ales commited on
Commit
7dee023
1 Parent(s): 2af2e4c

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.1,
3
+ "eval_loss": 0.07060841470956802,
4
+ "eval_runtime": 6849.8469,
5
+ "eval_samples_per_second": 2.317,
6
+ "eval_steps_per_second": 0.036,
7
+ "eval_wer": 6.3671568743912,
8
+ "train_loss": 0.08807742363711198,
9
+ "train_runtime": 202754.3941,
10
+ "train_samples_per_second": 3.788,
11
+ "train_steps_per_second": 0.059
12
+ }
eval_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.1,
3
+ "eval_loss": 0.07060841470956802,
4
+ "eval_runtime": 6849.8469,
5
+ "eval_samples_per_second": 2.317,
6
+ "eval_steps_per_second": 0.036,
7
+ "eval_wer": 6.3671568743912
8
+ }
runs/Dec14_11-27-46_129-213-88-66/events.out.tfevents.1671226969.129-213-88-66.335324.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad30b809964fb857b839f9fbf17f03e2e7a06b18cb5d06b546de0496ba086220
3
+ size 358
train_20221214-112746.log CHANGED
@@ -249,3 +249,7 @@ remote: LFS file scan complete.
249
  To https://huggingface.co/ales/whisper-small-belarusian
250
  72e5d42..1d54c35 main -> main
251
 
 
 
 
 
 
249
  To https://huggingface.co/ales/whisper-small-belarusian
250
  72e5d42..1d54c35 main -> main
251
 
252
+ 12/16/2022 19:48:36 - WARNING - huggingface_hub.repository - To https://huggingface.co/ales/whisper-small-belarusian
253
+ 1d54c35..dc59786 main -> main
254
+
255
+ 12/16/2022 19:48:39 - INFO - __main__ - *** Evaluate ***
train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.1,
3
+ "train_loss": 0.08807742363711198,
4
+ "train_runtime": 202754.3941,
5
+ "train_samples_per_second": 3.788,
6
+ "train_steps_per_second": 0.059
7
+ }
train_run_1.log CHANGED
@@ -33809,4 +33809,29 @@ remote: LFS file scan complete.
33809
  To https://huggingface.co/ales/whisper-small-belarusian
33810
  72e5d42..1d54c35 main -> main
33811
 
33812
-
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
33809
  To https://huggingface.co/ales/whisper-small-belarusian
33810
  72e5d42..1d54c35 main -> main
33811
 
33812
+
33813
+ To https://huggingface.co/ales/whisper-small-belarusian
33814
+ 1d54c35..dc59786 main -> main
33815
+
33816
+ 12/16/2022 19:48:36 - WARNING - huggingface_hub.repository - To https://huggingface.co/ales/whisper-small-belarusian
33817
+ 1d54c35..dc59786 main -> main
33818
+
33819
+ ***** train metrics *****
33820
+ epoch = 2.1
33821
+ train_loss = 0.0881
33822
+ train_runtime = 2 days, 8:19:14.39
33823
+ train_samples_per_second = 3.788
33824
+ train_steps_per_second = 0.059
33825
+ 12/16/2022 19:48:39 - INFO - __main__ - *** Evaluate ***
33826
+ [INFO|trainer.py:2955] 2022-12-16 19:48:39,226 >> ***** Running Evaluation *****
33827
+ [INFO|trainer.py:2959] 2022-12-16 19:48:39,226 >> Num examples: Unknown
33828
+ [INFO|trainer.py:2960] 2022-12-16 19:48:39,227 >> Batch size = 64
33829
+
33830
+ [INFO|trainer_utils.py:689] 2022-12-16 19:48:46,263 >> The following columns in the evaluation set don't have a corresponding argument in `WhisperForConditionalGeneration.forward` and have been ignored: labels_length, input_length, labels_truncated. If labels_length, input_length, labels_truncated are not expected by `WhisperForConditionalGeneration.forward`, you can safely ignore this message.
33831
+ [INFO|trainer.py:2700] 2022-12-16 21:42:49,079 >> Saving model checkpoint to ./
33832
+ [INFO|configuration_utils.py:447] 2022-12-16 21:42:49,080 >> Configuration saved in ./config.json
33833
+ [INFO|modeling_utils.py:1680] 2022-12-16 21:42:50,216 >> Model weights saved in ./pytorch_model.bin
33834
+ [INFO|feature_extraction_utils.py:368] 2022-12-16 21:42:50,232 >> Feature extractor saved in ./preprocessor_config.json
33835
+ [INFO|tokenization_utils_base.py:2157] 2022-12-16 21:42:50,233 >> tokenizer config file saved in ./tokenizer_config.json
33836
+ [INFO|tokenization_utils_base.py:2164] 2022-12-16 21:42:50,233 >> Special tokens file saved in ./special_tokens_map.json
33837
+ [INFO|tokenization_utils_base.py:2210] 2022-12-16 21:42:50,233 >> added tokens file saved in ./added_tokens.json
trainer_state.json ADDED
@@ -0,0 +1,1579 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 6.3671568743912,
3
+ "best_model_checkpoint": "./checkpoint-12000",
4
+ "epoch": 2.095833333333333,
5
+ "global_step": 12000,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.0,
12
+ "learning_rate": 0.0,
13
+ "loss": 1.9643,
14
+ "step": 1
15
+ },
16
+ {
17
+ "epoch": 0.0,
18
+ "learning_rate": 9.4e-06,
19
+ "loss": 1.184,
20
+ "step": 50
21
+ },
22
+ {
23
+ "epoch": 0.01,
24
+ "learning_rate": 1.94e-05,
25
+ "loss": 0.4478,
26
+ "step": 100
27
+ },
28
+ {
29
+ "epoch": 0.01,
30
+ "learning_rate": 2.94e-05,
31
+ "loss": 0.307,
32
+ "step": 150
33
+ },
34
+ {
35
+ "epoch": 0.02,
36
+ "learning_rate": 3.94e-05,
37
+ "loss": 0.2939,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.02,
42
+ "learning_rate": 4.94e-05,
43
+ "loss": 0.2464,
44
+ "step": 250
45
+ },
46
+ {
47
+ "epoch": 0.03,
48
+ "learning_rate": 5.94e-05,
49
+ "loss": 0.2403,
50
+ "step": 300
51
+ },
52
+ {
53
+ "epoch": 0.03,
54
+ "learning_rate": 6.939999999999999e-05,
55
+ "loss": 0.2404,
56
+ "step": 350
57
+ },
58
+ {
59
+ "epoch": 0.03,
60
+ "learning_rate": 7.94e-05,
61
+ "loss": 0.2563,
62
+ "step": 400
63
+ },
64
+ {
65
+ "epoch": 0.04,
66
+ "learning_rate": 8.94e-05,
67
+ "loss": 0.2366,
68
+ "step": 450
69
+ },
70
+ {
71
+ "epoch": 0.04,
72
+ "learning_rate": 9.94e-05,
73
+ "loss": 0.2237,
74
+ "step": 500
75
+ },
76
+ {
77
+ "epoch": 0.05,
78
+ "learning_rate": 9.959130434782609e-05,
79
+ "loss": 0.234,
80
+ "step": 550
81
+ },
82
+ {
83
+ "epoch": 0.05,
84
+ "learning_rate": 9.915652173913044e-05,
85
+ "loss": 0.2301,
86
+ "step": 600
87
+ },
88
+ {
89
+ "epoch": 0.05,
90
+ "learning_rate": 9.872173913043479e-05,
91
+ "loss": 0.2404,
92
+ "step": 650
93
+ },
94
+ {
95
+ "epoch": 0.06,
96
+ "learning_rate": 9.828695652173913e-05,
97
+ "loss": 0.2155,
98
+ "step": 700
99
+ },
100
+ {
101
+ "epoch": 0.06,
102
+ "learning_rate": 9.785217391304348e-05,
103
+ "loss": 0.1606,
104
+ "step": 750
105
+ },
106
+ {
107
+ "epoch": 0.07,
108
+ "learning_rate": 9.741739130434783e-05,
109
+ "loss": 0.1744,
110
+ "step": 800
111
+ },
112
+ {
113
+ "epoch": 0.07,
114
+ "learning_rate": 9.698260869565218e-05,
115
+ "loss": 0.2005,
116
+ "step": 850
117
+ },
118
+ {
119
+ "epoch": 0.07,
120
+ "learning_rate": 9.654782608695653e-05,
121
+ "loss": 0.2168,
122
+ "step": 900
123
+ },
124
+ {
125
+ "epoch": 0.08,
126
+ "learning_rate": 9.611304347826087e-05,
127
+ "loss": 0.1448,
128
+ "step": 950
129
+ },
130
+ {
131
+ "epoch": 0.08,
132
+ "learning_rate": 9.567826086956522e-05,
133
+ "loss": 0.1907,
134
+ "step": 1000
135
+ },
136
+ {
137
+ "epoch": 0.08,
138
+ "eval_loss": 0.2546018362045288,
139
+ "eval_runtime": 6940.5327,
140
+ "eval_samples_per_second": 2.287,
141
+ "eval_steps_per_second": 0.036,
142
+ "eval_wer": 25.463875887957045,
143
+ "step": 1000
144
+ },
145
+ {
146
+ "epoch": 0.09,
147
+ "learning_rate": 9.524347826086957e-05,
148
+ "loss": 0.1314,
149
+ "step": 1050
150
+ },
151
+ {
152
+ "epoch": 0.09,
153
+ "learning_rate": 9.480869565217392e-05,
154
+ "loss": 0.1387,
155
+ "step": 1100
156
+ },
157
+ {
158
+ "epoch": 0.1,
159
+ "learning_rate": 9.437391304347827e-05,
160
+ "loss": 0.139,
161
+ "step": 1150
162
+ },
163
+ {
164
+ "epoch": 0.1,
165
+ "learning_rate": 9.393913043478261e-05,
166
+ "loss": 0.1395,
167
+ "step": 1200
168
+ },
169
+ {
170
+ "epoch": 0.1,
171
+ "learning_rate": 9.350434782608696e-05,
172
+ "loss": 0.1268,
173
+ "step": 1250
174
+ },
175
+ {
176
+ "epoch": 0.11,
177
+ "learning_rate": 9.306956521739131e-05,
178
+ "loss": 0.177,
179
+ "step": 1300
180
+ },
181
+ {
182
+ "epoch": 0.11,
183
+ "learning_rate": 9.263478260869566e-05,
184
+ "loss": 0.1791,
185
+ "step": 1350
186
+ },
187
+ {
188
+ "epoch": 0.12,
189
+ "learning_rate": 9.22e-05,
190
+ "loss": 0.0979,
191
+ "step": 1400
192
+ },
193
+ {
194
+ "epoch": 0.12,
195
+ "learning_rate": 9.176521739130435e-05,
196
+ "loss": 0.09,
197
+ "step": 1450
198
+ },
199
+ {
200
+ "epoch": 0.12,
201
+ "learning_rate": 9.13304347826087e-05,
202
+ "loss": 0.138,
203
+ "step": 1500
204
+ },
205
+ {
206
+ "epoch": 0.13,
207
+ "learning_rate": 9.089565217391305e-05,
208
+ "loss": 0.1316,
209
+ "step": 1550
210
+ },
211
+ {
212
+ "epoch": 0.13,
213
+ "learning_rate": 9.04608695652174e-05,
214
+ "loss": 0.104,
215
+ "step": 1600
216
+ },
217
+ {
218
+ "epoch": 0.14,
219
+ "learning_rate": 9.002608695652174e-05,
220
+ "loss": 0.1265,
221
+ "step": 1650
222
+ },
223
+ {
224
+ "epoch": 0.14,
225
+ "learning_rate": 8.959130434782609e-05,
226
+ "loss": 0.1421,
227
+ "step": 1700
228
+ },
229
+ {
230
+ "epoch": 0.15,
231
+ "learning_rate": 8.915652173913044e-05,
232
+ "loss": 0.1799,
233
+ "step": 1750
234
+ },
235
+ {
236
+ "epoch": 0.15,
237
+ "learning_rate": 8.872173913043479e-05,
238
+ "loss": 0.1733,
239
+ "step": 1800
240
+ },
241
+ {
242
+ "epoch": 0.15,
243
+ "learning_rate": 8.828695652173914e-05,
244
+ "loss": 0.1525,
245
+ "step": 1850
246
+ },
247
+ {
248
+ "epoch": 0.16,
249
+ "learning_rate": 8.785217391304348e-05,
250
+ "loss": 0.1638,
251
+ "step": 1900
252
+ },
253
+ {
254
+ "epoch": 0.16,
255
+ "learning_rate": 8.741739130434783e-05,
256
+ "loss": 0.1642,
257
+ "step": 1950
258
+ },
259
+ {
260
+ "epoch": 0.17,
261
+ "learning_rate": 8.698260869565218e-05,
262
+ "loss": 0.1482,
263
+ "step": 2000
264
+ },
265
+ {
266
+ "epoch": 0.17,
267
+ "eval_loss": 0.1641348898410797,
268
+ "eval_runtime": 6931.8394,
269
+ "eval_samples_per_second": 2.29,
270
+ "eval_steps_per_second": 0.036,
271
+ "eval_wer": 17.16756551281747,
272
+ "step": 2000
273
+ },
274
+ {
275
+ "epoch": 0.17,
276
+ "learning_rate": 8.654782608695653e-05,
277
+ "loss": 0.1376,
278
+ "step": 2050
279
+ },
280
+ {
281
+ "epoch": 0.17,
282
+ "learning_rate": 8.611304347826088e-05,
283
+ "loss": 0.1323,
284
+ "step": 2100
285
+ },
286
+ {
287
+ "epoch": 0.18,
288
+ "learning_rate": 8.567826086956522e-05,
289
+ "loss": 0.1637,
290
+ "step": 2150
291
+ },
292
+ {
293
+ "epoch": 0.18,
294
+ "learning_rate": 8.524347826086957e-05,
295
+ "loss": 0.1518,
296
+ "step": 2200
297
+ },
298
+ {
299
+ "epoch": 0.19,
300
+ "learning_rate": 8.480869565217392e-05,
301
+ "loss": 0.1357,
302
+ "step": 2250
303
+ },
304
+ {
305
+ "epoch": 0.19,
306
+ "learning_rate": 8.437391304347827e-05,
307
+ "loss": 0.1373,
308
+ "step": 2300
309
+ },
310
+ {
311
+ "epoch": 0.2,
312
+ "learning_rate": 8.393913043478261e-05,
313
+ "loss": 0.1257,
314
+ "step": 2350
315
+ },
316
+ {
317
+ "epoch": 0.2,
318
+ "learning_rate": 8.350434782608696e-05,
319
+ "loss": 0.1336,
320
+ "step": 2400
321
+ },
322
+ {
323
+ "epoch": 0.2,
324
+ "learning_rate": 8.306956521739131e-05,
325
+ "loss": 0.1351,
326
+ "step": 2450
327
+ },
328
+ {
329
+ "epoch": 0.21,
330
+ "learning_rate": 8.263478260869566e-05,
331
+ "loss": 0.1159,
332
+ "step": 2500
333
+ },
334
+ {
335
+ "epoch": 0.21,
336
+ "learning_rate": 8.22e-05,
337
+ "loss": 0.1198,
338
+ "step": 2550
339
+ },
340
+ {
341
+ "epoch": 0.22,
342
+ "learning_rate": 8.176521739130435e-05,
343
+ "loss": 0.1179,
344
+ "step": 2600
345
+ },
346
+ {
347
+ "epoch": 0.22,
348
+ "learning_rate": 8.13304347826087e-05,
349
+ "loss": 0.128,
350
+ "step": 2650
351
+ },
352
+ {
353
+ "epoch": 0.23,
354
+ "learning_rate": 8.089565217391305e-05,
355
+ "loss": 0.1037,
356
+ "step": 2700
357
+ },
358
+ {
359
+ "epoch": 0.23,
360
+ "learning_rate": 8.04608695652174e-05,
361
+ "loss": 0.1172,
362
+ "step": 2750
363
+ },
364
+ {
365
+ "epoch": 0.23,
366
+ "learning_rate": 8.002608695652175e-05,
367
+ "loss": 0.1169,
368
+ "step": 2800
369
+ },
370
+ {
371
+ "epoch": 0.24,
372
+ "learning_rate": 7.959130434782609e-05,
373
+ "loss": 0.1165,
374
+ "step": 2850
375
+ },
376
+ {
377
+ "epoch": 0.24,
378
+ "learning_rate": 7.915652173913044e-05,
379
+ "loss": 0.117,
380
+ "step": 2900
381
+ },
382
+ {
383
+ "epoch": 0.25,
384
+ "learning_rate": 7.872173913043479e-05,
385
+ "loss": 0.1225,
386
+ "step": 2950
387
+ },
388
+ {
389
+ "epoch": 0.25,
390
+ "learning_rate": 7.828695652173914e-05,
391
+ "loss": 0.1175,
392
+ "step": 3000
393
+ },
394
+ {
395
+ "epoch": 0.25,
396
+ "eval_loss": 0.1453797072172165,
397
+ "eval_runtime": 6986.1082,
398
+ "eval_samples_per_second": 2.272,
399
+ "eval_steps_per_second": 0.035,
400
+ "eval_wer": 15.593990797716058,
401
+ "step": 3000
402
+ },
403
+ {
404
+ "epoch": 0.25,
405
+ "learning_rate": 7.785217391304348e-05,
406
+ "loss": 0.1011,
407
+ "step": 3050
408
+ },
409
+ {
410
+ "epoch": 0.26,
411
+ "learning_rate": 7.741739130434783e-05,
412
+ "loss": 0.1185,
413
+ "step": 3100
414
+ },
415
+ {
416
+ "epoch": 0.26,
417
+ "learning_rate": 7.698260869565218e-05,
418
+ "loss": 0.1094,
419
+ "step": 3150
420
+ },
421
+ {
422
+ "epoch": 0.27,
423
+ "learning_rate": 7.654782608695651e-05,
424
+ "loss": 0.1078,
425
+ "step": 3200
426
+ },
427
+ {
428
+ "epoch": 0.27,
429
+ "learning_rate": 7.611304347826088e-05,
430
+ "loss": 0.1171,
431
+ "step": 3250
432
+ },
433
+ {
434
+ "epoch": 0.28,
435
+ "learning_rate": 7.567826086956522e-05,
436
+ "loss": 0.1098,
437
+ "step": 3300
438
+ },
439
+ {
440
+ "epoch": 0.28,
441
+ "learning_rate": 7.524347826086957e-05,
442
+ "loss": 0.0955,
443
+ "step": 3350
444
+ },
445
+ {
446
+ "epoch": 0.28,
447
+ "learning_rate": 7.480869565217392e-05,
448
+ "loss": 0.1081,
449
+ "step": 3400
450
+ },
451
+ {
452
+ "epoch": 0.29,
453
+ "learning_rate": 7.437391304347825e-05,
454
+ "loss": 0.124,
455
+ "step": 3450
456
+ },
457
+ {
458
+ "epoch": 0.29,
459
+ "learning_rate": 7.393913043478262e-05,
460
+ "loss": 0.0987,
461
+ "step": 3500
462
+ },
463
+ {
464
+ "epoch": 0.3,
465
+ "learning_rate": 7.350434782608696e-05,
466
+ "loss": 0.0937,
467
+ "step": 3550
468
+ },
469
+ {
470
+ "epoch": 0.3,
471
+ "learning_rate": 7.306956521739131e-05,
472
+ "loss": 0.1307,
473
+ "step": 3600
474
+ },
475
+ {
476
+ "epoch": 0.3,
477
+ "learning_rate": 7.263478260869565e-05,
478
+ "loss": 0.0974,
479
+ "step": 3650
480
+ },
481
+ {
482
+ "epoch": 0.31,
483
+ "learning_rate": 7.22e-05,
484
+ "loss": 0.0971,
485
+ "step": 3700
486
+ },
487
+ {
488
+ "epoch": 0.31,
489
+ "learning_rate": 7.176521739130435e-05,
490
+ "loss": 0.1063,
491
+ "step": 3750
492
+ },
493
+ {
494
+ "epoch": 0.32,
495
+ "learning_rate": 7.13304347826087e-05,
496
+ "loss": 0.134,
497
+ "step": 3800
498
+ },
499
+ {
500
+ "epoch": 0.32,
501
+ "learning_rate": 7.089565217391305e-05,
502
+ "loss": 0.1118,
503
+ "step": 3850
504
+ },
505
+ {
506
+ "epoch": 0.33,
507
+ "learning_rate": 7.046086956521738e-05,
508
+ "loss": 0.109,
509
+ "step": 3900
510
+ },
511
+ {
512
+ "epoch": 0.33,
513
+ "learning_rate": 7.002608695652175e-05,
514
+ "loss": 0.1063,
515
+ "step": 3950
516
+ },
517
+ {
518
+ "epoch": 0.33,
519
+ "learning_rate": 6.95913043478261e-05,
520
+ "loss": 0.0958,
521
+ "step": 4000
522
+ },
523
+ {
524
+ "epoch": 0.33,
525
+ "eval_loss": 0.12609383463859558,
526
+ "eval_runtime": 6883.3335,
527
+ "eval_samples_per_second": 2.306,
528
+ "eval_steps_per_second": 0.036,
529
+ "eval_wer": 13.262534350177791,
530
+ "step": 4000
531
+ },
532
+ {
533
+ "epoch": 0.34,
534
+ "learning_rate": 6.915652173913044e-05,
535
+ "loss": 0.0891,
536
+ "step": 4050
537
+ },
538
+ {
539
+ "epoch": 0.34,
540
+ "learning_rate": 6.872173913043478e-05,
541
+ "loss": 0.1144,
542
+ "step": 4100
543
+ },
544
+ {
545
+ "epoch": 0.35,
546
+ "learning_rate": 6.828695652173914e-05,
547
+ "loss": 0.096,
548
+ "step": 4150
549
+ },
550
+ {
551
+ "epoch": 0.35,
552
+ "learning_rate": 6.785217391304349e-05,
553
+ "loss": 0.1078,
554
+ "step": 4200
555
+ },
556
+ {
557
+ "epoch": 0.35,
558
+ "learning_rate": 6.741739130434783e-05,
559
+ "loss": 0.0791,
560
+ "step": 4250
561
+ },
562
+ {
563
+ "epoch": 0.36,
564
+ "learning_rate": 6.698260869565218e-05,
565
+ "loss": 0.0946,
566
+ "step": 4300
567
+ },
568
+ {
569
+ "epoch": 0.36,
570
+ "learning_rate": 6.654782608695652e-05,
571
+ "loss": 0.0767,
572
+ "step": 4350
573
+ },
574
+ {
575
+ "epoch": 0.37,
576
+ "learning_rate": 6.611304347826088e-05,
577
+ "loss": 0.0753,
578
+ "step": 4400
579
+ },
580
+ {
581
+ "epoch": 0.37,
582
+ "learning_rate": 6.567826086956522e-05,
583
+ "loss": 0.0857,
584
+ "step": 4450
585
+ },
586
+ {
587
+ "epoch": 0.38,
588
+ "learning_rate": 6.524347826086957e-05,
589
+ "loss": 0.0887,
590
+ "step": 4500
591
+ },
592
+ {
593
+ "epoch": 0.38,
594
+ "learning_rate": 6.480869565217391e-05,
595
+ "loss": 0.0723,
596
+ "step": 4550
597
+ },
598
+ {
599
+ "epoch": 0.38,
600
+ "learning_rate": 6.437391304347827e-05,
601
+ "loss": 0.0952,
602
+ "step": 4600
603
+ },
604
+ {
605
+ "epoch": 0.39,
606
+ "learning_rate": 6.393913043478262e-05,
607
+ "loss": 0.0687,
608
+ "step": 4650
609
+ },
610
+ {
611
+ "epoch": 0.39,
612
+ "learning_rate": 6.350434782608696e-05,
613
+ "loss": 0.0967,
614
+ "step": 4700
615
+ },
616
+ {
617
+ "epoch": 0.4,
618
+ "learning_rate": 6.30695652173913e-05,
619
+ "loss": 0.0804,
620
+ "step": 4750
621
+ },
622
+ {
623
+ "epoch": 0.4,
624
+ "learning_rate": 6.263478260869565e-05,
625
+ "loss": 0.118,
626
+ "step": 4800
627
+ },
628
+ {
629
+ "epoch": 0.4,
630
+ "learning_rate": 6.220000000000001e-05,
631
+ "loss": 0.1065,
632
+ "step": 4850
633
+ },
634
+ {
635
+ "epoch": 0.41,
636
+ "learning_rate": 6.176521739130436e-05,
637
+ "loss": 0.1005,
638
+ "step": 4900
639
+ },
640
+ {
641
+ "epoch": 0.41,
642
+ "learning_rate": 6.13304347826087e-05,
643
+ "loss": 0.1045,
644
+ "step": 4950
645
+ },
646
+ {
647
+ "epoch": 0.42,
648
+ "learning_rate": 6.0895652173913044e-05,
649
+ "loss": 0.099,
650
+ "step": 5000
651
+ },
652
+ {
653
+ "epoch": 0.42,
654
+ "eval_loss": 0.10123514384031296,
655
+ "eval_runtime": 6887.4112,
656
+ "eval_samples_per_second": 2.304,
657
+ "eval_steps_per_second": 0.036,
658
+ "eval_wer": 10.61430392878921,
659
+ "step": 5000
660
+ },
661
+ {
662
+ "epoch": 0.42,
663
+ "learning_rate": 6.046086956521739e-05,
664
+ "loss": 0.1035,
665
+ "step": 5050
666
+ },
667
+ {
668
+ "epoch": 0.42,
669
+ "learning_rate": 6.002608695652174e-05,
670
+ "loss": 0.1015,
671
+ "step": 5100
672
+ },
673
+ {
674
+ "epoch": 0.43,
675
+ "learning_rate": 5.9591304347826095e-05,
676
+ "loss": 0.0909,
677
+ "step": 5150
678
+ },
679
+ {
680
+ "epoch": 0.43,
681
+ "learning_rate": 5.9156521739130436e-05,
682
+ "loss": 0.1056,
683
+ "step": 5200
684
+ },
685
+ {
686
+ "epoch": 0.44,
687
+ "learning_rate": 5.8721739130434784e-05,
688
+ "loss": 0.1031,
689
+ "step": 5250
690
+ },
691
+ {
692
+ "epoch": 0.44,
693
+ "learning_rate": 5.828695652173913e-05,
694
+ "loss": 0.0842,
695
+ "step": 5300
696
+ },
697
+ {
698
+ "epoch": 0.45,
699
+ "learning_rate": 5.7852173913043486e-05,
700
+ "loss": 0.0866,
701
+ "step": 5350
702
+ },
703
+ {
704
+ "epoch": 0.45,
705
+ "learning_rate": 5.7417391304347834e-05,
706
+ "loss": 0.0897,
707
+ "step": 5400
708
+ },
709
+ {
710
+ "epoch": 1.0,
711
+ "learning_rate": 5.6982608695652175e-05,
712
+ "loss": 0.067,
713
+ "step": 5450
714
+ },
715
+ {
716
+ "epoch": 1.01,
717
+ "learning_rate": 5.654782608695652e-05,
718
+ "loss": 0.0373,
719
+ "step": 5500
720
+ },
721
+ {
722
+ "epoch": 1.01,
723
+ "learning_rate": 5.611304347826087e-05,
724
+ "loss": 0.0368,
725
+ "step": 5550
726
+ },
727
+ {
728
+ "epoch": 1.01,
729
+ "learning_rate": 5.5678260869565226e-05,
730
+ "loss": 0.0372,
731
+ "step": 5600
732
+ },
733
+ {
734
+ "epoch": 1.02,
735
+ "learning_rate": 5.524347826086957e-05,
736
+ "loss": 0.0335,
737
+ "step": 5650
738
+ },
739
+ {
740
+ "epoch": 1.02,
741
+ "learning_rate": 5.4808695652173914e-05,
742
+ "loss": 0.0361,
743
+ "step": 5700
744
+ },
745
+ {
746
+ "epoch": 1.03,
747
+ "learning_rate": 5.437391304347826e-05,
748
+ "loss": 0.0345,
749
+ "step": 5750
750
+ },
751
+ {
752
+ "epoch": 1.03,
753
+ "learning_rate": 5.393913043478262e-05,
754
+ "loss": 0.0357,
755
+ "step": 5800
756
+ },
757
+ {
758
+ "epoch": 1.04,
759
+ "learning_rate": 5.3504347826086965e-05,
760
+ "loss": 0.0353,
761
+ "step": 5850
762
+ },
763
+ {
764
+ "epoch": 1.04,
765
+ "learning_rate": 5.3069565217391306e-05,
766
+ "loss": 0.0302,
767
+ "step": 5900
768
+ },
769
+ {
770
+ "epoch": 1.04,
771
+ "learning_rate": 5.2634782608695654e-05,
772
+ "loss": 0.0291,
773
+ "step": 5950
774
+ },
775
+ {
776
+ "epoch": 1.05,
777
+ "learning_rate": 5.22e-05,
778
+ "loss": 0.028,
779
+ "step": 6000
780
+ },
781
+ {
782
+ "epoch": 1.05,
783
+ "eval_loss": 0.1052699089050293,
784
+ "eval_runtime": 6861.1264,
785
+ "eval_samples_per_second": 2.313,
786
+ "eval_steps_per_second": 0.036,
787
+ "eval_wer": 9.879388309456495,
788
+ "step": 6000
789
+ },
790
+ {
791
+ "epoch": 1.05,
792
+ "learning_rate": 5.1765217391304356e-05,
793
+ "loss": 0.0341,
794
+ "step": 6050
795
+ },
796
+ {
797
+ "epoch": 1.06,
798
+ "learning_rate": 5.133043478260869e-05,
799
+ "loss": 0.058,
800
+ "step": 6100
801
+ },
802
+ {
803
+ "epoch": 1.06,
804
+ "learning_rate": 5.0895652173913045e-05,
805
+ "loss": 0.0644,
806
+ "step": 6150
807
+ },
808
+ {
809
+ "epoch": 1.06,
810
+ "learning_rate": 5.046086956521739e-05,
811
+ "loss": 0.0605,
812
+ "step": 6200
813
+ },
814
+ {
815
+ "epoch": 1.07,
816
+ "learning_rate": 5.002608695652175e-05,
817
+ "loss": 0.0509,
818
+ "step": 6250
819
+ },
820
+ {
821
+ "epoch": 1.07,
822
+ "learning_rate": 4.959130434782609e-05,
823
+ "loss": 0.0499,
824
+ "step": 6300
825
+ },
826
+ {
827
+ "epoch": 1.08,
828
+ "learning_rate": 4.915652173913044e-05,
829
+ "loss": 0.0505,
830
+ "step": 6350
831
+ },
832
+ {
833
+ "epoch": 1.08,
834
+ "learning_rate": 4.8721739130434785e-05,
835
+ "loss": 0.0574,
836
+ "step": 6400
837
+ },
838
+ {
839
+ "epoch": 1.09,
840
+ "learning_rate": 4.828695652173913e-05,
841
+ "loss": 0.0461,
842
+ "step": 6450
843
+ },
844
+ {
845
+ "epoch": 1.09,
846
+ "learning_rate": 4.785217391304348e-05,
847
+ "loss": 0.0473,
848
+ "step": 6500
849
+ },
850
+ {
851
+ "epoch": 1.09,
852
+ "learning_rate": 4.741739130434783e-05,
853
+ "loss": 0.049,
854
+ "step": 6550
855
+ },
856
+ {
857
+ "epoch": 1.1,
858
+ "learning_rate": 4.6982608695652176e-05,
859
+ "loss": 0.0507,
860
+ "step": 6600
861
+ },
862
+ {
863
+ "epoch": 1.1,
864
+ "learning_rate": 4.6547826086956524e-05,
865
+ "loss": 0.0481,
866
+ "step": 6650
867
+ },
868
+ {
869
+ "epoch": 1.11,
870
+ "learning_rate": 4.611304347826087e-05,
871
+ "loss": 0.0708,
872
+ "step": 6700
873
+ },
874
+ {
875
+ "epoch": 1.11,
876
+ "learning_rate": 4.567826086956522e-05,
877
+ "loss": 0.0758,
878
+ "step": 6750
879
+ },
880
+ {
881
+ "epoch": 1.11,
882
+ "learning_rate": 4.524347826086957e-05,
883
+ "loss": 0.051,
884
+ "step": 6800
885
+ },
886
+ {
887
+ "epoch": 1.12,
888
+ "learning_rate": 4.4808695652173915e-05,
889
+ "loss": 0.0557,
890
+ "step": 6850
891
+ },
892
+ {
893
+ "epoch": 1.12,
894
+ "learning_rate": 4.437391304347826e-05,
895
+ "loss": 0.0659,
896
+ "step": 6900
897
+ },
898
+ {
899
+ "epoch": 1.13,
900
+ "learning_rate": 4.393913043478261e-05,
901
+ "loss": 0.0795,
902
+ "step": 6950
903
+ },
904
+ {
905
+ "epoch": 1.13,
906
+ "learning_rate": 4.350434782608696e-05,
907
+ "loss": 0.0473,
908
+ "step": 7000
909
+ },
910
+ {
911
+ "epoch": 1.13,
912
+ "eval_loss": 0.10288114845752716,
913
+ "eval_runtime": 6910.216,
914
+ "eval_samples_per_second": 2.297,
915
+ "eval_steps_per_second": 0.036,
916
+ "eval_wer": 10.307825109089038,
917
+ "step": 7000
918
+ },
919
+ {
920
+ "epoch": 1.14,
921
+ "learning_rate": 4.306956521739131e-05,
922
+ "loss": 0.0691,
923
+ "step": 7050
924
+ },
925
+ {
926
+ "epoch": 1.14,
927
+ "learning_rate": 4.2634782608695655e-05,
928
+ "loss": 0.0436,
929
+ "step": 7100
930
+ },
931
+ {
932
+ "epoch": 1.14,
933
+ "learning_rate": 4.22e-05,
934
+ "loss": 0.0457,
935
+ "step": 7150
936
+ },
937
+ {
938
+ "epoch": 1.15,
939
+ "learning_rate": 4.176521739130435e-05,
940
+ "loss": 0.0455,
941
+ "step": 7200
942
+ },
943
+ {
944
+ "epoch": 1.15,
945
+ "learning_rate": 4.13304347826087e-05,
946
+ "loss": 0.0437,
947
+ "step": 7250
948
+ },
949
+ {
950
+ "epoch": 1.16,
951
+ "learning_rate": 4.0895652173913046e-05,
952
+ "loss": 0.0398,
953
+ "step": 7300
954
+ },
955
+ {
956
+ "epoch": 1.16,
957
+ "learning_rate": 4.0460869565217394e-05,
958
+ "loss": 0.0602,
959
+ "step": 7350
960
+ },
961
+ {
962
+ "epoch": 1.16,
963
+ "learning_rate": 4.002608695652174e-05,
964
+ "loss": 0.0653,
965
+ "step": 7400
966
+ },
967
+ {
968
+ "epoch": 1.17,
969
+ "learning_rate": 3.959130434782609e-05,
970
+ "loss": 0.0318,
971
+ "step": 7450
972
+ },
973
+ {
974
+ "epoch": 1.17,
975
+ "learning_rate": 3.915652173913044e-05,
976
+ "loss": 0.0295,
977
+ "step": 7500
978
+ },
979
+ {
980
+ "epoch": 1.18,
981
+ "learning_rate": 3.8721739130434785e-05,
982
+ "loss": 0.0442,
983
+ "step": 7550
984
+ },
985
+ {
986
+ "epoch": 1.18,
987
+ "learning_rate": 3.828695652173913e-05,
988
+ "loss": 0.0443,
989
+ "step": 7600
990
+ },
991
+ {
992
+ "epoch": 1.19,
993
+ "learning_rate": 3.785217391304348e-05,
994
+ "loss": 0.0351,
995
+ "step": 7650
996
+ },
997
+ {
998
+ "epoch": 1.19,
999
+ "learning_rate": 3.741739130434783e-05,
1000
+ "loss": 0.0447,
1001
+ "step": 7700
1002
+ },
1003
+ {
1004
+ "epoch": 1.19,
1005
+ "learning_rate": 3.698260869565218e-05,
1006
+ "loss": 0.0494,
1007
+ "step": 7750
1008
+ },
1009
+ {
1010
+ "epoch": 1.2,
1011
+ "learning_rate": 3.6547826086956525e-05,
1012
+ "loss": 0.0481,
1013
+ "step": 7800
1014
+ },
1015
+ {
1016
+ "epoch": 1.2,
1017
+ "learning_rate": 3.611304347826087e-05,
1018
+ "loss": 0.0404,
1019
+ "step": 7850
1020
+ },
1021
+ {
1022
+ "epoch": 1.21,
1023
+ "learning_rate": 3.5678260869565214e-05,
1024
+ "loss": 0.0482,
1025
+ "step": 7900
1026
+ },
1027
+ {
1028
+ "epoch": 1.21,
1029
+ "learning_rate": 3.524347826086957e-05,
1030
+ "loss": 0.0416,
1031
+ "step": 7950
1032
+ },
1033
+ {
1034
+ "epoch": 1.21,
1035
+ "learning_rate": 3.4808695652173916e-05,
1036
+ "loss": 0.0391,
1037
+ "step": 8000
1038
+ },
1039
+ {
1040
+ "epoch": 1.21,
1041
+ "eval_loss": 0.09238624572753906,
1042
+ "eval_runtime": 6861.6911,
1043
+ "eval_samples_per_second": 2.313,
1044
+ "eval_steps_per_second": 0.036,
1045
+ "eval_wer": 9.24188068708275,
1046
+ "step": 8000
1047
+ },
1048
+ {
1049
+ "epoch": 1.22,
1050
+ "learning_rate": 3.4373913043478264e-05,
1051
+ "loss": 0.0473,
1052
+ "step": 8050
1053
+ },
1054
+ {
1055
+ "epoch": 1.22,
1056
+ "learning_rate": 3.393913043478261e-05,
1057
+ "loss": 0.0423,
1058
+ "step": 8100
1059
+ },
1060
+ {
1061
+ "epoch": 1.23,
1062
+ "learning_rate": 3.350434782608696e-05,
1063
+ "loss": 0.0365,
1064
+ "step": 8150
1065
+ },
1066
+ {
1067
+ "epoch": 1.23,
1068
+ "learning_rate": 3.306956521739131e-05,
1069
+ "loss": 0.0428,
1070
+ "step": 8200
1071
+ },
1072
+ {
1073
+ "epoch": 1.24,
1074
+ "learning_rate": 3.263478260869565e-05,
1075
+ "loss": 0.0479,
1076
+ "step": 8250
1077
+ },
1078
+ {
1079
+ "epoch": 1.24,
1080
+ "learning_rate": 3.2200000000000003e-05,
1081
+ "loss": 0.0365,
1082
+ "step": 8300
1083
+ },
1084
+ {
1085
+ "epoch": 1.24,
1086
+ "learning_rate": 3.1765217391304345e-05,
1087
+ "loss": 0.035,
1088
+ "step": 8350
1089
+ },
1090
+ {
1091
+ "epoch": 1.25,
1092
+ "learning_rate": 3.13304347826087e-05,
1093
+ "loss": 0.0523,
1094
+ "step": 8400
1095
+ },
1096
+ {
1097
+ "epoch": 1.25,
1098
+ "learning_rate": 3.089565217391304e-05,
1099
+ "loss": 0.0602,
1100
+ "step": 8450
1101
+ },
1102
+ {
1103
+ "epoch": 1.26,
1104
+ "learning_rate": 3.046086956521739e-05,
1105
+ "loss": 0.0504,
1106
+ "step": 8500
1107
+ },
1108
+ {
1109
+ "epoch": 1.26,
1110
+ "learning_rate": 3.0026086956521743e-05,
1111
+ "loss": 0.05,
1112
+ "step": 8550
1113
+ },
1114
+ {
1115
+ "epoch": 1.26,
1116
+ "learning_rate": 2.9591304347826087e-05,
1117
+ "loss": 0.0579,
1118
+ "step": 8600
1119
+ },
1120
+ {
1121
+ "epoch": 1.27,
1122
+ "learning_rate": 2.915652173913044e-05,
1123
+ "loss": 0.0509,
1124
+ "step": 8650
1125
+ },
1126
+ {
1127
+ "epoch": 1.27,
1128
+ "learning_rate": 2.8721739130434783e-05,
1129
+ "loss": 0.0439,
1130
+ "step": 8700
1131
+ },
1132
+ {
1133
+ "epoch": 1.28,
1134
+ "learning_rate": 2.8286956521739134e-05,
1135
+ "loss": 0.0435,
1136
+ "step": 8750
1137
+ },
1138
+ {
1139
+ "epoch": 1.28,
1140
+ "learning_rate": 2.785217391304348e-05,
1141
+ "loss": 0.0419,
1142
+ "step": 8800
1143
+ },
1144
+ {
1145
+ "epoch": 1.29,
1146
+ "learning_rate": 2.7417391304347827e-05,
1147
+ "loss": 0.0615,
1148
+ "step": 8850
1149
+ },
1150
+ {
1151
+ "epoch": 1.29,
1152
+ "learning_rate": 2.6982608695652174e-05,
1153
+ "loss": 0.0404,
1154
+ "step": 8900
1155
+ },
1156
+ {
1157
+ "epoch": 1.29,
1158
+ "learning_rate": 2.6547826086956522e-05,
1159
+ "loss": 0.045,
1160
+ "step": 8950
1161
+ },
1162
+ {
1163
+ "epoch": 1.3,
1164
+ "learning_rate": 2.6113043478260874e-05,
1165
+ "loss": 0.0423,
1166
+ "step": 9000
1167
+ },
1168
+ {
1169
+ "epoch": 1.3,
1170
+ "eval_loss": 0.0796533077955246,
1171
+ "eval_runtime": 6857.0536,
1172
+ "eval_samples_per_second": 2.315,
1173
+ "eval_steps_per_second": 0.036,
1174
+ "eval_wer": 7.924892890800092,
1175
+ "step": 9000
1176
+ },
1177
+ {
1178
+ "epoch": 1.3,
1179
+ "learning_rate": 2.5678260869565218e-05,
1180
+ "loss": 0.031,
1181
+ "step": 9050
1182
+ },
1183
+ {
1184
+ "epoch": 1.31,
1185
+ "learning_rate": 2.524347826086957e-05,
1186
+ "loss": 0.0355,
1187
+ "step": 9100
1188
+ },
1189
+ {
1190
+ "epoch": 1.31,
1191
+ "learning_rate": 2.4808695652173914e-05,
1192
+ "loss": 0.0278,
1193
+ "step": 9150
1194
+ },
1195
+ {
1196
+ "epoch": 1.31,
1197
+ "learning_rate": 2.437391304347826e-05,
1198
+ "loss": 0.0295,
1199
+ "step": 9200
1200
+ },
1201
+ {
1202
+ "epoch": 1.32,
1203
+ "learning_rate": 2.3939130434782613e-05,
1204
+ "loss": 0.0323,
1205
+ "step": 9250
1206
+ },
1207
+ {
1208
+ "epoch": 1.32,
1209
+ "learning_rate": 2.3504347826086957e-05,
1210
+ "loss": 0.0313,
1211
+ "step": 9300
1212
+ },
1213
+ {
1214
+ "epoch": 1.33,
1215
+ "learning_rate": 2.3069565217391305e-05,
1216
+ "loss": 0.029,
1217
+ "step": 9350
1218
+ },
1219
+ {
1220
+ "epoch": 1.33,
1221
+ "learning_rate": 2.2634782608695653e-05,
1222
+ "loss": 0.0327,
1223
+ "step": 9400
1224
+ },
1225
+ {
1226
+ "epoch": 1.34,
1227
+ "learning_rate": 2.22e-05,
1228
+ "loss": 0.0246,
1229
+ "step": 9450
1230
+ },
1231
+ {
1232
+ "epoch": 1.34,
1233
+ "learning_rate": 2.176521739130435e-05,
1234
+ "loss": 0.0333,
1235
+ "step": 9500
1236
+ },
1237
+ {
1238
+ "epoch": 1.34,
1239
+ "learning_rate": 2.1330434782608697e-05,
1240
+ "loss": 0.031,
1241
+ "step": 9550
1242
+ },
1243
+ {
1244
+ "epoch": 1.35,
1245
+ "learning_rate": 2.0895652173913044e-05,
1246
+ "loss": 0.0471,
1247
+ "step": 9600
1248
+ },
1249
+ {
1250
+ "epoch": 1.35,
1251
+ "learning_rate": 2.0460869565217392e-05,
1252
+ "loss": 0.0725,
1253
+ "step": 9650
1254
+ },
1255
+ {
1256
+ "epoch": 1.36,
1257
+ "learning_rate": 2.002608695652174e-05,
1258
+ "loss": 0.0773,
1259
+ "step": 9700
1260
+ },
1261
+ {
1262
+ "epoch": 1.36,
1263
+ "learning_rate": 1.9591304347826088e-05,
1264
+ "loss": 0.0616,
1265
+ "step": 9750
1266
+ },
1267
+ {
1268
+ "epoch": 1.36,
1269
+ "learning_rate": 1.9156521739130436e-05,
1270
+ "loss": 0.0698,
1271
+ "step": 9800
1272
+ },
1273
+ {
1274
+ "epoch": 1.37,
1275
+ "learning_rate": 1.8721739130434784e-05,
1276
+ "loss": 0.0631,
1277
+ "step": 9850
1278
+ },
1279
+ {
1280
+ "epoch": 1.37,
1281
+ "learning_rate": 1.828695652173913e-05,
1282
+ "loss": 0.0592,
1283
+ "step": 9900
1284
+ },
1285
+ {
1286
+ "epoch": 1.38,
1287
+ "learning_rate": 1.785217391304348e-05,
1288
+ "loss": 0.0583,
1289
+ "step": 9950
1290
+ },
1291
+ {
1292
+ "epoch": 1.38,
1293
+ "learning_rate": 1.7417391304347827e-05,
1294
+ "loss": 0.0604,
1295
+ "step": 10000
1296
+ },
1297
+ {
1298
+ "epoch": 1.38,
1299
+ "eval_loss": 0.06884036958217621,
1300
+ "eval_runtime": 6836.4142,
1301
+ "eval_samples_per_second": 2.322,
1302
+ "eval_steps_per_second": 0.036,
1303
+ "eval_wer": 7.014959650915081,
1304
+ "step": 10000
1305
+ },
1306
+ {
1307
+ "epoch": 1.39,
1308
+ "learning_rate": 1.6982608695652175e-05,
1309
+ "loss": 0.0542,
1310
+ "step": 10050
1311
+ },
1312
+ {
1313
+ "epoch": 1.39,
1314
+ "learning_rate": 1.6556521739130434e-05,
1315
+ "loss": 0.0485,
1316
+ "step": 10100
1317
+ },
1318
+ {
1319
+ "epoch": 1.39,
1320
+ "learning_rate": 1.612173913043478e-05,
1321
+ "loss": 0.0483,
1322
+ "step": 10150
1323
+ },
1324
+ {
1325
+ "epoch": 1.4,
1326
+ "learning_rate": 1.5686956521739133e-05,
1327
+ "loss": 0.0484,
1328
+ "step": 10200
1329
+ },
1330
+ {
1331
+ "epoch": 1.4,
1332
+ "learning_rate": 1.5252173913043479e-05,
1333
+ "loss": 0.0484,
1334
+ "step": 10250
1335
+ },
1336
+ {
1337
+ "epoch": 1.41,
1338
+ "learning_rate": 1.4817391304347827e-05,
1339
+ "loss": 0.0385,
1340
+ "step": 10300
1341
+ },
1342
+ {
1343
+ "epoch": 1.41,
1344
+ "learning_rate": 1.4382608695652175e-05,
1345
+ "loss": 0.0295,
1346
+ "step": 10350
1347
+ },
1348
+ {
1349
+ "epoch": 1.41,
1350
+ "learning_rate": 1.3947826086956523e-05,
1351
+ "loss": 0.035,
1352
+ "step": 10400
1353
+ },
1354
+ {
1355
+ "epoch": 1.42,
1356
+ "learning_rate": 1.351304347826087e-05,
1357
+ "loss": 0.028,
1358
+ "step": 10450
1359
+ },
1360
+ {
1361
+ "epoch": 1.42,
1362
+ "learning_rate": 1.3078260869565218e-05,
1363
+ "loss": 0.0436,
1364
+ "step": 10500
1365
+ },
1366
+ {
1367
+ "epoch": 1.43,
1368
+ "learning_rate": 1.2643478260869565e-05,
1369
+ "loss": 0.0334,
1370
+ "step": 10550
1371
+ },
1372
+ {
1373
+ "epoch": 1.43,
1374
+ "learning_rate": 1.2208695652173914e-05,
1375
+ "loss": 0.032,
1376
+ "step": 10600
1377
+ },
1378
+ {
1379
+ "epoch": 1.44,
1380
+ "learning_rate": 1.1773913043478262e-05,
1381
+ "loss": 0.0324,
1382
+ "step": 10650
1383
+ },
1384
+ {
1385
+ "epoch": 1.44,
1386
+ "learning_rate": 1.133913043478261e-05,
1387
+ "loss": 0.0309,
1388
+ "step": 10700
1389
+ },
1390
+ {
1391
+ "epoch": 1.44,
1392
+ "learning_rate": 1.0904347826086956e-05,
1393
+ "loss": 0.0269,
1394
+ "step": 10750
1395
+ },
1396
+ {
1397
+ "epoch": 1.45,
1398
+ "learning_rate": 1.0469565217391304e-05,
1399
+ "loss": 0.0341,
1400
+ "step": 10800
1401
+ },
1402
+ {
1403
+ "epoch": 1.45,
1404
+ "learning_rate": 1.0043478260869566e-05,
1405
+ "loss": 0.0326,
1406
+ "step": 10850
1407
+ },
1408
+ {
1409
+ "epoch": 2.0,
1410
+ "learning_rate": 9.608695652173914e-06,
1411
+ "loss": 0.0175,
1412
+ "step": 10900
1413
+ },
1414
+ {
1415
+ "epoch": 2.01,
1416
+ "learning_rate": 9.173913043478261e-06,
1417
+ "loss": 0.0117,
1418
+ "step": 10950
1419
+ },
1420
+ {
1421
+ "epoch": 2.01,
1422
+ "learning_rate": 8.73913043478261e-06,
1423
+ "loss": 0.0121,
1424
+ "step": 11000
1425
+ },
1426
+ {
1427
+ "epoch": 2.01,
1428
+ "eval_loss": 0.06960244476795197,
1429
+ "eval_runtime": 6836.8996,
1430
+ "eval_samples_per_second": 2.322,
1431
+ "eval_steps_per_second": 0.036,
1432
+ "eval_wer": 6.463772936415545,
1433
+ "step": 11000
1434
+ },
1435
+ {
1436
+ "epoch": 2.02,
1437
+ "learning_rate": 8.304347826086957e-06,
1438
+ "loss": 0.0119,
1439
+ "step": 11050
1440
+ },
1441
+ {
1442
+ "epoch": 2.02,
1443
+ "learning_rate": 7.869565217391305e-06,
1444
+ "loss": 0.0163,
1445
+ "step": 11100
1446
+ },
1447
+ {
1448
+ "epoch": 2.02,
1449
+ "learning_rate": 7.434782608695652e-06,
1450
+ "loss": 0.0136,
1451
+ "step": 11150
1452
+ },
1453
+ {
1454
+ "epoch": 2.03,
1455
+ "learning_rate": 7.000000000000001e-06,
1456
+ "loss": 0.0124,
1457
+ "step": 11200
1458
+ },
1459
+ {
1460
+ "epoch": 2.03,
1461
+ "learning_rate": 6.565217391304349e-06,
1462
+ "loss": 0.0124,
1463
+ "step": 11250
1464
+ },
1465
+ {
1466
+ "epoch": 2.04,
1467
+ "learning_rate": 6.1304347826086965e-06,
1468
+ "loss": 0.0135,
1469
+ "step": 11300
1470
+ },
1471
+ {
1472
+ "epoch": 2.04,
1473
+ "learning_rate": 5.6956521739130435e-06,
1474
+ "loss": 0.0111,
1475
+ "step": 11350
1476
+ },
1477
+ {
1478
+ "epoch": 2.05,
1479
+ "learning_rate": 5.260869565217392e-06,
1480
+ "loss": 0.0128,
1481
+ "step": 11400
1482
+ },
1483
+ {
1484
+ "epoch": 2.05,
1485
+ "learning_rate": 4.826086956521739e-06,
1486
+ "loss": 0.0142,
1487
+ "step": 11450
1488
+ },
1489
+ {
1490
+ "epoch": 2.05,
1491
+ "learning_rate": 4.391304347826087e-06,
1492
+ "loss": 0.0136,
1493
+ "step": 11500
1494
+ },
1495
+ {
1496
+ "epoch": 2.06,
1497
+ "learning_rate": 3.956521739130435e-06,
1498
+ "loss": 0.0157,
1499
+ "step": 11550
1500
+ },
1501
+ {
1502
+ "epoch": 2.06,
1503
+ "learning_rate": 3.521739130434783e-06,
1504
+ "loss": 0.0154,
1505
+ "step": 11600
1506
+ },
1507
+ {
1508
+ "epoch": 2.07,
1509
+ "learning_rate": 3.0869565217391307e-06,
1510
+ "loss": 0.016,
1511
+ "step": 11650
1512
+ },
1513
+ {
1514
+ "epoch": 2.07,
1515
+ "learning_rate": 2.6521739130434785e-06,
1516
+ "loss": 0.0146,
1517
+ "step": 11700
1518
+ },
1519
+ {
1520
+ "epoch": 2.08,
1521
+ "learning_rate": 2.217391304347826e-06,
1522
+ "loss": 0.0133,
1523
+ "step": 11750
1524
+ },
1525
+ {
1526
+ "epoch": 2.08,
1527
+ "learning_rate": 1.7826086956521738e-06,
1528
+ "loss": 0.0142,
1529
+ "step": 11800
1530
+ },
1531
+ {
1532
+ "epoch": 2.08,
1533
+ "learning_rate": 1.3478260869565217e-06,
1534
+ "loss": 0.0155,
1535
+ "step": 11850
1536
+ },
1537
+ {
1538
+ "epoch": 2.09,
1539
+ "learning_rate": 9.130434782608697e-07,
1540
+ "loss": 0.012,
1541
+ "step": 11900
1542
+ },
1543
+ {
1544
+ "epoch": 2.09,
1545
+ "learning_rate": 4.782608695652174e-07,
1546
+ "loss": 0.0149,
1547
+ "step": 11950
1548
+ },
1549
+ {
1550
+ "epoch": 2.1,
1551
+ "learning_rate": 4.347826086956522e-08,
1552
+ "loss": 0.0155,
1553
+ "step": 12000
1554
+ },
1555
+ {
1556
+ "epoch": 2.1,
1557
+ "eval_loss": 0.07060841470956802,
1558
+ "eval_runtime": 6855.2466,
1559
+ "eval_samples_per_second": 2.315,
1560
+ "eval_steps_per_second": 0.036,
1561
+ "eval_wer": 6.3671568743912,
1562
+ "step": 12000
1563
+ },
1564
+ {
1565
+ "epoch": 2.1,
1566
+ "step": 12000,
1567
+ "total_flos": 2.2160934627213312e+20,
1568
+ "train_loss": 0.08807742363711198,
1569
+ "train_runtime": 202754.3941,
1570
+ "train_samples_per_second": 3.788,
1571
+ "train_steps_per_second": 0.059
1572
+ }
1573
+ ],
1574
+ "max_steps": 12000,
1575
+ "num_train_epochs": 9223372036854775807,
1576
+ "total_flos": 2.2160934627213312e+20,
1577
+ "trial_name": null,
1578
+ "trial_params": null
1579
+ }