AleBurzio commited on
Commit
6181dda
1 Parent(s): 0456239

End of training

Browse files
Files changed (4) hide show
  1. all_results.json +15 -0
  2. eval_results.json +10 -0
  3. train_results.json +8 -0
  4. trainer_state.json +2329 -0
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.6594206731193033,
4
+ "eval_loss": 4.810673236846924,
5
+ "eval_runtime": 4.2065,
6
+ "eval_samples": 111,
7
+ "eval_samples_per_second": 26.388,
8
+ "eval_steps_per_second": 13.313,
9
+ "perplexity": 122.81427278462422,
10
+ "train_loss": 0.47322619671273664,
11
+ "train_runtime": 757.4177,
12
+ "train_samples": 383,
13
+ "train_samples_per_second": 5.057,
14
+ "train_steps_per_second": 2.535
15
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "eval_accuracy": 0.6594206731193033,
4
+ "eval_loss": 4.810673236846924,
5
+ "eval_runtime": 4.2065,
6
+ "eval_samples": 111,
7
+ "eval_samples_per_second": 26.388,
8
+ "eval_steps_per_second": 13.313,
9
+ "perplexity": 122.81427278462422
10
+ }
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 10.0,
3
+ "train_loss": 0.47322619671273664,
4
+ "train_runtime": 757.4177,
5
+ "train_samples": 383,
6
+ "train_samples_per_second": 5.057,
7
+ "train_steps_per_second": 2.535
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,2329 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 10.0,
5
+ "global_step": 1920,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.03,
12
+ "learning_rate": 1.25e-05,
13
+ "loss": 2.1623,
14
+ "step": 5
15
+ },
16
+ {
17
+ "epoch": 0.05,
18
+ "learning_rate": 2.5e-05,
19
+ "loss": 2.0298,
20
+ "step": 10
21
+ },
22
+ {
23
+ "epoch": 0.08,
24
+ "learning_rate": 3.7500000000000003e-05,
25
+ "loss": 1.9422,
26
+ "step": 15
27
+ },
28
+ {
29
+ "epoch": 0.1,
30
+ "learning_rate": 5e-05,
31
+ "loss": 2.0168,
32
+ "step": 20
33
+ },
34
+ {
35
+ "epoch": 0.13,
36
+ "learning_rate": 4.999914564160437e-05,
37
+ "loss": 1.9103,
38
+ "step": 25
39
+ },
40
+ {
41
+ "epoch": 0.16,
42
+ "learning_rate": 4.9996582624811725e-05,
43
+ "loss": 2.1249,
44
+ "step": 30
45
+ },
46
+ {
47
+ "epoch": 0.18,
48
+ "learning_rate": 4.9992311124800875e-05,
49
+ "loss": 2.002,
50
+ "step": 35
51
+ },
52
+ {
53
+ "epoch": 0.21,
54
+ "learning_rate": 4.9986331433523156e-05,
55
+ "loss": 2.004,
56
+ "step": 40
57
+ },
58
+ {
59
+ "epoch": 0.23,
60
+ "learning_rate": 4.997864395968252e-05,
61
+ "loss": 1.9562,
62
+ "step": 45
63
+ },
64
+ {
65
+ "epoch": 0.26,
66
+ "learning_rate": 4.996924922870762e-05,
67
+ "loss": 2.078,
68
+ "step": 50
69
+ },
70
+ {
71
+ "epoch": 0.29,
72
+ "learning_rate": 4.995814788271582e-05,
73
+ "loss": 1.9552,
74
+ "step": 55
75
+ },
76
+ {
77
+ "epoch": 0.31,
78
+ "learning_rate": 4.994534068046937e-05,
79
+ "loss": 2.033,
80
+ "step": 60
81
+ },
82
+ {
83
+ "epoch": 0.34,
84
+ "learning_rate": 4.9930828497323526e-05,
85
+ "loss": 1.9229,
86
+ "step": 65
87
+ },
88
+ {
89
+ "epoch": 0.36,
90
+ "learning_rate": 4.991461232516675e-05,
91
+ "loss": 1.8923,
92
+ "step": 70
93
+ },
94
+ {
95
+ "epoch": 0.39,
96
+ "learning_rate": 4.9896693272352846e-05,
97
+ "loss": 2.0122,
98
+ "step": 75
99
+ },
100
+ {
101
+ "epoch": 0.42,
102
+ "learning_rate": 4.9877072563625285e-05,
103
+ "loss": 1.8244,
104
+ "step": 80
105
+ },
106
+ {
107
+ "epoch": 0.44,
108
+ "learning_rate": 4.985575154003345e-05,
109
+ "loss": 1.976,
110
+ "step": 85
111
+ },
112
+ {
113
+ "epoch": 0.47,
114
+ "learning_rate": 4.9832731658840956e-05,
115
+ "loss": 1.8806,
116
+ "step": 90
117
+ },
118
+ {
119
+ "epoch": 0.49,
120
+ "learning_rate": 4.980801449342613e-05,
121
+ "loss": 1.7659,
122
+ "step": 95
123
+ },
124
+ {
125
+ "epoch": 0.52,
126
+ "learning_rate": 4.978160173317438e-05,
127
+ "loss": 1.8351,
128
+ "step": 100
129
+ },
130
+ {
131
+ "epoch": 0.55,
132
+ "learning_rate": 4.9753495183362796e-05,
133
+ "loss": 1.7661,
134
+ "step": 105
135
+ },
136
+ {
137
+ "epoch": 0.57,
138
+ "learning_rate": 4.972369676503672e-05,
139
+ "loss": 1.9104,
140
+ "step": 110
141
+ },
142
+ {
143
+ "epoch": 0.6,
144
+ "learning_rate": 4.9692208514878444e-05,
145
+ "loss": 1.9618,
146
+ "step": 115
147
+ },
148
+ {
149
+ "epoch": 0.62,
150
+ "learning_rate": 4.965903258506806e-05,
151
+ "loss": 1.874,
152
+ "step": 120
153
+ },
154
+ {
155
+ "epoch": 0.65,
156
+ "learning_rate": 4.96241712431363e-05,
157
+ "loss": 1.7711,
158
+ "step": 125
159
+ },
160
+ {
161
+ "epoch": 0.68,
162
+ "learning_rate": 4.958762687180956e-05,
163
+ "loss": 1.7548,
164
+ "step": 130
165
+ },
166
+ {
167
+ "epoch": 0.7,
168
+ "learning_rate": 4.95494019688471e-05,
169
+ "loss": 1.9196,
170
+ "step": 135
171
+ },
172
+ {
173
+ "epoch": 0.73,
174
+ "learning_rate": 4.9509499146870236e-05,
175
+ "loss": 1.8337,
176
+ "step": 140
177
+ },
178
+ {
179
+ "epoch": 0.76,
180
+ "learning_rate": 4.946792113318386e-05,
181
+ "loss": 2.0006,
182
+ "step": 145
183
+ },
184
+ {
185
+ "epoch": 0.78,
186
+ "learning_rate": 4.9424670769589984e-05,
187
+ "loss": 1.8703,
188
+ "step": 150
189
+ },
190
+ {
191
+ "epoch": 0.81,
192
+ "learning_rate": 4.93797510121935e-05,
193
+ "loss": 1.8343,
194
+ "step": 155
195
+ },
196
+ {
197
+ "epoch": 0.83,
198
+ "learning_rate": 4.933316493120015e-05,
199
+ "loss": 1.8636,
200
+ "step": 160
201
+ },
202
+ {
203
+ "epoch": 0.86,
204
+ "learning_rate": 4.9284915710706695e-05,
205
+ "loss": 1.9264,
206
+ "step": 165
207
+ },
208
+ {
209
+ "epoch": 0.89,
210
+ "learning_rate": 4.923500664848326e-05,
211
+ "loss": 1.9231,
212
+ "step": 170
213
+ },
214
+ {
215
+ "epoch": 0.91,
216
+ "learning_rate": 4.918344115574796e-05,
217
+ "loss": 1.6458,
218
+ "step": 175
219
+ },
220
+ {
221
+ "epoch": 0.94,
222
+ "learning_rate": 4.913022275693372e-05,
223
+ "loss": 1.8824,
224
+ "step": 180
225
+ },
226
+ {
227
+ "epoch": 0.96,
228
+ "learning_rate": 4.907535508944741e-05,
229
+ "loss": 1.7897,
230
+ "step": 185
231
+ },
232
+ {
233
+ "epoch": 0.99,
234
+ "learning_rate": 4.901884190342121e-05,
235
+ "loss": 1.8587,
236
+ "step": 190
237
+ },
238
+ {
239
+ "epoch": 1.02,
240
+ "learning_rate": 4.8960687061456324e-05,
241
+ "loss": 1.4371,
242
+ "step": 195
243
+ },
244
+ {
245
+ "epoch": 1.04,
246
+ "learning_rate": 4.8900894538358944e-05,
247
+ "loss": 1.3405,
248
+ "step": 200
249
+ },
250
+ {
251
+ "epoch": 1.07,
252
+ "learning_rate": 4.8839468420868606e-05,
253
+ "loss": 1.1991,
254
+ "step": 205
255
+ },
256
+ {
257
+ "epoch": 1.09,
258
+ "learning_rate": 4.877641290737884e-05,
259
+ "loss": 1.2276,
260
+ "step": 210
261
+ },
262
+ {
263
+ "epoch": 1.12,
264
+ "learning_rate": 4.871173230765024e-05,
265
+ "loss": 1.3785,
266
+ "step": 215
267
+ },
268
+ {
269
+ "epoch": 1.15,
270
+ "learning_rate": 4.864543104251587e-05,
271
+ "loss": 1.2454,
272
+ "step": 220
273
+ },
274
+ {
275
+ "epoch": 1.17,
276
+ "learning_rate": 4.857751364357913e-05,
277
+ "loss": 1.1988,
278
+ "step": 225
279
+ },
280
+ {
281
+ "epoch": 1.2,
282
+ "learning_rate": 4.850798475290403e-05,
283
+ "loss": 1.2279,
284
+ "step": 230
285
+ },
286
+ {
287
+ "epoch": 1.22,
288
+ "learning_rate": 4.8436849122697883e-05,
289
+ "loss": 1.2478,
290
+ "step": 235
291
+ },
292
+ {
293
+ "epoch": 1.25,
294
+ "learning_rate": 4.8364111614986527e-05,
295
+ "loss": 1.3029,
296
+ "step": 240
297
+ },
298
+ {
299
+ "epoch": 1.28,
300
+ "learning_rate": 4.8289777201281974e-05,
301
+ "loss": 1.1754,
302
+ "step": 245
303
+ },
304
+ {
305
+ "epoch": 1.3,
306
+ "learning_rate": 4.821385096224268e-05,
307
+ "loss": 1.2921,
308
+ "step": 250
309
+ },
310
+ {
311
+ "epoch": 1.33,
312
+ "learning_rate": 4.8136338087326216e-05,
313
+ "loss": 1.3902,
314
+ "step": 255
315
+ },
316
+ {
317
+ "epoch": 1.35,
318
+ "learning_rate": 4.805724387443462e-05,
319
+ "loss": 1.3123,
320
+ "step": 260
321
+ },
322
+ {
323
+ "epoch": 1.38,
324
+ "learning_rate": 4.797657372955228e-05,
325
+ "loss": 1.2955,
326
+ "step": 265
327
+ },
328
+ {
329
+ "epoch": 1.41,
330
+ "learning_rate": 4.789433316637644e-05,
331
+ "loss": 1.3642,
332
+ "step": 270
333
+ },
334
+ {
335
+ "epoch": 1.43,
336
+ "learning_rate": 4.781052780594034e-05,
337
+ "loss": 1.3868,
338
+ "step": 275
339
+ },
340
+ {
341
+ "epoch": 1.46,
342
+ "learning_rate": 4.7725163376229064e-05,
343
+ "loss": 1.2758,
344
+ "step": 280
345
+ },
346
+ {
347
+ "epoch": 1.48,
348
+ "learning_rate": 4.763824571178798e-05,
349
+ "loss": 1.327,
350
+ "step": 285
351
+ },
352
+ {
353
+ "epoch": 1.51,
354
+ "learning_rate": 4.754978075332398e-05,
355
+ "loss": 1.1598,
356
+ "step": 290
357
+ },
358
+ {
359
+ "epoch": 1.54,
360
+ "learning_rate": 4.7459774547299475e-05,
361
+ "loss": 1.2948,
362
+ "step": 295
363
+ },
364
+ {
365
+ "epoch": 1.56,
366
+ "learning_rate": 4.736823324551909e-05,
367
+ "loss": 1.2907,
368
+ "step": 300
369
+ },
370
+ {
371
+ "epoch": 1.59,
372
+ "learning_rate": 4.72751631047092e-05,
373
+ "loss": 1.2316,
374
+ "step": 305
375
+ },
376
+ {
377
+ "epoch": 1.61,
378
+ "learning_rate": 4.71805704860903e-05,
379
+ "loss": 1.3136,
380
+ "step": 310
381
+ },
382
+ {
383
+ "epoch": 1.64,
384
+ "learning_rate": 4.708446185494222e-05,
385
+ "loss": 1.3186,
386
+ "step": 315
387
+ },
388
+ {
389
+ "epoch": 1.67,
390
+ "learning_rate": 4.698684378016222e-05,
391
+ "loss": 1.4358,
392
+ "step": 320
393
+ },
394
+ {
395
+ "epoch": 1.69,
396
+ "learning_rate": 4.6887722933816076e-05,
397
+ "loss": 1.2789,
398
+ "step": 325
399
+ },
400
+ {
401
+ "epoch": 1.72,
402
+ "learning_rate": 4.678710609068193e-05,
403
+ "loss": 1.3904,
404
+ "step": 330
405
+ },
406
+ {
407
+ "epoch": 1.74,
408
+ "learning_rate": 4.668500012778738e-05,
409
+ "loss": 1.3282,
410
+ "step": 335
411
+ },
412
+ {
413
+ "epoch": 1.77,
414
+ "learning_rate": 4.6581412023939354e-05,
415
+ "loss": 1.1926,
416
+ "step": 340
417
+ },
418
+ {
419
+ "epoch": 1.8,
420
+ "learning_rate": 4.6476348859247134e-05,
421
+ "loss": 1.3098,
422
+ "step": 345
423
+ },
424
+ {
425
+ "epoch": 1.82,
426
+ "learning_rate": 4.6369817814638475e-05,
427
+ "loss": 1.1817,
428
+ "step": 350
429
+ },
430
+ {
431
+ "epoch": 1.85,
432
+ "learning_rate": 4.6261826171368774e-05,
433
+ "loss": 1.1436,
434
+ "step": 355
435
+ },
436
+ {
437
+ "epoch": 1.88,
438
+ "learning_rate": 4.6152381310523387e-05,
439
+ "loss": 1.1436,
440
+ "step": 360
441
+ },
442
+ {
443
+ "epoch": 1.9,
444
+ "learning_rate": 4.604149071251318e-05,
445
+ "loss": 1.2308,
446
+ "step": 365
447
+ },
448
+ {
449
+ "epoch": 1.93,
450
+ "learning_rate": 4.592916195656322e-05,
451
+ "loss": 1.2855,
452
+ "step": 370
453
+ },
454
+ {
455
+ "epoch": 1.95,
456
+ "learning_rate": 4.581540272019476e-05,
457
+ "loss": 1.2501,
458
+ "step": 375
459
+ },
460
+ {
461
+ "epoch": 1.98,
462
+ "learning_rate": 4.5700220778700504e-05,
463
+ "loss": 1.3435,
464
+ "step": 380
465
+ },
466
+ {
467
+ "epoch": 2.01,
468
+ "learning_rate": 4.5583624004613145e-05,
469
+ "loss": 1.2623,
470
+ "step": 385
471
+ },
472
+ {
473
+ "epoch": 2.03,
474
+ "learning_rate": 4.546562036716732e-05,
475
+ "loss": 0.6914,
476
+ "step": 390
477
+ },
478
+ {
479
+ "epoch": 2.06,
480
+ "learning_rate": 4.534621793175487e-05,
481
+ "loss": 0.6295,
482
+ "step": 395
483
+ },
484
+ {
485
+ "epoch": 2.08,
486
+ "learning_rate": 4.522542485937369e-05,
487
+ "loss": 0.7132,
488
+ "step": 400
489
+ },
490
+ {
491
+ "epoch": 2.11,
492
+ "learning_rate": 4.510324940606979e-05,
493
+ "loss": 0.6092,
494
+ "step": 405
495
+ },
496
+ {
497
+ "epoch": 2.14,
498
+ "learning_rate": 4.497969992237312e-05,
499
+ "loss": 0.6314,
500
+ "step": 410
501
+ },
502
+ {
503
+ "epoch": 2.16,
504
+ "learning_rate": 4.4854784852726776e-05,
505
+ "loss": 0.7211,
506
+ "step": 415
507
+ },
508
+ {
509
+ "epoch": 2.19,
510
+ "learning_rate": 4.4728512734909844e-05,
511
+ "loss": 0.6992,
512
+ "step": 420
513
+ },
514
+ {
515
+ "epoch": 2.21,
516
+ "learning_rate": 4.460089219945382e-05,
517
+ "loss": 0.7503,
518
+ "step": 425
519
+ },
520
+ {
521
+ "epoch": 2.24,
522
+ "learning_rate": 4.4471931969052816e-05,
523
+ "loss": 0.6703,
524
+ "step": 430
525
+ },
526
+ {
527
+ "epoch": 2.27,
528
+ "learning_rate": 4.434164085796724e-05,
529
+ "loss": 0.6396,
530
+ "step": 435
531
+ },
532
+ {
533
+ "epoch": 2.29,
534
+ "learning_rate": 4.421002777142148e-05,
535
+ "loss": 0.6972,
536
+ "step": 440
537
+ },
538
+ {
539
+ "epoch": 2.32,
540
+ "learning_rate": 4.4077101704995166e-05,
541
+ "loss": 0.6558,
542
+ "step": 445
543
+ },
544
+ {
545
+ "epoch": 2.34,
546
+ "learning_rate": 4.3942871744008374e-05,
547
+ "loss": 0.7016,
548
+ "step": 450
549
+ },
550
+ {
551
+ "epoch": 2.37,
552
+ "learning_rate": 4.3807347062900624e-05,
553
+ "loss": 0.6722,
554
+ "step": 455
555
+ },
556
+ {
557
+ "epoch": 2.4,
558
+ "learning_rate": 4.367053692460385e-05,
559
+ "loss": 0.7364,
560
+ "step": 460
561
+ },
562
+ {
563
+ "epoch": 2.42,
564
+ "learning_rate": 4.3532450679909274e-05,
565
+ "loss": 0.6927,
566
+ "step": 465
567
+ },
568
+ {
569
+ "epoch": 2.45,
570
+ "learning_rate": 4.3393097766828293e-05,
571
+ "loss": 0.6704,
572
+ "step": 470
573
+ },
574
+ {
575
+ "epoch": 2.47,
576
+ "learning_rate": 4.325248770994741e-05,
577
+ "loss": 0.6888,
578
+ "step": 475
579
+ },
580
+ {
581
+ "epoch": 2.5,
582
+ "learning_rate": 4.311063011977723e-05,
583
+ "loss": 0.6381,
584
+ "step": 480
585
+ },
586
+ {
587
+ "epoch": 2.53,
588
+ "learning_rate": 4.29675346920956e-05,
589
+ "loss": 0.7126,
590
+ "step": 485
591
+ },
592
+ {
593
+ "epoch": 2.55,
594
+ "learning_rate": 4.282321120728493e-05,
595
+ "loss": 0.7115,
596
+ "step": 490
597
+ },
598
+ {
599
+ "epoch": 2.58,
600
+ "learning_rate": 4.267766952966369e-05,
601
+ "loss": 0.7172,
602
+ "step": 495
603
+ },
604
+ {
605
+ "epoch": 2.6,
606
+ "learning_rate": 4.2530919606812216e-05,
607
+ "loss": 0.6488,
608
+ "step": 500
609
+ },
610
+ {
611
+ "epoch": 2.63,
612
+ "learning_rate": 4.2382971468892806e-05,
613
+ "loss": 0.6716,
614
+ "step": 505
615
+ },
616
+ {
617
+ "epoch": 2.66,
618
+ "learning_rate": 4.223383522796415e-05,
619
+ "loss": 0.6279,
620
+ "step": 510
621
+ },
622
+ {
623
+ "epoch": 2.68,
624
+ "learning_rate": 4.2083521077290213e-05,
625
+ "loss": 0.6189,
626
+ "step": 515
627
+ },
628
+ {
629
+ "epoch": 2.71,
630
+ "learning_rate": 4.193203929064353e-05,
631
+ "loss": 0.671,
632
+ "step": 520
633
+ },
634
+ {
635
+ "epoch": 2.73,
636
+ "learning_rate": 4.177940022160299e-05,
637
+ "loss": 0.7658,
638
+ "step": 525
639
+ },
640
+ {
641
+ "epoch": 2.76,
642
+ "learning_rate": 4.16256143028462e-05,
643
+ "loss": 0.673,
644
+ "step": 530
645
+ },
646
+ {
647
+ "epoch": 2.79,
648
+ "learning_rate": 4.147069204543645e-05,
649
+ "loss": 0.7758,
650
+ "step": 535
651
+ },
652
+ {
653
+ "epoch": 2.81,
654
+ "learning_rate": 4.131464403810422e-05,
655
+ "loss": 0.6823,
656
+ "step": 540
657
+ },
658
+ {
659
+ "epoch": 2.84,
660
+ "learning_rate": 4.115748094652352e-05,
661
+ "loss": 0.7035,
662
+ "step": 545
663
+ },
664
+ {
665
+ "epoch": 2.86,
666
+ "learning_rate": 4.099921351258292e-05,
667
+ "loss": 0.6478,
668
+ "step": 550
669
+ },
670
+ {
671
+ "epoch": 2.89,
672
+ "learning_rate": 4.0839852553651265e-05,
673
+ "loss": 0.8168,
674
+ "step": 555
675
+ },
676
+ {
677
+ "epoch": 2.92,
678
+ "learning_rate": 4.067940896183843e-05,
679
+ "loss": 0.7212,
680
+ "step": 560
681
+ },
682
+ {
683
+ "epoch": 2.94,
684
+ "learning_rate": 4.051789370325078e-05,
685
+ "loss": 0.6609,
686
+ "step": 565
687
+ },
688
+ {
689
+ "epoch": 2.97,
690
+ "learning_rate": 4.03553178172417e-05,
691
+ "loss": 0.7815,
692
+ "step": 570
693
+ },
694
+ {
695
+ "epoch": 2.99,
696
+ "learning_rate": 4.019169241565703e-05,
697
+ "loss": 0.6535,
698
+ "step": 575
699
+ },
700
+ {
701
+ "epoch": 3.02,
702
+ "learning_rate": 4.002702868207563e-05,
703
+ "loss": 0.4454,
704
+ "step": 580
705
+ },
706
+ {
707
+ "epoch": 3.05,
708
+ "learning_rate": 3.9861337871044954e-05,
709
+ "loss": 0.3374,
710
+ "step": 585
711
+ },
712
+ {
713
+ "epoch": 3.07,
714
+ "learning_rate": 3.969463130731183e-05,
715
+ "loss": 0.3974,
716
+ "step": 590
717
+ },
718
+ {
719
+ "epoch": 3.1,
720
+ "learning_rate": 3.952692038504846e-05,
721
+ "loss": 0.3395,
722
+ "step": 595
723
+ },
724
+ {
725
+ "epoch": 3.12,
726
+ "learning_rate": 3.935821656707359e-05,
727
+ "loss": 0.3572,
728
+ "step": 600
729
+ },
730
+ {
731
+ "epoch": 3.15,
732
+ "learning_rate": 3.9188531384069096e-05,
733
+ "loss": 0.3072,
734
+ "step": 605
735
+ },
736
+ {
737
+ "epoch": 3.18,
738
+ "learning_rate": 3.901787643379182e-05,
739
+ "loss": 0.3527,
740
+ "step": 610
741
+ },
742
+ {
743
+ "epoch": 3.2,
744
+ "learning_rate": 3.884626338028094e-05,
745
+ "loss": 0.338,
746
+ "step": 615
747
+ },
748
+ {
749
+ "epoch": 3.23,
750
+ "learning_rate": 3.867370395306068e-05,
751
+ "loss": 0.357,
752
+ "step": 620
753
+ },
754
+ {
755
+ "epoch": 3.26,
756
+ "learning_rate": 3.850020994633868e-05,
757
+ "loss": 0.3689,
758
+ "step": 625
759
+ },
760
+ {
761
+ "epoch": 3.28,
762
+ "learning_rate": 3.832579321819985e-05,
763
+ "loss": 0.3469,
764
+ "step": 630
765
+ },
766
+ {
767
+ "epoch": 3.31,
768
+ "learning_rate": 3.8150465689795854e-05,
769
+ "loss": 0.3679,
770
+ "step": 635
771
+ },
772
+ {
773
+ "epoch": 3.33,
774
+ "learning_rate": 3.797423934453038e-05,
775
+ "loss": 0.3651,
776
+ "step": 640
777
+ },
778
+ {
779
+ "epoch": 3.36,
780
+ "learning_rate": 3.779712622724003e-05,
781
+ "loss": 0.3828,
782
+ "step": 645
783
+ },
784
+ {
785
+ "epoch": 3.39,
786
+ "learning_rate": 3.76191384433711e-05,
787
+ "loss": 0.3727,
788
+ "step": 650
789
+ },
790
+ {
791
+ "epoch": 3.41,
792
+ "learning_rate": 3.7440288158152187e-05,
793
+ "loss": 0.3554,
794
+ "step": 655
795
+ },
796
+ {
797
+ "epoch": 3.44,
798
+ "learning_rate": 3.726058759576271e-05,
799
+ "loss": 0.3658,
800
+ "step": 660
801
+ },
802
+ {
803
+ "epoch": 3.46,
804
+ "learning_rate": 3.708004903849741e-05,
805
+ "loss": 0.3765,
806
+ "step": 665
807
+ },
808
+ {
809
+ "epoch": 3.49,
810
+ "learning_rate": 3.689868482592684e-05,
811
+ "loss": 0.3733,
812
+ "step": 670
813
+ },
814
+ {
815
+ "epoch": 3.52,
816
+ "learning_rate": 3.671650735405404e-05,
817
+ "loss": 0.4086,
818
+ "step": 675
819
+ },
820
+ {
821
+ "epoch": 3.54,
822
+ "learning_rate": 3.65335290744672e-05,
823
+ "loss": 0.3753,
824
+ "step": 680
825
+ },
826
+ {
827
+ "epoch": 3.57,
828
+ "learning_rate": 3.634976249348867e-05,
829
+ "loss": 0.3704,
830
+ "step": 685
831
+ },
832
+ {
833
+ "epoch": 3.59,
834
+ "learning_rate": 3.616522017132017e-05,
835
+ "loss": 0.3939,
836
+ "step": 690
837
+ },
838
+ {
839
+ "epoch": 3.62,
840
+ "learning_rate": 3.597991472118426e-05,
841
+ "loss": 0.3691,
842
+ "step": 695
843
+ },
844
+ {
845
+ "epoch": 3.65,
846
+ "learning_rate": 3.579385880846232e-05,
847
+ "loss": 0.3804,
848
+ "step": 700
849
+ },
850
+ {
851
+ "epoch": 3.67,
852
+ "learning_rate": 3.5607065149828843e-05,
853
+ "loss": 0.385,
854
+ "step": 705
855
+ },
856
+ {
857
+ "epoch": 3.7,
858
+ "learning_rate": 3.5419546512382266e-05,
859
+ "loss": 0.3773,
860
+ "step": 710
861
+ },
862
+ {
863
+ "epoch": 3.72,
864
+ "learning_rate": 3.523131571277235e-05,
865
+ "loss": 0.402,
866
+ "step": 715
867
+ },
868
+ {
869
+ "epoch": 3.75,
870
+ "learning_rate": 3.504238561632424e-05,
871
+ "loss": 0.3561,
872
+ "step": 720
873
+ },
874
+ {
875
+ "epoch": 3.78,
876
+ "learning_rate": 3.485276913615905e-05,
877
+ "loss": 0.3542,
878
+ "step": 725
879
+ },
880
+ {
881
+ "epoch": 3.8,
882
+ "learning_rate": 3.4662479232311306e-05,
883
+ "loss": 0.354,
884
+ "step": 730
885
+ },
886
+ {
887
+ "epoch": 3.83,
888
+ "learning_rate": 3.447152891084319e-05,
889
+ "loss": 0.4085,
890
+ "step": 735
891
+ },
892
+ {
893
+ "epoch": 3.85,
894
+ "learning_rate": 3.427993122295552e-05,
895
+ "loss": 0.3853,
896
+ "step": 740
897
+ },
898
+ {
899
+ "epoch": 3.88,
900
+ "learning_rate": 3.4087699264095745e-05,
901
+ "loss": 0.3909,
902
+ "step": 745
903
+ },
904
+ {
905
+ "epoch": 3.91,
906
+ "learning_rate": 3.389484617306292e-05,
907
+ "loss": 0.3882,
908
+ "step": 750
909
+ },
910
+ {
911
+ "epoch": 3.93,
912
+ "learning_rate": 3.3701385131109616e-05,
913
+ "loss": 0.3917,
914
+ "step": 755
915
+ },
916
+ {
917
+ "epoch": 3.96,
918
+ "learning_rate": 3.350732936104108e-05,
919
+ "loss": 0.3825,
920
+ "step": 760
921
+ },
922
+ {
923
+ "epoch": 3.98,
924
+ "learning_rate": 3.3312692126311425e-05,
925
+ "loss": 0.379,
926
+ "step": 765
927
+ },
928
+ {
929
+ "epoch": 4.01,
930
+ "learning_rate": 3.311748673011709e-05,
931
+ "loss": 0.3357,
932
+ "step": 770
933
+ },
934
+ {
935
+ "epoch": 4.04,
936
+ "learning_rate": 3.2921726514487614e-05,
937
+ "loss": 0.2284,
938
+ "step": 775
939
+ },
940
+ {
941
+ "epoch": 4.06,
942
+ "learning_rate": 3.272542485937369e-05,
943
+ "loss": 0.2536,
944
+ "step": 780
945
+ },
946
+ {
947
+ "epoch": 4.09,
948
+ "learning_rate": 3.252859518173269e-05,
949
+ "loss": 0.2258,
950
+ "step": 785
951
+ },
952
+ {
953
+ "epoch": 4.11,
954
+ "learning_rate": 3.2331250934611624e-05,
955
+ "loss": 0.2522,
956
+ "step": 790
957
+ },
958
+ {
959
+ "epoch": 4.14,
960
+ "learning_rate": 3.213340560622763e-05,
961
+ "loss": 0.2166,
962
+ "step": 795
963
+ },
964
+ {
965
+ "epoch": 4.17,
966
+ "learning_rate": 3.1935072719046115e-05,
967
+ "loss": 0.2441,
968
+ "step": 800
969
+ },
970
+ {
971
+ "epoch": 4.19,
972
+ "learning_rate": 3.173626582885645e-05,
973
+ "loss": 0.2381,
974
+ "step": 805
975
+ },
976
+ {
977
+ "epoch": 4.22,
978
+ "learning_rate": 3.1536998523845494e-05,
979
+ "loss": 0.2642,
980
+ "step": 810
981
+ },
982
+ {
983
+ "epoch": 4.24,
984
+ "learning_rate": 3.133728442366885e-05,
985
+ "loss": 0.2315,
986
+ "step": 815
987
+ },
988
+ {
989
+ "epoch": 4.27,
990
+ "learning_rate": 3.1137137178519985e-05,
991
+ "loss": 0.2342,
992
+ "step": 820
993
+ },
994
+ {
995
+ "epoch": 4.3,
996
+ "learning_rate": 3.093657046819722e-05,
997
+ "loss": 0.2378,
998
+ "step": 825
999
+ },
1000
+ {
1001
+ "epoch": 4.32,
1002
+ "learning_rate": 3.073559800116879e-05,
1003
+ "loss": 0.2383,
1004
+ "step": 830
1005
+ },
1006
+ {
1007
+ "epoch": 4.35,
1008
+ "learning_rate": 3.053423351363586e-05,
1009
+ "loss": 0.2597,
1010
+ "step": 835
1011
+ },
1012
+ {
1013
+ "epoch": 4.38,
1014
+ "learning_rate": 3.0332490768593675e-05,
1015
+ "loss": 0.2262,
1016
+ "step": 840
1017
+ },
1018
+ {
1019
+ "epoch": 4.4,
1020
+ "learning_rate": 3.0130383554890856e-05,
1021
+ "loss": 0.2244,
1022
+ "step": 845
1023
+ },
1024
+ {
1025
+ "epoch": 4.43,
1026
+ "learning_rate": 2.9927925686287006e-05,
1027
+ "loss": 0.2385,
1028
+ "step": 850
1029
+ },
1030
+ {
1031
+ "epoch": 4.45,
1032
+ "learning_rate": 2.972513100050851e-05,
1033
+ "loss": 0.2521,
1034
+ "step": 855
1035
+ },
1036
+ {
1037
+ "epoch": 4.48,
1038
+ "learning_rate": 2.952201335830275e-05,
1039
+ "loss": 0.2357,
1040
+ "step": 860
1041
+ },
1042
+ {
1043
+ "epoch": 4.51,
1044
+ "learning_rate": 2.9318586642490763e-05,
1045
+ "loss": 0.249,
1046
+ "step": 865
1047
+ },
1048
+ {
1049
+ "epoch": 4.53,
1050
+ "learning_rate": 2.9114864757018352e-05,
1051
+ "loss": 0.2807,
1052
+ "step": 870
1053
+ },
1054
+ {
1055
+ "epoch": 4.56,
1056
+ "learning_rate": 2.8910861626005776e-05,
1057
+ "loss": 0.2273,
1058
+ "step": 875
1059
+ },
1060
+ {
1061
+ "epoch": 4.58,
1062
+ "learning_rate": 2.870659119279605e-05,
1063
+ "loss": 0.253,
1064
+ "step": 880
1065
+ },
1066
+ {
1067
+ "epoch": 4.61,
1068
+ "learning_rate": 2.850206741900195e-05,
1069
+ "loss": 0.2429,
1070
+ "step": 885
1071
+ },
1072
+ {
1073
+ "epoch": 4.64,
1074
+ "learning_rate": 2.8297304283551728e-05,
1075
+ "loss": 0.2613,
1076
+ "step": 890
1077
+ },
1078
+ {
1079
+ "epoch": 4.66,
1080
+ "learning_rate": 2.8092315781733696e-05,
1081
+ "loss": 0.223,
1082
+ "step": 895
1083
+ },
1084
+ {
1085
+ "epoch": 4.69,
1086
+ "learning_rate": 2.788711592423966e-05,
1087
+ "loss": 0.2265,
1088
+ "step": 900
1089
+ },
1090
+ {
1091
+ "epoch": 4.71,
1092
+ "learning_rate": 2.7681718736207298e-05,
1093
+ "loss": 0.2338,
1094
+ "step": 905
1095
+ },
1096
+ {
1097
+ "epoch": 4.74,
1098
+ "learning_rate": 2.7476138256261575e-05,
1099
+ "loss": 0.2393,
1100
+ "step": 910
1101
+ },
1102
+ {
1103
+ "epoch": 4.77,
1104
+ "learning_rate": 2.727038853555521e-05,
1105
+ "loss": 0.2447,
1106
+ "step": 915
1107
+ },
1108
+ {
1109
+ "epoch": 4.79,
1110
+ "learning_rate": 2.7064483636808313e-05,
1111
+ "loss": 0.2319,
1112
+ "step": 920
1113
+ },
1114
+ {
1115
+ "epoch": 4.82,
1116
+ "learning_rate": 2.6858437633347194e-05,
1117
+ "loss": 0.2762,
1118
+ "step": 925
1119
+ },
1120
+ {
1121
+ "epoch": 4.84,
1122
+ "learning_rate": 2.6652264608142484e-05,
1123
+ "loss": 0.24,
1124
+ "step": 930
1125
+ },
1126
+ {
1127
+ "epoch": 4.87,
1128
+ "learning_rate": 2.6445978652846602e-05,
1129
+ "loss": 0.2366,
1130
+ "step": 935
1131
+ },
1132
+ {
1133
+ "epoch": 4.9,
1134
+ "learning_rate": 2.623959386683056e-05,
1135
+ "loss": 0.2354,
1136
+ "step": 940
1137
+ },
1138
+ {
1139
+ "epoch": 4.92,
1140
+ "learning_rate": 2.6033124356220328e-05,
1141
+ "loss": 0.247,
1142
+ "step": 945
1143
+ },
1144
+ {
1145
+ "epoch": 4.95,
1146
+ "learning_rate": 2.5826584232932706e-05,
1147
+ "loss": 0.2143,
1148
+ "step": 950
1149
+ },
1150
+ {
1151
+ "epoch": 4.97,
1152
+ "learning_rate": 2.5619987613710756e-05,
1153
+ "loss": 0.2412,
1154
+ "step": 955
1155
+ },
1156
+ {
1157
+ "epoch": 5.0,
1158
+ "learning_rate": 2.5413348619158967e-05,
1159
+ "loss": 0.2407,
1160
+ "step": 960
1161
+ },
1162
+ {
1163
+ "epoch": 5.03,
1164
+ "learning_rate": 2.5206681372778124e-05,
1165
+ "loss": 0.1533,
1166
+ "step": 965
1167
+ },
1168
+ {
1169
+ "epoch": 5.05,
1170
+ "learning_rate": 2.5e-05,
1171
+ "loss": 0.1639,
1172
+ "step": 970
1173
+ },
1174
+ {
1175
+ "epoch": 5.08,
1176
+ "learning_rate": 2.4793318627221878e-05,
1177
+ "loss": 0.1375,
1178
+ "step": 975
1179
+ },
1180
+ {
1181
+ "epoch": 5.1,
1182
+ "learning_rate": 2.458665138084104e-05,
1183
+ "loss": 0.1514,
1184
+ "step": 980
1185
+ },
1186
+ {
1187
+ "epoch": 5.13,
1188
+ "learning_rate": 2.438001238628925e-05,
1189
+ "loss": 0.1482,
1190
+ "step": 985
1191
+ },
1192
+ {
1193
+ "epoch": 5.16,
1194
+ "learning_rate": 2.4173415767067297e-05,
1195
+ "loss": 0.1462,
1196
+ "step": 990
1197
+ },
1198
+ {
1199
+ "epoch": 5.18,
1200
+ "learning_rate": 2.3966875643779667e-05,
1201
+ "loss": 0.1428,
1202
+ "step": 995
1203
+ },
1204
+ {
1205
+ "epoch": 5.21,
1206
+ "learning_rate": 2.3760406133169443e-05,
1207
+ "loss": 0.1342,
1208
+ "step": 1000
1209
+ },
1210
+ {
1211
+ "epoch": 5.23,
1212
+ "learning_rate": 2.3554021347153403e-05,
1213
+ "loss": 0.1363,
1214
+ "step": 1005
1215
+ },
1216
+ {
1217
+ "epoch": 5.26,
1218
+ "learning_rate": 2.334773539185752e-05,
1219
+ "loss": 0.1293,
1220
+ "step": 1010
1221
+ },
1222
+ {
1223
+ "epoch": 5.29,
1224
+ "learning_rate": 2.3141562366652812e-05,
1225
+ "loss": 0.1418,
1226
+ "step": 1015
1227
+ },
1228
+ {
1229
+ "epoch": 5.31,
1230
+ "learning_rate": 2.2935516363191693e-05,
1231
+ "loss": 0.14,
1232
+ "step": 1020
1233
+ },
1234
+ {
1235
+ "epoch": 5.34,
1236
+ "learning_rate": 2.2729611464444794e-05,
1237
+ "loss": 0.1384,
1238
+ "step": 1025
1239
+ },
1240
+ {
1241
+ "epoch": 5.36,
1242
+ "learning_rate": 2.2523861743738434e-05,
1243
+ "loss": 0.1404,
1244
+ "step": 1030
1245
+ },
1246
+ {
1247
+ "epoch": 5.39,
1248
+ "learning_rate": 2.231828126379271e-05,
1249
+ "loss": 0.1367,
1250
+ "step": 1035
1251
+ },
1252
+ {
1253
+ "epoch": 5.42,
1254
+ "learning_rate": 2.2112884075760347e-05,
1255
+ "loss": 0.1497,
1256
+ "step": 1040
1257
+ },
1258
+ {
1259
+ "epoch": 5.44,
1260
+ "learning_rate": 2.190768421826631e-05,
1261
+ "loss": 0.1347,
1262
+ "step": 1045
1263
+ },
1264
+ {
1265
+ "epoch": 5.47,
1266
+ "learning_rate": 2.1702695716448278e-05,
1267
+ "loss": 0.1437,
1268
+ "step": 1050
1269
+ },
1270
+ {
1271
+ "epoch": 5.49,
1272
+ "learning_rate": 2.1497932580998053e-05,
1273
+ "loss": 0.1393,
1274
+ "step": 1055
1275
+ },
1276
+ {
1277
+ "epoch": 5.52,
1278
+ "learning_rate": 2.1293408807203947e-05,
1279
+ "loss": 0.1673,
1280
+ "step": 1060
1281
+ },
1282
+ {
1283
+ "epoch": 5.55,
1284
+ "learning_rate": 2.1089138373994223e-05,
1285
+ "loss": 0.1413,
1286
+ "step": 1065
1287
+ },
1288
+ {
1289
+ "epoch": 5.57,
1290
+ "learning_rate": 2.088513524298165e-05,
1291
+ "loss": 0.1402,
1292
+ "step": 1070
1293
+ },
1294
+ {
1295
+ "epoch": 5.6,
1296
+ "learning_rate": 2.068141335750925e-05,
1297
+ "loss": 0.1345,
1298
+ "step": 1075
1299
+ },
1300
+ {
1301
+ "epoch": 5.62,
1302
+ "learning_rate": 2.047798664169726e-05,
1303
+ "loss": 0.144,
1304
+ "step": 1080
1305
+ },
1306
+ {
1307
+ "epoch": 5.65,
1308
+ "learning_rate": 2.02748689994915e-05,
1309
+ "loss": 0.1436,
1310
+ "step": 1085
1311
+ },
1312
+ {
1313
+ "epoch": 5.68,
1314
+ "learning_rate": 2.0072074313712997e-05,
1315
+ "loss": 0.1416,
1316
+ "step": 1090
1317
+ },
1318
+ {
1319
+ "epoch": 5.7,
1320
+ "learning_rate": 1.9869616445109147e-05,
1321
+ "loss": 0.1463,
1322
+ "step": 1095
1323
+ },
1324
+ {
1325
+ "epoch": 5.73,
1326
+ "learning_rate": 1.9667509231406334e-05,
1327
+ "loss": 0.1383,
1328
+ "step": 1100
1329
+ },
1330
+ {
1331
+ "epoch": 5.76,
1332
+ "learning_rate": 1.9465766486364143e-05,
1333
+ "loss": 0.1288,
1334
+ "step": 1105
1335
+ },
1336
+ {
1337
+ "epoch": 5.78,
1338
+ "learning_rate": 1.9264401998831213e-05,
1339
+ "loss": 0.147,
1340
+ "step": 1110
1341
+ },
1342
+ {
1343
+ "epoch": 5.81,
1344
+ "learning_rate": 1.9063429531802786e-05,
1345
+ "loss": 0.1297,
1346
+ "step": 1115
1347
+ },
1348
+ {
1349
+ "epoch": 5.83,
1350
+ "learning_rate": 1.8862862821480025e-05,
1351
+ "loss": 0.1242,
1352
+ "step": 1120
1353
+ },
1354
+ {
1355
+ "epoch": 5.86,
1356
+ "learning_rate": 1.866271557633115e-05,
1357
+ "loss": 0.14,
1358
+ "step": 1125
1359
+ },
1360
+ {
1361
+ "epoch": 5.89,
1362
+ "learning_rate": 1.8463001476154508e-05,
1363
+ "loss": 0.1331,
1364
+ "step": 1130
1365
+ },
1366
+ {
1367
+ "epoch": 5.91,
1368
+ "learning_rate": 1.826373417114355e-05,
1369
+ "loss": 0.1461,
1370
+ "step": 1135
1371
+ },
1372
+ {
1373
+ "epoch": 5.94,
1374
+ "learning_rate": 1.806492728095389e-05,
1375
+ "loss": 0.1305,
1376
+ "step": 1140
1377
+ },
1378
+ {
1379
+ "epoch": 5.96,
1380
+ "learning_rate": 1.7866594393772373e-05,
1381
+ "loss": 0.1244,
1382
+ "step": 1145
1383
+ },
1384
+ {
1385
+ "epoch": 5.99,
1386
+ "learning_rate": 1.7668749065388385e-05,
1387
+ "loss": 0.1427,
1388
+ "step": 1150
1389
+ },
1390
+ {
1391
+ "epoch": 6.02,
1392
+ "learning_rate": 1.7471404818267316e-05,
1393
+ "loss": 0.0886,
1394
+ "step": 1155
1395
+ },
1396
+ {
1397
+ "epoch": 6.04,
1398
+ "learning_rate": 1.7274575140626318e-05,
1399
+ "loss": 0.0576,
1400
+ "step": 1160
1401
+ },
1402
+ {
1403
+ "epoch": 6.07,
1404
+ "learning_rate": 1.7078273485512392e-05,
1405
+ "loss": 0.0725,
1406
+ "step": 1165
1407
+ },
1408
+ {
1409
+ "epoch": 6.09,
1410
+ "learning_rate": 1.6882513269882917e-05,
1411
+ "loss": 0.0488,
1412
+ "step": 1170
1413
+ },
1414
+ {
1415
+ "epoch": 6.12,
1416
+ "learning_rate": 1.668730787368858e-05,
1417
+ "loss": 0.0484,
1418
+ "step": 1175
1419
+ },
1420
+ {
1421
+ "epoch": 6.15,
1422
+ "learning_rate": 1.6492670638958924e-05,
1423
+ "loss": 0.0465,
1424
+ "step": 1180
1425
+ },
1426
+ {
1427
+ "epoch": 6.17,
1428
+ "learning_rate": 1.6298614868890387e-05,
1429
+ "loss": 0.0607,
1430
+ "step": 1185
1431
+ },
1432
+ {
1433
+ "epoch": 6.2,
1434
+ "learning_rate": 1.6105153826937085e-05,
1435
+ "loss": 0.0452,
1436
+ "step": 1190
1437
+ },
1438
+ {
1439
+ "epoch": 6.22,
1440
+ "learning_rate": 1.591230073590425e-05,
1441
+ "loss": 0.0572,
1442
+ "step": 1195
1443
+ },
1444
+ {
1445
+ "epoch": 6.25,
1446
+ "learning_rate": 1.5720068777044476e-05,
1447
+ "loss": 0.0626,
1448
+ "step": 1200
1449
+ },
1450
+ {
1451
+ "epoch": 6.28,
1452
+ "learning_rate": 1.5528471089156804e-05,
1453
+ "loss": 0.0487,
1454
+ "step": 1205
1455
+ },
1456
+ {
1457
+ "epoch": 6.3,
1458
+ "learning_rate": 1.5337520767688703e-05,
1459
+ "loss": 0.0539,
1460
+ "step": 1210
1461
+ },
1462
+ {
1463
+ "epoch": 6.33,
1464
+ "learning_rate": 1.5147230863840966e-05,
1465
+ "loss": 0.0514,
1466
+ "step": 1215
1467
+ },
1468
+ {
1469
+ "epoch": 6.35,
1470
+ "learning_rate": 1.495761438367577e-05,
1471
+ "loss": 0.0626,
1472
+ "step": 1220
1473
+ },
1474
+ {
1475
+ "epoch": 6.38,
1476
+ "learning_rate": 1.4768684287227652e-05,
1477
+ "loss": 0.0611,
1478
+ "step": 1225
1479
+ },
1480
+ {
1481
+ "epoch": 6.41,
1482
+ "learning_rate": 1.4580453487617745e-05,
1483
+ "loss": 0.0717,
1484
+ "step": 1230
1485
+ },
1486
+ {
1487
+ "epoch": 6.43,
1488
+ "learning_rate": 1.439293485017116e-05,
1489
+ "loss": 0.0608,
1490
+ "step": 1235
1491
+ },
1492
+ {
1493
+ "epoch": 6.46,
1494
+ "learning_rate": 1.4206141191537682e-05,
1495
+ "loss": 0.0665,
1496
+ "step": 1240
1497
+ },
1498
+ {
1499
+ "epoch": 6.48,
1500
+ "learning_rate": 1.4020085278815745e-05,
1501
+ "loss": 0.0489,
1502
+ "step": 1245
1503
+ },
1504
+ {
1505
+ "epoch": 6.51,
1506
+ "learning_rate": 1.383477982867984e-05,
1507
+ "loss": 0.052,
1508
+ "step": 1250
1509
+ },
1510
+ {
1511
+ "epoch": 6.54,
1512
+ "learning_rate": 1.3650237506511331e-05,
1513
+ "loss": 0.0664,
1514
+ "step": 1255
1515
+ },
1516
+ {
1517
+ "epoch": 6.56,
1518
+ "learning_rate": 1.346647092553281e-05,
1519
+ "loss": 0.0651,
1520
+ "step": 1260
1521
+ },
1522
+ {
1523
+ "epoch": 6.59,
1524
+ "learning_rate": 1.3283492645945966e-05,
1525
+ "loss": 0.0488,
1526
+ "step": 1265
1527
+ },
1528
+ {
1529
+ "epoch": 6.61,
1530
+ "learning_rate": 1.3101315174073162e-05,
1531
+ "loss": 0.0669,
1532
+ "step": 1270
1533
+ },
1534
+ {
1535
+ "epoch": 6.64,
1536
+ "learning_rate": 1.2919950961502603e-05,
1537
+ "loss": 0.0561,
1538
+ "step": 1275
1539
+ },
1540
+ {
1541
+ "epoch": 6.67,
1542
+ "learning_rate": 1.2739412404237306e-05,
1543
+ "loss": 0.0586,
1544
+ "step": 1280
1545
+ },
1546
+ {
1547
+ "epoch": 6.69,
1548
+ "learning_rate": 1.255971184184783e-05,
1549
+ "loss": 0.0632,
1550
+ "step": 1285
1551
+ },
1552
+ {
1553
+ "epoch": 6.72,
1554
+ "learning_rate": 1.2380861556628915e-05,
1555
+ "loss": 0.0618,
1556
+ "step": 1290
1557
+ },
1558
+ {
1559
+ "epoch": 6.74,
1560
+ "learning_rate": 1.2202873772759981e-05,
1561
+ "loss": 0.0526,
1562
+ "step": 1295
1563
+ },
1564
+ {
1565
+ "epoch": 6.77,
1566
+ "learning_rate": 1.202576065546963e-05,
1567
+ "loss": 0.0604,
1568
+ "step": 1300
1569
+ },
1570
+ {
1571
+ "epoch": 6.8,
1572
+ "learning_rate": 1.1849534310204152e-05,
1573
+ "loss": 0.0538,
1574
+ "step": 1305
1575
+ },
1576
+ {
1577
+ "epoch": 6.82,
1578
+ "learning_rate": 1.1674206781800162e-05,
1579
+ "loss": 0.062,
1580
+ "step": 1310
1581
+ },
1582
+ {
1583
+ "epoch": 6.85,
1584
+ "learning_rate": 1.1499790053661327e-05,
1585
+ "loss": 0.062,
1586
+ "step": 1315
1587
+ },
1588
+ {
1589
+ "epoch": 6.88,
1590
+ "learning_rate": 1.1326296046939333e-05,
1591
+ "loss": 0.0588,
1592
+ "step": 1320
1593
+ },
1594
+ {
1595
+ "epoch": 6.9,
1596
+ "learning_rate": 1.1153736619719077e-05,
1597
+ "loss": 0.0506,
1598
+ "step": 1325
1599
+ },
1600
+ {
1601
+ "epoch": 6.93,
1602
+ "learning_rate": 1.0982123566208185e-05,
1603
+ "loss": 0.0606,
1604
+ "step": 1330
1605
+ },
1606
+ {
1607
+ "epoch": 6.95,
1608
+ "learning_rate": 1.0811468615930911e-05,
1609
+ "loss": 0.0594,
1610
+ "step": 1335
1611
+ },
1612
+ {
1613
+ "epoch": 6.98,
1614
+ "learning_rate": 1.064178343292641e-05,
1615
+ "loss": 0.0514,
1616
+ "step": 1340
1617
+ },
1618
+ {
1619
+ "epoch": 7.01,
1620
+ "learning_rate": 1.0473079614951545e-05,
1621
+ "loss": 0.0521,
1622
+ "step": 1345
1623
+ },
1624
+ {
1625
+ "epoch": 7.03,
1626
+ "learning_rate": 1.0305368692688174e-05,
1627
+ "loss": 0.0196,
1628
+ "step": 1350
1629
+ },
1630
+ {
1631
+ "epoch": 7.06,
1632
+ "learning_rate": 1.0138662128955053e-05,
1633
+ "loss": 0.0296,
1634
+ "step": 1355
1635
+ },
1636
+ {
1637
+ "epoch": 7.08,
1638
+ "learning_rate": 9.972971317924374e-06,
1639
+ "loss": 0.0168,
1640
+ "step": 1360
1641
+ },
1642
+ {
1643
+ "epoch": 7.11,
1644
+ "learning_rate": 9.808307584342971e-06,
1645
+ "loss": 0.0255,
1646
+ "step": 1365
1647
+ },
1648
+ {
1649
+ "epoch": 7.14,
1650
+ "learning_rate": 9.644682182758306e-06,
1651
+ "loss": 0.0287,
1652
+ "step": 1370
1653
+ },
1654
+ {
1655
+ "epoch": 7.16,
1656
+ "learning_rate": 9.482106296749221e-06,
1657
+ "loss": 0.0238,
1658
+ "step": 1375
1659
+ },
1660
+ {
1661
+ "epoch": 7.19,
1662
+ "learning_rate": 9.320591038161574e-06,
1663
+ "loss": 0.0243,
1664
+ "step": 1380
1665
+ },
1666
+ {
1667
+ "epoch": 7.21,
1668
+ "learning_rate": 9.160147446348739e-06,
1669
+ "loss": 0.0198,
1670
+ "step": 1385
1671
+ },
1672
+ {
1673
+ "epoch": 7.24,
1674
+ "learning_rate": 9.000786487417085e-06,
1675
+ "loss": 0.0231,
1676
+ "step": 1390
1677
+ },
1678
+ {
1679
+ "epoch": 7.27,
1680
+ "learning_rate": 8.842519053476476e-06,
1681
+ "loss": 0.0266,
1682
+ "step": 1395
1683
+ },
1684
+ {
1685
+ "epoch": 7.29,
1686
+ "learning_rate": 8.685355961895784e-06,
1687
+ "loss": 0.0184,
1688
+ "step": 1400
1689
+ },
1690
+ {
1691
+ "epoch": 7.32,
1692
+ "learning_rate": 8.52930795456355e-06,
1693
+ "loss": 0.03,
1694
+ "step": 1405
1695
+ },
1696
+ {
1697
+ "epoch": 7.34,
1698
+ "learning_rate": 8.374385697153792e-06,
1699
+ "loss": 0.0203,
1700
+ "step": 1410
1701
+ },
1702
+ {
1703
+ "epoch": 7.37,
1704
+ "learning_rate": 8.220599778397017e-06,
1705
+ "loss": 0.0264,
1706
+ "step": 1415
1707
+ },
1708
+ {
1709
+ "epoch": 7.4,
1710
+ "learning_rate": 8.067960709356478e-06,
1711
+ "loss": 0.0237,
1712
+ "step": 1420
1713
+ },
1714
+ {
1715
+ "epoch": 7.42,
1716
+ "learning_rate": 7.91647892270979e-06,
1717
+ "loss": 0.0251,
1718
+ "step": 1425
1719
+ },
1720
+ {
1721
+ "epoch": 7.45,
1722
+ "learning_rate": 7.766164772035856e-06,
1723
+ "loss": 0.0265,
1724
+ "step": 1430
1725
+ },
1726
+ {
1727
+ "epoch": 7.47,
1728
+ "learning_rate": 7.617028531107201e-06,
1729
+ "loss": 0.0248,
1730
+ "step": 1435
1731
+ },
1732
+ {
1733
+ "epoch": 7.5,
1734
+ "learning_rate": 7.469080393187786e-06,
1735
+ "loss": 0.0242,
1736
+ "step": 1440
1737
+ },
1738
+ {
1739
+ "epoch": 7.53,
1740
+ "learning_rate": 7.3223304703363135e-06,
1741
+ "loss": 0.0234,
1742
+ "step": 1445
1743
+ },
1744
+ {
1745
+ "epoch": 7.55,
1746
+ "learning_rate": 7.176788792715075e-06,
1747
+ "loss": 0.0193,
1748
+ "step": 1450
1749
+ },
1750
+ {
1751
+ "epoch": 7.58,
1752
+ "learning_rate": 7.032465307904404e-06,
1753
+ "loss": 0.0322,
1754
+ "step": 1455
1755
+ },
1756
+ {
1757
+ "epoch": 7.6,
1758
+ "learning_rate": 6.889369880222776e-06,
1759
+ "loss": 0.0263,
1760
+ "step": 1460
1761
+ },
1762
+ {
1763
+ "epoch": 7.63,
1764
+ "learning_rate": 6.747512290052596e-06,
1765
+ "loss": 0.0209,
1766
+ "step": 1465
1767
+ },
1768
+ {
1769
+ "epoch": 7.66,
1770
+ "learning_rate": 6.606902233171711e-06,
1771
+ "loss": 0.023,
1772
+ "step": 1470
1773
+ },
1774
+ {
1775
+ "epoch": 7.68,
1776
+ "learning_rate": 6.46754932009073e-06,
1777
+ "loss": 0.0266,
1778
+ "step": 1475
1779
+ },
1780
+ {
1781
+ "epoch": 7.71,
1782
+ "learning_rate": 6.329463075396161e-06,
1783
+ "loss": 0.0228,
1784
+ "step": 1480
1785
+ },
1786
+ {
1787
+ "epoch": 7.73,
1788
+ "learning_rate": 6.192652937099388e-06,
1789
+ "loss": 0.0268,
1790
+ "step": 1485
1791
+ },
1792
+ {
1793
+ "epoch": 7.76,
1794
+ "learning_rate": 6.057128255991637e-06,
1795
+ "loss": 0.0261,
1796
+ "step": 1490
1797
+ },
1798
+ {
1799
+ "epoch": 7.79,
1800
+ "learning_rate": 5.9228982950048416e-06,
1801
+ "loss": 0.0198,
1802
+ "step": 1495
1803
+ },
1804
+ {
1805
+ "epoch": 7.81,
1806
+ "learning_rate": 5.78997222857853e-06,
1807
+ "loss": 0.0219,
1808
+ "step": 1500
1809
+ },
1810
+ {
1811
+ "epoch": 7.84,
1812
+ "learning_rate": 5.6583591420327684e-06,
1813
+ "loss": 0.0194,
1814
+ "step": 1505
1815
+ },
1816
+ {
1817
+ "epoch": 7.86,
1818
+ "learning_rate": 5.528068030947192e-06,
1819
+ "loss": 0.0227,
1820
+ "step": 1510
1821
+ },
1822
+ {
1823
+ "epoch": 7.89,
1824
+ "learning_rate": 5.399107800546177e-06,
1825
+ "loss": 0.0297,
1826
+ "step": 1515
1827
+ },
1828
+ {
1829
+ "epoch": 7.92,
1830
+ "learning_rate": 5.271487265090163e-06,
1831
+ "loss": 0.0197,
1832
+ "step": 1520
1833
+ },
1834
+ {
1835
+ "epoch": 7.94,
1836
+ "learning_rate": 5.145215147273224e-06,
1837
+ "loss": 0.0336,
1838
+ "step": 1525
1839
+ },
1840
+ {
1841
+ "epoch": 7.97,
1842
+ "learning_rate": 5.0203000776268825e-06,
1843
+ "loss": 0.0132,
1844
+ "step": 1530
1845
+ },
1846
+ {
1847
+ "epoch": 7.99,
1848
+ "learning_rate": 4.896750593930216e-06,
1849
+ "loss": 0.0226,
1850
+ "step": 1535
1851
+ },
1852
+ {
1853
+ "epoch": 8.02,
1854
+ "learning_rate": 4.7745751406263165e-06,
1855
+ "loss": 0.0075,
1856
+ "step": 1540
1857
+ },
1858
+ {
1859
+ "epoch": 8.05,
1860
+ "learning_rate": 4.653782068245127e-06,
1861
+ "loss": 0.0124,
1862
+ "step": 1545
1863
+ },
1864
+ {
1865
+ "epoch": 8.07,
1866
+ "learning_rate": 4.534379632832692e-06,
1867
+ "loss": 0.0083,
1868
+ "step": 1550
1869
+ },
1870
+ {
1871
+ "epoch": 8.1,
1872
+ "learning_rate": 4.416375995386857e-06,
1873
+ "loss": 0.0104,
1874
+ "step": 1555
1875
+ },
1876
+ {
1877
+ "epoch": 8.12,
1878
+ "learning_rate": 4.299779221299499e-06,
1879
+ "loss": 0.009,
1880
+ "step": 1560
1881
+ },
1882
+ {
1883
+ "epoch": 8.15,
1884
+ "learning_rate": 4.184597279805241e-06,
1885
+ "loss": 0.0115,
1886
+ "step": 1565
1887
+ },
1888
+ {
1889
+ "epoch": 8.18,
1890
+ "learning_rate": 4.070838043436786e-06,
1891
+ "loss": 0.013,
1892
+ "step": 1570
1893
+ },
1894
+ {
1895
+ "epoch": 8.2,
1896
+ "learning_rate": 3.958509287486823e-06,
1897
+ "loss": 0.0132,
1898
+ "step": 1575
1899
+ },
1900
+ {
1901
+ "epoch": 8.23,
1902
+ "learning_rate": 3.847618689476612e-06,
1903
+ "loss": 0.01,
1904
+ "step": 1580
1905
+ },
1906
+ {
1907
+ "epoch": 8.26,
1908
+ "learning_rate": 3.738173828631228e-06,
1909
+ "loss": 0.0113,
1910
+ "step": 1585
1911
+ },
1912
+ {
1913
+ "epoch": 8.28,
1914
+ "learning_rate": 3.630182185361522e-06,
1915
+ "loss": 0.0105,
1916
+ "step": 1590
1917
+ },
1918
+ {
1919
+ "epoch": 8.31,
1920
+ "learning_rate": 3.523651140752868e-06,
1921
+ "loss": 0.0085,
1922
+ "step": 1595
1923
+ },
1924
+ {
1925
+ "epoch": 8.33,
1926
+ "learning_rate": 3.418587976060653e-06,
1927
+ "loss": 0.0201,
1928
+ "step": 1600
1929
+ },
1930
+ {
1931
+ "epoch": 8.36,
1932
+ "learning_rate": 3.314999872212618e-06,
1933
+ "loss": 0.011,
1934
+ "step": 1605
1935
+ },
1936
+ {
1937
+ "epoch": 8.39,
1938
+ "learning_rate": 3.2128939093180655e-06,
1939
+ "loss": 0.0106,
1940
+ "step": 1610
1941
+ },
1942
+ {
1943
+ "epoch": 8.41,
1944
+ "learning_rate": 3.11227706618393e-06,
1945
+ "loss": 0.0146,
1946
+ "step": 1615
1947
+ },
1948
+ {
1949
+ "epoch": 8.44,
1950
+ "learning_rate": 3.013156219837776e-06,
1951
+ "loss": 0.0123,
1952
+ "step": 1620
1953
+ },
1954
+ {
1955
+ "epoch": 8.46,
1956
+ "learning_rate": 2.9155381450577863e-06,
1957
+ "loss": 0.0128,
1958
+ "step": 1625
1959
+ },
1960
+ {
1961
+ "epoch": 8.49,
1962
+ "learning_rate": 2.8194295139097048e-06,
1963
+ "loss": 0.0093,
1964
+ "step": 1630
1965
+ },
1966
+ {
1967
+ "epoch": 8.52,
1968
+ "learning_rate": 2.7248368952908053e-06,
1969
+ "loss": 0.0085,
1970
+ "step": 1635
1971
+ },
1972
+ {
1973
+ "epoch": 8.54,
1974
+ "learning_rate": 2.6317667544809134e-06,
1975
+ "loss": 0.0072,
1976
+ "step": 1640
1977
+ },
1978
+ {
1979
+ "epoch": 8.57,
1980
+ "learning_rate": 2.5402254527005287e-06,
1981
+ "loss": 0.0103,
1982
+ "step": 1645
1983
+ },
1984
+ {
1985
+ "epoch": 8.59,
1986
+ "learning_rate": 2.4502192466760276e-06,
1987
+ "loss": 0.0156,
1988
+ "step": 1650
1989
+ },
1990
+ {
1991
+ "epoch": 8.62,
1992
+ "learning_rate": 2.361754288212031e-06,
1993
+ "loss": 0.0128,
1994
+ "step": 1655
1995
+ },
1996
+ {
1997
+ "epoch": 8.65,
1998
+ "learning_rate": 2.2748366237709374e-06,
1999
+ "loss": 0.0113,
2000
+ "step": 1660
2001
+ },
2002
+ {
2003
+ "epoch": 8.67,
2004
+ "learning_rate": 2.1894721940596554e-06,
2005
+ "loss": 0.0076,
2006
+ "step": 1665
2007
+ },
2008
+ {
2009
+ "epoch": 8.7,
2010
+ "learning_rate": 2.1056668336235622e-06,
2011
+ "loss": 0.01,
2012
+ "step": 1670
2013
+ },
2014
+ {
2015
+ "epoch": 8.72,
2016
+ "learning_rate": 2.023426270447723e-06,
2017
+ "loss": 0.0104,
2018
+ "step": 1675
2019
+ },
2020
+ {
2021
+ "epoch": 8.75,
2022
+ "learning_rate": 1.9427561255653816e-06,
2023
+ "loss": 0.0078,
2024
+ "step": 1680
2025
+ },
2026
+ {
2027
+ "epoch": 8.78,
2028
+ "learning_rate": 1.8636619126737892e-06,
2029
+ "loss": 0.0122,
2030
+ "step": 1685
2031
+ },
2032
+ {
2033
+ "epoch": 8.8,
2034
+ "learning_rate": 1.7861490377573258e-06,
2035
+ "loss": 0.008,
2036
+ "step": 1690
2037
+ },
2038
+ {
2039
+ "epoch": 8.83,
2040
+ "learning_rate": 1.710222798718028e-06,
2041
+ "loss": 0.0112,
2042
+ "step": 1695
2043
+ },
2044
+ {
2045
+ "epoch": 8.85,
2046
+ "learning_rate": 1.6358883850134816e-06,
2047
+ "loss": 0.0087,
2048
+ "step": 1700
2049
+ },
2050
+ {
2051
+ "epoch": 8.88,
2052
+ "learning_rate": 1.5631508773021165e-06,
2053
+ "loss": 0.0102,
2054
+ "step": 1705
2055
+ },
2056
+ {
2057
+ "epoch": 8.91,
2058
+ "learning_rate": 1.4920152470959707e-06,
2059
+ "loss": 0.0115,
2060
+ "step": 1710
2061
+ },
2062
+ {
2063
+ "epoch": 8.93,
2064
+ "learning_rate": 1.4224863564208684e-06,
2065
+ "loss": 0.0081,
2066
+ "step": 1715
2067
+ },
2068
+ {
2069
+ "epoch": 8.96,
2070
+ "learning_rate": 1.3545689574841342e-06,
2071
+ "loss": 0.0113,
2072
+ "step": 1720
2073
+ },
2074
+ {
2075
+ "epoch": 8.98,
2076
+ "learning_rate": 1.288267692349765e-06,
2077
+ "loss": 0.0119,
2078
+ "step": 1725
2079
+ },
2080
+ {
2081
+ "epoch": 9.01,
2082
+ "learning_rate": 1.2235870926211619e-06,
2083
+ "loss": 0.0062,
2084
+ "step": 1730
2085
+ },
2086
+ {
2087
+ "epoch": 9.04,
2088
+ "learning_rate": 1.1605315791313964e-06,
2089
+ "loss": 0.0063,
2090
+ "step": 1735
2091
+ },
2092
+ {
2093
+ "epoch": 9.06,
2094
+ "learning_rate": 1.0991054616410589e-06,
2095
+ "loss": 0.005,
2096
+ "step": 1740
2097
+ },
2098
+ {
2099
+ "epoch": 9.09,
2100
+ "learning_rate": 1.0393129385436824e-06,
2101
+ "loss": 0.0064,
2102
+ "step": 1745
2103
+ },
2104
+ {
2105
+ "epoch": 9.11,
2106
+ "learning_rate": 9.811580965787965e-07,
2107
+ "loss": 0.0061,
2108
+ "step": 1750
2109
+ },
2110
+ {
2111
+ "epoch": 9.14,
2112
+ "learning_rate": 9.246449105525995e-07,
2113
+ "loss": 0.0054,
2114
+ "step": 1755
2115
+ },
2116
+ {
2117
+ "epoch": 9.17,
2118
+ "learning_rate": 8.697772430662859e-07,
2119
+ "loss": 0.0052,
2120
+ "step": 1760
2121
+ },
2122
+ {
2123
+ "epoch": 9.19,
2124
+ "learning_rate": 8.165588442520439e-07,
2125
+ "loss": 0.004,
2126
+ "step": 1765
2127
+ },
2128
+ {
2129
+ "epoch": 9.22,
2130
+ "learning_rate": 7.649933515167407e-07,
2131
+ "loss": 0.0041,
2132
+ "step": 1770
2133
+ },
2134
+ {
2135
+ "epoch": 9.24,
2136
+ "learning_rate": 7.150842892933107e-07,
2137
+ "loss": 0.006,
2138
+ "step": 1775
2139
+ },
2140
+ {
2141
+ "epoch": 9.27,
2142
+ "learning_rate": 6.668350687998565e-07,
2143
+ "loss": 0.0059,
2144
+ "step": 1780
2145
+ },
2146
+ {
2147
+ "epoch": 9.3,
2148
+ "learning_rate": 6.202489878065071e-07,
2149
+ "loss": 0.0059,
2150
+ "step": 1785
2151
+ },
2152
+ {
2153
+ "epoch": 9.32,
2154
+ "learning_rate": 5.753292304100183e-07,
2155
+ "loss": 0.007,
2156
+ "step": 1790
2157
+ },
2158
+ {
2159
+ "epoch": 9.35,
2160
+ "learning_rate": 5.32078866816138e-07,
2161
+ "loss": 0.0077,
2162
+ "step": 1795
2163
+ },
2164
+ {
2165
+ "epoch": 9.38,
2166
+ "learning_rate": 4.905008531297661e-07,
2167
+ "loss": 0.0072,
2168
+ "step": 1800
2169
+ },
2170
+ {
2171
+ "epoch": 9.4,
2172
+ "learning_rate": 4.505980311529101e-07,
2173
+ "loss": 0.0028,
2174
+ "step": 1805
2175
+ },
2176
+ {
2177
+ "epoch": 9.43,
2178
+ "learning_rate": 4.1237312819044085e-07,
2179
+ "loss": 0.0073,
2180
+ "step": 1810
2181
+ },
2182
+ {
2183
+ "epoch": 9.45,
2184
+ "learning_rate": 3.758287568637081e-07,
2185
+ "loss": 0.0058,
2186
+ "step": 1815
2187
+ },
2188
+ {
2189
+ "epoch": 9.48,
2190
+ "learning_rate": 3.4096741493194197e-07,
2191
+ "loss": 0.0044,
2192
+ "step": 1820
2193
+ },
2194
+ {
2195
+ "epoch": 9.51,
2196
+ "learning_rate": 3.077914851215585e-07,
2197
+ "loss": 0.0078,
2198
+ "step": 1825
2199
+ },
2200
+ {
2201
+ "epoch": 9.53,
2202
+ "learning_rate": 2.763032349632877e-07,
2203
+ "loss": 0.0075,
2204
+ "step": 1830
2205
+ },
2206
+ {
2207
+ "epoch": 9.56,
2208
+ "learning_rate": 2.4650481663720525e-07,
2209
+ "loss": 0.0054,
2210
+ "step": 1835
2211
+ },
2212
+ {
2213
+ "epoch": 9.58,
2214
+ "learning_rate": 2.1839826682562015e-07,
2215
+ "loss": 0.0061,
2216
+ "step": 1840
2217
+ },
2218
+ {
2219
+ "epoch": 9.61,
2220
+ "learning_rate": 1.919855065738746e-07,
2221
+ "loss": 0.0041,
2222
+ "step": 1845
2223
+ },
2224
+ {
2225
+ "epoch": 9.64,
2226
+ "learning_rate": 1.6726834115904643e-07,
2227
+ "loss": 0.0068,
2228
+ "step": 1850
2229
+ },
2230
+ {
2231
+ "epoch": 9.66,
2232
+ "learning_rate": 1.4424845996655888e-07,
2233
+ "loss": 0.008,
2234
+ "step": 1855
2235
+ },
2236
+ {
2237
+ "epoch": 9.69,
2238
+ "learning_rate": 1.229274363747146e-07,
2239
+ "loss": 0.0107,
2240
+ "step": 1860
2241
+ },
2242
+ {
2243
+ "epoch": 9.71,
2244
+ "learning_rate": 1.0330672764715387e-07,
2245
+ "loss": 0.0071,
2246
+ "step": 1865
2247
+ },
2248
+ {
2249
+ "epoch": 9.74,
2250
+ "learning_rate": 8.538767483325383e-08,
2251
+ "loss": 0.0091,
2252
+ "step": 1870
2253
+ },
2254
+ {
2255
+ "epoch": 9.77,
2256
+ "learning_rate": 6.91715026764711e-08,
2257
+ "loss": 0.0057,
2258
+ "step": 1875
2259
+ },
2260
+ {
2261
+ "epoch": 9.79,
2262
+ "learning_rate": 5.4659319530636633e-08,
2263
+ "loss": 0.0064,
2264
+ "step": 1880
2265
+ },
2266
+ {
2267
+ "epoch": 9.82,
2268
+ "learning_rate": 4.18521172841857e-08,
2269
+ "loss": 0.0073,
2270
+ "step": 1885
2271
+ },
2272
+ {
2273
+ "epoch": 9.84,
2274
+ "learning_rate": 3.075077129238158e-08,
2275
+ "loss": 0.0066,
2276
+ "step": 1890
2277
+ },
2278
+ {
2279
+ "epoch": 9.87,
2280
+ "learning_rate": 2.1356040317474512e-08,
2281
+ "loss": 0.0052,
2282
+ "step": 1895
2283
+ },
2284
+ {
2285
+ "epoch": 9.9,
2286
+ "learning_rate": 1.3668566476848777e-08,
2287
+ "loss": 0.0054,
2288
+ "step": 1900
2289
+ },
2290
+ {
2291
+ "epoch": 9.92,
2292
+ "learning_rate": 7.688875199132751e-09,
2293
+ "loss": 0.0069,
2294
+ "step": 1905
2295
+ },
2296
+ {
2297
+ "epoch": 9.95,
2298
+ "learning_rate": 3.417375188274896e-09,
2299
+ "loss": 0.0045,
2300
+ "step": 1910
2301
+ },
2302
+ {
2303
+ "epoch": 9.97,
2304
+ "learning_rate": 8.543583956355239e-10,
2305
+ "loss": 0.0043,
2306
+ "step": 1915
2307
+ },
2308
+ {
2309
+ "epoch": 10.0,
2310
+ "learning_rate": 0.0,
2311
+ "loss": 0.0066,
2312
+ "step": 1920
2313
+ },
2314
+ {
2315
+ "epoch": 10.0,
2316
+ "step": 1920,
2317
+ "total_flos": 3556947756318720.0,
2318
+ "train_loss": 0.47322619671273664,
2319
+ "train_runtime": 757.4177,
2320
+ "train_samples_per_second": 5.057,
2321
+ "train_steps_per_second": 2.535
2322
+ }
2323
+ ],
2324
+ "max_steps": 1920,
2325
+ "num_train_epochs": 10,
2326
+ "total_flos": 3556947756318720.0,
2327
+ "trial_name": null,
2328
+ "trial_params": null
2329
+ }