
330
+ "loss": 1.1196,
331
+ "step": 225
332
+ },
333
+ {
334
+ "epoch": 0.24216899183995788,
335
+ "grad_norm": 0.158620581331537,
336
+ "learning_rate": 0.0001879197151581702,
337
+ "loss": 1.0786,
338
+ "step": 230
339
+ },
340
+ {
341
+ "epoch": 0.24743353514082653,
342
+ "grad_norm": 0.1591976970503649,
343
+ "learning_rate": 0.00018702852410301554,
344
+ "loss": 1.0861,
345
+ "step": 235
346
+ },
347
+ {
348
+ "epoch": 0.2526980784416952,
349
+ "grad_norm": 0.16271741933565712,
350
+ "learning_rate": 0.00018610789063576913,
351
+ "loss": 1.077,
352
+ "step": 240
353
+ },
354
+ {
355
+ "epoch": 0.25796262174256385,
356
+ "grad_norm": 0.16691583214883504,
357
+ "learning_rate": 0.00018515812621373997,
358
+ "loss": 1.0931,
359
+ "step": 245
360
+ },
361
+ {
362
+ "epoch": 0.26322716504343247,
363
+ "grad_norm": 0.15795453416798677,
364
+ "learning_rate": 0.00018417955214946092,
365
+ "loss": 1.0929,
366
+ "step": 250
367
+ },
368
+ {
369
+ "epoch": 0.26849170834430114,
370
+ "grad_norm": 0.15734119895920037,
371
+ "learning_rate": 0.00018317249950198597,
372
+ "loss": 1.086,
373
+ "step": 255
374
+ },
375
+ {
376
+ "epoch": 0.27375625164516976,
377
+ "grad_norm": 0.15815121783491273,
378
+ "learning_rate": 0.0001821373089648906,
379
+ "loss": 1.1142,
380
+ "step": 260
381
+ },
382
+ {
383
+ "epoch": 0.27902079494603843,
384
+ "grad_norm": 0.15790684873329372,
385
+ "learning_rate": 0.00018107433075101252,
386
+ "loss": 1.0907,
387
+ "step": 265
388
+ },
389
+ {
390
+ "epoch": 0.2842853382469071,
391
+ "grad_norm": 0.1603612919235376,
392
+ "learning_rate": 0.00017998392447397197,
393
+ "loss": 1.103,
394
+ "step": 270
395
+ },
396
+ {
397
+ "epoch": 0.2895498815477757,
398
+ "grad_norm": 0.1935643212000403,
399
+ "learning_rate": 0.00017886645902651167,
400
+ "loss": 1.1207,
401
+ "step": 275
402
+ },
403
+ {
404
+ "epoch": 0.2948144248486444,
405
+ "grad_norm": 0.16197395404790052,
406
+ "learning_rate": 0.0001777223124556978,
407
+ "loss": 1.1036,
408
+ "step": 280
409
+ },
410
+ {
411
+ "epoch": 0.300078968149513,
412
+ "grad_norm": 0.16503760296000086,
413
+ "learning_rate": 0.00017655187183502344,
414
+ "loss": 1.0647,
415
+ "step": 285
416
+ },
417
+ {
418
+ "epoch": 0.3053435114503817,
419
+ "grad_norm": 0.1772283442967409,
420
+ "learning_rate": 0.00017535553313345904,
421
+ "loss": 1.1075,
422
+ "step": 290
423
+ },
424
+ {
425
+ "epoch": 0.3106080547512503,
426
+ "grad_norm": 0.16282645295325013,
427
+ "learning_rate": 0.00017413370108149286,
428
+ "loss": 1.1094,
429
+ "step": 295
430
+ },
431
+ {
432
+ "epoch": 0.315872598052119,
433
+ "grad_norm": 0.15561402718068354,
434
+ "learning_rate": 0.00017288678903420762,
435
+ "loss": 1.0776,
436
+ "step": 300
437
+ },
438
+ {
439
+ "epoch": 0.32113714135298765,
440
+ "grad_norm": 0.15594031474920508,
441
+ "learning_rate": 0.00017161521883143934,
442
+ "loss": 1.1078,
443
+ "step": 305
444
+ },
445
+ {
446
+ "epoch": 0.32640168465385627,
447
+ "grad_norm": 0.1570395383591175,
448
+ "learning_rate": 0.00017031942065506576,
449
+ "loss": 1.1124,
450
+ "step": 310
451
+ },
452
+ {
453
+ "epoch": 0.33166622795472495,
454
+ "grad_norm": 0.15773152315944608,
455
+ "learning_rate": 0.00016899983288347248,
456
+ "loss": 1.0913,
457
+ "step": 315
458
+ },
459
+ {
460
+ "epoch": 0.33693077125559356,
461
+ "grad_norm": 0.15310806808595664,
462
+ "learning_rate": 0.00016765690194324616,
463
+ "loss": 1.0845,
464
+ "step": 320
465
+ },
466
+ {
467
+ "epoch": 0.34219531455646224,
468
+ "grad_norm": 0.16384678369715433,
469
+ "learning_rate": 0.00016629108215814525,
470
+ "loss": 1.1173,
471
+ "step": 325
472
+ },
473
+ {
474
+ "epoch": 0.34745985785733086,
475
+ "grad_norm": 0.165818325184464,
476
+ "learning_rate": 0.00016490283559539838,
477
+ "loss": 1.1014,
478
+ "step": 330
479
+ },
480
+ {
481
+ "epoch": 0.35272440115819953,
482
+ "grad_norm": 0.15456003221800826,
483
+ "learning_rate": 0.000163492631909384,
484
+ "loss": 1.0915,
485
+ "step": 335
486
+ },
487
+ {
488
+ "epoch": 0.3579889444590682,
489
+ "grad_norm": 0.16059867173233644,
490
+ "learning_rate": 0.00016206094818274229,
491
+ "loss": 1.0969,
492
+ "step": 340
493
+ },
494
+ {
495
+ "epoch": 0.3632534877599368,
496
+ "grad_norm": 0.17415674474557066,
497
+ "learning_rate": 0.00016060826876497478,
498
+ "loss": 1.1145,
499
+ "step": 345
500
+ },
501
+ {
502
+ "epoch": 0.3685180310608055,
503
+ "grad_norm": 0.16440403677512835,
504
+ "learning_rate": 0.0001591350851085851,
505
+ "loss": 1.0683,
506
+ "step": 350
507
+ },
508
+ {
509
+ "epoch": 0.3737825743616741,
510
+ "grad_norm": 0.15901493438320982,
511
+ "learning_rate": 0.00015764189560281677,
512
+ "loss": 1.1199,
513
+ "step": 355
514
+ },
515
+ {
516
+ "epoch": 0.3790471176625428,
517
+ "grad_norm": 0.15988293404570103,
518
+ "learning_rate": 0.00015612920540504453,
519
+ "loss": 1.0709,
520
+ "step": 360
521
+ },
522
+ {
523
+ "epoch": 0.3843116609634114,
524
+ "grad_norm": 0.1616109424204681,
525
+ "learning_rate": 0.00015459752626987563,
526
+ "loss": 1.1027,
527
+ "step": 365
528
+ },
529
+ {
530
+ "epoch": 0.3895762042642801,
531
+ "grad_norm": 0.1513607201651111,
532
+ "learning_rate": 0.00015304737637601926,
533
+ "loss": 1.0956,
534
+ "step": 370
535
+ },
536
+ {
537
+ "epoch": 0.3948407475651487,
538
+ "grad_norm": 0.15452619863423803,
539
+ "learning_rate": 0.0001514792801509831,
540
+ "loss": 1.0952,
541
+ "step": 375
542
+ },
543
+ {
544
+ "epoch": 0.40010529086601737,
545
+ "grad_norm": 0.15418975657555584,
546
+ "learning_rate": 0.00014989376809365493,
547
+ "loss": 1.0934,
548
+ "step": 380
549
+ },
550
+ {
551
+ "epoch": 0.40536983416688604,
552
+ "grad_norm": 0.15158447263390024,
553
+ "learning_rate": 0.00014829137659483143,
554
+ "loss": 1.0981,
555
+ "step": 385
556
+ },
557
+ {
558
+ "epoch": 0.41063437746775466,
559
+ "grad_norm": 0.15420702474431047,
560
+ "learning_rate": 0.0001466726477557527,
561
+ "loss": 1.1013,
562
+ "step": 390
563
+ },
564
+ {
565
+ "epoch": 0.41589892076862334,
566
+ "grad_norm": 0.1513401762569788,
567
+ "learning_rate": 0.00014503812920470534,
568
+ "loss": 1.1128,
569
+ "step": 395
570
+ },
571
+ {
572
+ "epoch": 0.42116346406949196,
573
+ "grad_norm": 0.1759021276212348,
574
+ "learning_rate": 0.00014338837391175582,
575
+ "loss": 1.0793,
576
+ "step": 400
577
+ },
578
+ {
579
+ "epoch": 0.42642800737036063,
580
+ "grad_norm": 0.15639002655528358,
581
+ "learning_rate": 0.00014172394000167623,
582
+ "loss": 1.1126,
583
+ "step": 405
584
+ },
585
+ {
586
+ "epoch": 0.43169255067122925,
587
+ "grad_norm": 0.1558922751326013,
588
+ "learning_rate": 0.00014004539056512667,
589
+ "loss": 1.0864,
590
+ "step": 410
591
+ },
592
+ {
593
+ "epoch": 0.4369570939720979,
594
+ "grad_norm": 0.15449223519766864,
595
+ "learning_rate": 0.00013835329346815716,
596
+ "loss": 1.1161,
597
+ "step": 415
598
+ },
599
+ {
600
+ "epoch": 0.4422216372729666,
601
+ "grad_norm": 0.15398779214531882,
602
+ "learning_rate": 0.0001366482211600945,
603
+ "loss": 1.113,
604
+ "step": 420
605
+ },
606
+ {
607
+ "epoch": 0.4474861805738352,
608
+ "grad_norm": 0.15902962443654645,
609
+ "learning_rate": 0.000134930750479878,
610
+ "loss": 1.0783,
611
+ "step": 425
612
+ },
613
+ {
614
+ "epoch": 0.4527507238747039,
615
+ "grad_norm": 0.15614703146804315,
616
+ "learning_rate": 0.00013320146246091074,
617
+ "loss": 1.0891,
618
+ "step": 430
619
+ },
620
+ {
621
+ "epoch": 0.4580152671755725,
622
+ "grad_norm": 0.151735228198923,
623
+ "learning_rate": 0.00013146094213449148,
624
+ "loss": 1.1006,
625
+ "step": 435
626
+ },
627
+ {
628
+ "epoch": 0.4632798104764412,
629
+ "grad_norm": 0.1633743946888902,
630
+ "learning_rate": 0.00012970977833189393,
631
+ "loss": 1.0717,
632
+ "step": 440
633
+ },
634
+ {
635
+ "epoch": 0.4685443537773098,
636
+ "grad_norm": 0.16534257355481496,
637
+ "learning_rate": 0.00012794856348516095,
638
+ "loss": 1.0778,
639
+ "step": 445
640
+ },
641
+ {
642
+ "epoch": 0.47380889707817847,
643
+ "grad_norm": 0.1856142828881669,
644
+ "learning_rate": 0.00012617789342668004,
645
+ "loss": 1.0859,
646
+ "step": 450
647
+ },
648
+ {
649
+ "epoch": 0.47907344037904714,
650
+ "grad_norm": 0.15229515578033356,
651
+ "learning_rate": 0.00012439836718760886,
652
+ "loss": 1.0761,
653
+ "step": 455
654
+ },
655
+ {
656
+ "epoch": 0.48433798367991576,
657
+ "grad_norm": 0.15984985984562605,
658
+ "learning_rate": 0.00012261058679521834,
659
+ "loss": 1.0926,
660
+ "step": 460
661
+ },
662
+ {
663
+ "epoch": 0.48960252698078444,
664
+ "grad_norm": 0.14896040772758903,
665
+ "learning_rate": 0.00012081515706922227,
666
+ "loss": 1.0834,
667
+ "step": 465
668
+ },
669
+ {
670
+ "epoch": 0.49486707028165305,
671
+ "grad_norm": 0.1514924492192347,
672
+ "learning_rate": 0.00011901268541716224,
673
+ "loss": 1.0885,
674
+ "step": 470
675
+ },
676
+ {
677
+ "epoch": 0.5001316135825217,
678
+ "grad_norm": 0.1513889418015892,
679
+ "learning_rate": 0.00011720378162891708,
680
+ "loss": 1.1001,
681
+ "step": 475
682
+ },
683
+ {
684
+ "epoch": 0.5053961568833903,
685
+ "grad_norm": 0.15159825336613816,
686
+ "learning_rate": 0.0001153890576704062,
687
+ "loss": 1.1082,
688
+ "step": 480
689
+ },
690
+ {
691
+ "epoch": 0.510660700184259,
692
+ "grad_norm": 0.15427722774659086,
693
+ "learning_rate": 0.00011356912747655685,
694
+ "loss": 1.0843,
695
+ "step": 485
696
+ },
697
+ {
698
+ "epoch": 0.5159252434851277,
699
+ "grad_norm": 0.14639500931900093,
700
+ "learning_rate": 0.00011174460674360549,
701
+ "loss": 1.1058,
702
+ "step": 490
703
+ },
704
+ {
705
+ "epoch": 0.5211897867859964,
706
+ "grad_norm": 0.15320269723203808,
707
+ "learning_rate": 0.00010991611272080269,
708
+ "loss": 1.1125,
709
+ "step": 495
710
+ },
711
+ {
712
+ "epoch": 0.5264543300868649,
713
+ "grad_norm": 0.15092814943890553,
714
+ "learning_rate": 0.00010808426400159338,
715
+ "loss": 1.0898,
716
+ "step": 500
717
+ },
718
+ {
719
+ "epoch": 0.5317188733877336,
720
+ "grad_norm": 0.14712598563479434,
721
+ "learning_rate": 0.00010624968031434173,
722
+ "loss": 1.0975,
723
+ "step": 505
724
+ },
725
+ {
726
+ "epoch": 0.5369834166886023,
727
+ "grad_norm": 0.1506174008648404,
728
+ "learning_rate": 0.00010441298231267242,
729
+ "loss": 1.0789,
730
+ "step": 510
731
+ },
732
+ {
733
+ "epoch": 0.542247959989471,
734
+ "grad_norm": 0.14915164476738402,
735
+ "learning_rate": 0.00010257479136549889,
736
+ "loss": 1.088,
737
+ "step": 515
738
+ },
739
+ {
740
+ "epoch": 0.5475125032903395,
741
+ "grad_norm": 0.14933216158522156,
742
+ "learning_rate": 0.00010073572934680919,
743
+ "loss": 1.1012,
744
+ "step": 520
745
+ },
746
+ {
747
+ "epoch": 0.5527770465912082,
748
+ "grad_norm": 0.1623395783916047,
749
+ "learning_rate": 9.889641842528178e-05,
750
+ "loss": 1.0992,
751
+ "step": 525
752
+ },
753
+ {
754
+ "epoch": 0.5580415898920769,
755
+ "grad_norm": 0.15524883773019818,
756
+ "learning_rate": 9.70574808538006e-05,
757
+ "loss": 1.0558,
758
+ "step": 530
759
+ },
760
+ {
761
+ "epoch": 0.5633061331929455,
762
+ "grad_norm": 0.14879516385003932,
763
+ "learning_rate": 9.521953875894257e-05,
764
+ "loss": 1.0634,
765
+ "step": 535
766
+ },
767
+ {
768
+ "epoch": 0.5685706764938142,
769
+ "grad_norm": 0.14856407933911947,
770
+ "learning_rate": 9.338321393050719e-05,
771
+ "loss": 1.0513,
772
+ "step": 540
773
+ },
774
+ {
775
+ "epoch": 0.5738352197946828,
776
+ "grad_norm": 0.1514919636398635,
777
+ "learning_rate": 9.154912761116056e-05,
778
+ "loss": 1.0899,
779
+ "step": 545
780
+ },
781
+ {
782
+ "epoch": 0.5790997630955514,
783
+ "grad_norm": 0.15005939408454377,
784
+ "learning_rate": 8.971790028626395e-05,
785
+ "loss": 1.09,
786
+ "step": 550
787
+ },
788
+ {
789
+ "epoch": 0.5843643063964201,
790
+ "grad_norm": 0.1541140355049706,
791
+ "learning_rate": 8.789015147395919e-05,
792
+ "loss": 1.072,
793
+ "step": 555
794
+ },
795
+ {
796
+ "epoch": 0.5896288496972888,
797
+ "grad_norm": 0.14756189100480177,
798
+ "learning_rate": 8.606649951558073e-05,
799
+ "loss": 1.0548,
800
+ "step": 560
801
+ },
802
+ {
803
+ "epoch": 0.5948933929981574,
804
+ "grad_norm": 0.14468591274130843,
805
+ "learning_rate": 8.424756136646623e-05,
806
+ "loss": 1.056,
807
+ "step": 565
808
+ },
809
+ {
810
+ "epoch": 0.600157936299026,
811
+ "grad_norm": 0.1510683202100121,
812
+ "learning_rate": 8.243395238723571e-05,
813
+ "loss": 1.0999,
814
+ "step": 570
815
+ },
816
+ {
817
+ "epoch": 0.6054224795998947,
818
+ "grad_norm": 0.14942489035639112,
819
+ "learning_rate": 8.062628613561051e-05,
820
+ "loss": 1.08,
821
+ "step": 575
822
+ },
823
+ {
824
+ "epoch": 0.6106870229007634,
825
+ "grad_norm": 0.14792710995590722,
826
+ "learning_rate": 7.8825174158842e-05,
827
+ "loss": 1.0916,
828
+ "step": 580
829
+ },
830
+ {
831
+ "epoch": 0.615951566201632,
832
+ "grad_norm": 0.14543568608581728,
833
+ "learning_rate": 7.703122578682046e-05,
834
+ "loss": 1.061,
835
+ "step": 585
836
+ },
837
+ {
838
+ "epoch": 0.6212161095025006,
839
+ "grad_norm": 0.14792849899325772,
840
+ "learning_rate": 7.524504792593419e-05,
841
+ "loss": 1.1101,
842
+ "step": 590
843
+ },
844
+ {
845
+ "epoch": 0.6264806528033693,
846
+ "grad_norm": 0.14574924924348462,
847
+ "learning_rate": 7.346724485374837e-05,
848
+ "loss": 1.0687,
849
+ "step": 595
850
+ },
851
+ {
852
+ "epoch": 0.631745196104238,
853
+ "grad_norm": 0.1434166906369258,
854
+ "learning_rate": 7.169841801457347e-05,
855
+ "loss": 1.0825,
856
+ "step": 600
857
+ },
858
+ {
859
+ "epoch": 0.6370097394051066,
860
+ "grad_norm": 0.14254720323207454,
861
+ "learning_rate": 6.993916581599202e-05,
862
+ "loss": 1.0896,
863
+ "step": 605
864
+ },
865
+ {
866
+ "epoch": 0.6422742827059753,
867
+ "grad_norm": 0.14534591022474969,
868
+ "learning_rate": 6.819008342641273e-05,
869
+ "loss": 1.0805,
870
+ "step": 610
871
+ },
872
+ {
873
+ "epoch": 0.6475388260068439,
874
+ "grad_norm": 0.1471482502229213,
875
+ "learning_rate": 6.645176257372055e-05,
876
+ "loss": 1.0933,
877
+ "step": 615
878
+ },
879
+ {
880
+ "epoch": 0.6528033693077125,
881
+ "grad_norm": 0.14967562406928056,
882
+ "learning_rate": 6.472479134509052e-05,
883
+ "loss": 1.0987,
884
+ "step": 620
885
+ },
886
+ {
887
+ "epoch": 0.6580679126085812,
888
+ "grad_norm": 0.14756218985788289,
889
+ "learning_rate": 6.300975398803362e-05,
890
+ "loss": 1.0862,
891
+ "step": 625
892
+ },
893
+ {
894
+ "epoch": 0.6633324559094499,
895
+ "grad_norm": 0.14358810278632364,
896
+ "learning_rate": 6.130723071274107e-05,
897
+ "loss": 1.0736,
898
+ "step": 630
899
+ },
900
+ {
901
+ "epoch": 0.6685969992103185,
902
+ "grad_norm": 0.14508119820046267,
903
+ "learning_rate": 5.961779749579516e-05,
904
+ "loss": 1.077,
905
+ "step": 635
906
+ },
907
+ {
908
+ "epoch": 0.6738615425111871,
909
+ "grad_norm": 0.14868475648668983,
910
+ "learning_rate": 5.794202588531166e-05,
911
+ "loss": 1.0921,
912
+ "step": 640
913
+ },
914
+ {
915
+ "epoch": 0.6791260858120558,
916
+ "grad_norm": 0.14136660751737096,
917
+ "learning_rate": 5.628048280758096e-05,
918
+ "loss": 1.0967,
919
+ "step": 645
920
+ },
921
+ {
922
+ "epoch": 0.6843906291129245,
923
+ "grad_norm": 0.14429824406995242,
924
+ "learning_rate": 5.4633730375272594e-05,
925
+ "loss": 1.094,
926
+ "step": 650
927
+ },
928
+ {
929
+ "epoch": 0.6896551724137931,
930
+ "grad_norm": 0.1435583500936634,
931
+ "learning_rate": 5.300232569726804e-05,
932
+ "loss": 1.0796,
933
+ "step": 655
934
+ },
935
+ {
936
+ "epoch": 0.6949197157146617,
937
+ "grad_norm": 0.14917594264214823,
938
+ "learning_rate": 5.13868206901867e-05,
939
+ "loss": 1.0813,
940
+ "step": 660
941
+ },
942
+ {
943
+ "epoch": 0.7001842590155304,
944
+ "grad_norm": 0.14484547003342338,
945
+ "learning_rate": 4.9787761891668397e-05,
946
+ "loss": 1.0833,
947
+ "step": 665
948
+ },
949
+ {
950
+ "epoch": 0.7054488023163991,
951
+ "grad_norm": 0.14125281408090304,
952
+ "learning_rate": 4.820569027547533e-05,
953
+ "loss": 1.0813,
954
+ "step": 670
955
+ },
956
+ {
957
+ "epoch": 0.7107133456172677,
958
+ "grad_norm": 0.1408995053360923,
959
+ "learning_rate": 4.6641141068476666e-05,
960
+ "loss": 1.0752,
961
+ "step": 675
962
+ },
963
+ {
964
+ "epoch": 0.7159778889181364,
965
+ "grad_norm": 0.1414179653044325,
966
+ "learning_rate": 4.5094643569577186e-05,
967
+ "loss": 1.054,
968
+ "step": 680
969
+ },
970
+ {
971
+ "epoch": 0.721242432219005,
972
+ "grad_norm": 0.14582058548503438,
973
+ "learning_rate": 4.356672097065134e-05,
974
+ "loss": 1.1048,
975
+ "step": 685
976
+ },
977
+ {
978
+ "epoch": 0.7265069755198736,
979
+ "grad_norm": 0.14009606861616825,
980
+ "learning_rate": 4.205789017954364e-05,
981
+ "loss": 1.0683,
982
+ "step": 690
983
+ },
984
+ {
985
+ "epoch": 0.7317715188207423,
986
+ "grad_norm": 0.14586506040118713,
987
+ "learning_rate": 4.056866164519465e-05,
988
+ "loss": 1.0728,
989
+ "step": 695
990
+ },
991
+ {
992
+ "epoch": 0.737036062121611,
993
+ "grad_norm": 0.14168474565307407,
994
+ "learning_rate": 3.909953918495234e-05,
995
+ "loss": 1.0476,
996
+ "step": 700
997
+ },
998
+ {
999
+ "epoch": 0.7423006054224796,
1000
+ "grad_norm": 0.14476382479542646,
1001
+ "learning_rate": 3.7651019814126654e-05,
1002
+ "loss": 1.05,
1003
+ "step": 705
1004
+ },
1005
+ {
1006
+ "epoch": 0.7475651487233482,
1007
+ "grad_norm": 0.14528550784733454,
1008
+ "learning_rate": 3.622359357784569e-05,
1009
+ "loss": 1.0611,
1010
+ "step": 710
1011
+ },
1012
+ {
1013
+ "epoch": 0.7528296920242169,
1014
+ "grad_norm": 0.14781069746763306,
1015
+ "learning_rate": 3.481774338526954e-05,
1016
+ "loss": 1.0952,
1017
+ "step": 715
1018
+ },
1019
+ {
1020
+ "epoch": 0.7580942353250856,
1021
+ "grad_norm": 0.15618197530507127,
1022
+ "learning_rate": 3.343394484621855e-05,
1023
+ "loss": 1.0836,
1024
+ "step": 720
1025
+ },
1026
+ {
1027
+ "epoch": 0.7633587786259542,
1028
+ "grad_norm": 0.22087793925041818,
1029
+ "learning_rate": 3.207266611027069e-05,
1030
+ "loss": 1.0727,
1031
+ "step": 725
1032
+ },
1033
+ {
1034
+ "epoch": 0.7686233219268228,
1035
+ "grad_norm": 0.14674869869141435,
1036
+ "learning_rate": 3.0734367708383294e-05,
1037
+ "loss": 1.0712,
1038
+ "step": 730
1039
+ },
1040
+ {
1041
+ "epoch": 0.7738878652276915,
1042
+ "grad_norm": 0.14673826341334423,
1043
+ "learning_rate": 2.9419502397091713e-05,
1044
+ "loss": 1.0852,
1045
+ "step": 735
1046
+ },
1047
+ {
1048
+ "epoch": 0.7791524085285602,
1049
+ "grad_norm": 0.1426087824509766,
1050
+ "learning_rate": 2.812851500533843e-05,
1051
+ "loss": 1.0604,
1052
+ "step": 740
1053
+ },
1054
+ {
1055
+ "epoch": 0.7844169518294288,
1056
+ "grad_norm": 0.1446320144127932,
1057
+ "learning_rate": 2.6861842283983953e-05,
1058
+ "loss": 1.0537,
1059
+ "step": 745
1060
+ },
1061
+ {
1062
+ "epoch": 0.7896814951302974,
1063
+ "grad_norm": 0.14326111319394175,
1064
+ "learning_rate": 2.5619912758050725e-05,
1065
+ "loss": 1.0942,
1066
+ "step": 750
1067
+ },
1068
+ {
1069
+ "epoch": 0.7949460384311661,
1070
+ "grad_norm": 0.14149919988871043,
1071
+ "learning_rate": 2.4403146581749925e-05,
1072
+ "loss": 1.0578,
1073
+ "step": 755
1074
+ },
1075
+ {
1076
+ "epoch": 0.8002105817320347,
1077
+ "grad_norm": 0.14034086298796508,
1078
+ "learning_rate": 2.3211955396340002e-05,
1079
+ "loss": 1.0808,
1080
+ "step": 760
1081
+ },
1082
+ {
1083
+ "epoch": 0.8054751250329034,
1084
+ "grad_norm": 0.1433790314655123,
1085
+ "learning_rate": 2.204674219086531e-05,
1086
+ "loss": 1.0906,
1087
+ "step": 765
1088
+ },
1089
+ {
1090
+ "epoch": 0.8107396683337721,
1091
+ "grad_norm": 0.138618618401559,
1092
+ "learning_rate": 2.090790116582191e-05,
1093
+ "loss": 1.0559,
1094
+ "step": 770
1095
+ },
1096
+ {
1097
+ "epoch": 0.8160042116346407,
1098
+ "grad_norm": 0.1429827381187093,
1099
+ "learning_rate": 1.9795817599796418e-05,
1100
+ "loss": 1.0792,
1101
+ "step": 775
1102
+ },
1103
+ {
1104
+ "epoch": 0.8212687549355093,
1105
+ "grad_norm": 0.14200271718072968,
1106
+ "learning_rate": 1.871086771912348e-05,
1107
+ "loss": 1.0702,
1108
+ "step": 780
1109
+ },
1110
+ {
1111
+ "epoch": 0.826533298236378,
1112
+ "grad_norm": 0.1429932480295589,
1113
+ "learning_rate": 1.7653418570605475e-05,
1114
+ "loss": 1.0715,
1115
+ "step": 785
1116
+ },
1117
+ {
1118
+ "epoch": 0.8317978415372467,
1119
+ "grad_norm": 0.14431467515210814,
1120
+ "learning_rate": 1.6623827897337762e-05,
1121
+ "loss": 1.0713,
1122
+ "step": 790
1123
+ },
1124
+ {
1125
+ "epoch": 0.8370623848381153,
1126
+ "grad_norm": 0.15238820455432608,
1127
+ "learning_rate": 1.562244401768144e-05,
1128
+ "loss": 1.0824,
1129
+ "step": 795
1130
+ },
1131
+ {
1132
+ "epoch": 0.8423269281389839,
1133
+ "grad_norm": 0.14830242766673976,
1134
+ "learning_rate": 1.4649605707424707e-05,
1135
+ "loss": 1.0787,
1136
+ "step": 800
1137
+ },
1138
+ {
1139
+ "epoch": 0.8475914714398526,
1140
+ "grad_norm": 0.14468170557092047,
1141
+ "learning_rate": 1.3705642085172366e-05,
1142
+ "loss": 1.0737,
1143
+ "step": 805
1144
+ },
1145
+ {
1146
+ "epoch": 0.8528560147407213,
1147
+ "grad_norm": 0.14674968769736463,
1148
+ "learning_rate": 1.2790872501002472e-05,
1149
+ "loss": 1.0577,
1150
+ "step": 810
1151
+ },
1152
+ {
1153
+ "epoch": 0.8581205580415899,
1154
+ "grad_norm": 0.14311627432536864,
1155
+ "learning_rate": 1.1905606428427774e-05,
1156
+ "loss": 1.0692,
1157
+ "step": 815
1158
+ },
1159
+ {
1160
+ "epoch": 0.8633851013424585,
1161
+ "grad_norm": 0.14558376197107287,
1162
+ "learning_rate": 1.105014335969855e-05,
1163
+ "loss": 1.0934,
1164
+ "step": 820
1165
+ },
1166
+ {
1167
+ "epoch": 0.8686496446433272,
1168
+ "grad_norm": 0.14414555681497093,
1169
+ "learning_rate": 1.0224772704482033e-05,
1170
+ "loss": 1.0875,
1171
+ "step": 825
1172
+ },
1173
+ {
1174
+ "epoch": 0.8739141879441958,
1175
+ "grad_norm": 0.1399627142514978,
1176
+ "learning_rate": 9.429773691952858e-06,
1177
+ "loss": 1.082,
1178
+ "step": 830
1179
+ },
1180
+ {
1181
+ "epoch": 0.8791787312450645,
1182
+ "grad_norm": 0.1392001373823857,
1183
+ "learning_rate": 8.665415276327871e-06,
1184
+ "loss": 1.0573,
1185
+ "step": 835
1186
+ },
1187
+ {
1188
+ "epoch": 0.8844432745459332,
1189
+ "grad_norm": 0.13993969105859186,
1190
+ "learning_rate": 7.931956045876688e-06,
1191
+ "loss": 1.0448,
1192
+ "step": 840
1193
+ },
1194
+ {
1195
+ "epoch": 0.8897078178468018,
1196
+ "grad_norm": 0.16741517197447736,
1197
+ "learning_rate": 7.229644135439473e-06,
1198
+ "loss": 1.104,
1199
+ "step": 845
1200
+ },
1201
+ {
1202
+ "epoch": 0.8949723611476704,
1203
+ "grad_norm": 0.14123729142229655,
1204
+ "learning_rate": 6.558717142480919e-06,
1205
+ "loss": 1.0808,
1206
+ "step": 850
1207
+ },
1208
+ {
1209
+ "epoch": 0.9002369044485391,
1210
+ "grad_norm": 0.1424278055064695,
1211
+ "learning_rate": 5.919402046709288e-06,
1212
+ "loss": 1.0709,
1213
+ "step": 855
1214
+ },
1215
+ {
1216
+ "epoch": 0.9055014477494078,
1217
+ "grad_norm": 0.13993993967003346,
1218
+ "learning_rate": 5.311915133287415e-06,
1219
+ "loss": 1.0941,
1220
+ "step": 860
1221
+ },
1222
+ {
1223
+ "epoch": 0.9107659910502763,
1224
+ "grad_norm": 0.14557850289664284,
1225
+ "learning_rate": 4.7364619196617495e-06,
1226
+ "loss": 1.0492,
1227
+ "step": 865
1228
+ },
1229
+ {
1230
+ "epoch": 0.916030534351145,
1231
+ "grad_norm": 0.1450177459066908,
1232
+ "learning_rate": 4.193237086034351e-06,
1233
+ "loss": 1.0972,
1234
+ "step": 870
1235
+ },
1236
+ {
1237
+ "epoch": 0.9212950776520137,
1238
+ "grad_norm": 0.1570091074884799,
1239
+ "learning_rate": 3.6824244095010065e-06,
1240
+ "loss": 1.0695,
1241
+ "step": 875
1242
+ },
1243
+ {
1244
+ "epoch": 0.9265596209528824,
1245
+ "grad_norm": 0.14097561405495265,
1246
+ "learning_rate": 3.2041967018780707e-06,
1247
+ "loss": 1.0948,
1248
+ "step": 880
1249
+ },
1250
+ {
1251
+ "epoch": 0.931824164253751,
1252
+ "grad_norm": 0.1420984285291773,
1253
+ "learning_rate": 2.7587157512388718e-06,
1254
+ "loss": 1.0573,
1255
+ "step": 885
1256
+ },
1257
+ {
1258
+ "epoch": 0.9370887075546196,
1259
+ "grad_norm": 0.1545471738706476,
1260
+ "learning_rate": 2.346132267179646e-06,
1261
+ "loss": 1.0786,
1262
+ "step": 890
1263
+ },
1264
+ {
1265
+ "epoch": 0.9423532508554883,
1266
+ "grad_norm": 0.14481364480205125,
1267
+ "learning_rate": 1.9665858298333005e-06,
1268
+ "loss": 1.0939,
1269
+ "step": 895
1270
+ },
1271
+ {
1272
+ "epoch": 0.9476177941563569,
1273
+ "grad_norm": 0.1446556897144525,
1274
+ "learning_rate": 1.6202048426483651e-06,
1275
+ "loss": 1.0752,
1276
+ "step": 900
1277
+ },
1278
+ {
1279
+ "epoch": 0.9528823374572256,
1280
+ "grad_norm": 0.13840641658264988,
1281
+ "learning_rate": 1.3071064889491724e-06,
1282
+ "loss": 1.0757,
1283
+ "step": 905
1284
+ },
1285
+ {
1286
+ "epoch": 0.9581468807580943,
1287
+ "grad_norm": 0.1405867091258211,
1288
+ "learning_rate": 1.0273966922918155e-06,
1289
+ "loss": 1.0886,
1290
+ "step": 910
1291
+ },
1292
+ {
1293
+ "epoch": 0.9634114240589629,
1294
+ "grad_norm": 0.15143973079201015,
1295
+ "learning_rate": 7.81170080629412e-07,
1296
+ "loss": 1.0337,
1297
+ "step": 915
1298
+ },
1299
+ {
1300
+ "epoch": 0.9686759673598315,
1301
+ "grad_norm": 0.15113893856195346,
1302
+ "learning_rate": 5.68509954298757e-07,
1303
+ "loss": 1.099,
1304
+ "step": 920
1305
+ },
1306
+ {
1307
+ "epoch": 0.9739405106607002,
1308
+ "grad_norm": 0.1436446854214333,
1309
+ "learning_rate": 3.8948825783918784e-07,
1310
+ "loss": 1.0595,
1311
+ "step": 925
1312
+ },
1313
+ {
1314
+ "epoch": 0.9792050539615689,
1315
+ "grad_norm": 0.14373165990559605,
1316
+ "learning_rate": 2.4416555565318635e-07,
1317
+ "loss": 1.0815,
1318
+ "step": 930
1319
+ },
1320
+ {
1321
+ "epoch": 0.9844695972624374,
1322
+ "grad_norm": 0.14233020784379563,
1323
+ "learning_rate": 1.3259101151694708e-07,
1324
+ "loss": 1.0569,
1325
+ "step": 935
1326
+ },
1327
+ {
1328
+ "epoch": 0.9897341405633061,
1329
+ "grad_norm": 0.13823967108377017,
1330
+ "learning_rate": 5.480237194799287e-08,
1331
+ "loss": 1.0689,
1332
+ "step": 940
1333
+ },
1334
+ {
1335
+ "epoch": 0.9949986838641748,
1336
+ "grad_norm": 0.1431568671824589,
1337
+ "learning_rate": 1.0825953435122938e-08,
1338
+ "loss": 1.0709,
1339
+ "step": 945
1340
+ },
1341
+ {
1342
+ "epoch": 0.9992103185048697,
1343
+ "eval_loss": 1.07915198802948,
1344
+ "eval_runtime": 3821.2872,
1345
+ "eval_samples_per_second": 3.522,
1346
+ "eval_steps_per_second": 0.881,
1347
+ "step": 949
1348
+ },
1349
+ {
1350
+ "epoch": 0.9992103185048697,
1351
+ "step": 949,
1352
+ "total_flos": 1959448100732928.0,
1353
+ "train_loss": 1.0930153100081064,
1354
+ "train_runtime": 22340.3866,
1355
+ "train_samples_per_second": 2.72,
1356
+ "train_steps_per_second": 0.042
1357
+ }
1358
+ ],
1359
+ "logging_steps": 5,
1360
+ "max_steps": 949,
1361
+ "num_input_tokens_seen": 0,
1362
+ "num_train_epochs": 1,
1363
+ "save_steps": 100,
1364
+ "stateful_callbacks": {
1365
+ "TrainerControl": {
1366
+ "args": {
1367
+ "should_epoch_stop": false,
1368
+ "should_evaluate": false,
1369
+ "should_log": false,
1370
+ "should_save": true,
1371
+ "should_training_stop": true
1372
+ },
1373
+ "attributes": {}
1374
+ }
1375
+ },
1376
+ "total_flos": 1959448100732928.0,
1377
+ "train_batch_size": 4,
1378
+ "trial_name": null,
1379
+ "trial_params": null
1380
+ }