infinitejoy commited on
Commit
9352f65
1 Parent(s): 75ecf49

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.19587931036949158,
4
+ "eval_runtime": 168.5422,
5
+ "eval_samples": 4686,
6
+ "eval_samples_per_second": 27.803,
7
+ "eval_steps_per_second": 1.738,
8
+ "eval_wer": 0.24543618426163288,
9
+ "train_loss": 1.2943066022937424,
10
+ "train_runtime": 28854.2781,
11
+ "train_samples": 13247,
12
+ "train_samples_per_second": 22.955,
13
+ "train_steps_per_second": 0.717
14
+ }
eval_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "eval_loss": 0.19587931036949158,
4
+ "eval_runtime": 168.5422,
5
+ "eval_samples": 4686,
6
+ "eval_samples_per_second": 27.803,
7
+ "eval_steps_per_second": 1.738,
8
+ "eval_wer": 0.24543618426163288
9
+ }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f4a20a063e02613c3a6229c458c2b336acaa3c046e998b5a3f3ecf0bd5146c5e
3
  size 1262145073
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:25f144236d73b276af60997ba3e53e23dab0fca4dc081a995b0a6dfef078e4dd
3
  size 1262145073
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 50.0,
3
+ "train_loss": 1.2943066022937424,
4
+ "train_runtime": 28854.2781,
5
+ "train_samples": 13247,
6
+ "train_samples_per_second": 22.955,
7
+ "train_steps_per_second": 0.717
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 50.0,
5
+ "global_step": 20700,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 2.42,
12
+ "learning_rate": 1.74475e-05,
13
+ "loss": 7.0856,
14
+ "step": 1000
15
+ },
16
+ {
17
+ "epoch": 4.83,
18
+ "learning_rate": 3.49475e-05,
19
+ "loss": 2.771,
20
+ "step": 2000
21
+ },
22
+ {
23
+ "epoch": 7.25,
24
+ "learning_rate": 5.243e-05,
25
+ "loss": 1.4051,
26
+ "step": 3000
27
+ },
28
+ {
29
+ "epoch": 9.66,
30
+ "learning_rate": 6.99125e-05,
31
+ "loss": 1.173,
32
+ "step": 4000
33
+ },
34
+ {
35
+ "epoch": 9.66,
36
+ "eval_loss": 0.29204532504081726,
37
+ "eval_runtime": 168.5565,
38
+ "eval_samples_per_second": 27.801,
39
+ "eval_steps_per_second": 1.738,
40
+ "eval_wer": 0.36075506874854346,
41
+ "step": 4000
42
+ },
43
+ {
44
+ "epoch": 12.08,
45
+ "learning_rate": 6.582934131736526e-05,
46
+ "loss": 1.082,
47
+ "step": 5000
48
+ },
49
+ {
50
+ "epoch": 14.49,
51
+ "learning_rate": 6.164191616766467e-05,
52
+ "loss": 1.0256,
53
+ "step": 6000
54
+ },
55
+ {
56
+ "epoch": 16.91,
57
+ "learning_rate": 5.74502994011976e-05,
58
+ "loss": 0.9779,
59
+ "step": 7000
60
+ },
61
+ {
62
+ "epoch": 19.32,
63
+ "learning_rate": 5.3262874251497e-05,
64
+ "loss": 0.9433,
65
+ "step": 8000
66
+ },
67
+ {
68
+ "epoch": 19.32,
69
+ "eval_loss": 0.23364943265914917,
70
+ "eval_runtime": 167.8504,
71
+ "eval_samples_per_second": 27.918,
72
+ "eval_steps_per_second": 1.746,
73
+ "eval_wer": 0.30261011419249595,
74
+ "step": 8000
75
+ },
76
+ {
77
+ "epoch": 21.74,
78
+ "learning_rate": 4.907125748502993e-05,
79
+ "loss": 0.9211,
80
+ "step": 9000
81
+ },
82
+ {
83
+ "epoch": 24.15,
84
+ "learning_rate": 4.48880239520958e-05,
85
+ "loss": 0.8969,
86
+ "step": 10000
87
+ },
88
+ {
89
+ "epoch": 26.57,
90
+ "learning_rate": 4.069640718562874e-05,
91
+ "loss": 0.8727,
92
+ "step": 11000
93
+ },
94
+ {
95
+ "epoch": 28.99,
96
+ "learning_rate": 3.6504790419161677e-05,
97
+ "loss": 0.8552,
98
+ "step": 12000
99
+ },
100
+ {
101
+ "epoch": 28.99,
102
+ "eval_loss": 0.22206313908100128,
103
+ "eval_runtime": 169.6004,
104
+ "eval_samples_per_second": 27.63,
105
+ "eval_steps_per_second": 1.728,
106
+ "eval_wer": 0.27988813796317874,
107
+ "step": 12000
108
+ },
109
+ {
110
+ "epoch": 31.4,
111
+ "learning_rate": 3.231317365269461e-05,
112
+ "loss": 0.8401,
113
+ "step": 13000
114
+ },
115
+ {
116
+ "epoch": 33.82,
117
+ "learning_rate": 2.8121556886227542e-05,
118
+ "loss": 0.8178,
119
+ "step": 14000
120
+ },
121
+ {
122
+ "epoch": 36.23,
123
+ "learning_rate": 2.3934131736526942e-05,
124
+ "loss": 0.8002,
125
+ "step": 15000
126
+ },
127
+ {
128
+ "epoch": 38.65,
129
+ "learning_rate": 1.974251497005988e-05,
130
+ "loss": 0.7863,
131
+ "step": 16000
132
+ },
133
+ {
134
+ "epoch": 38.65,
135
+ "eval_loss": 0.1952943205833435,
136
+ "eval_runtime": 169.2585,
137
+ "eval_samples_per_second": 27.685,
138
+ "eval_steps_per_second": 1.731,
139
+ "eval_wer": 0.24792200730210517,
140
+ "step": 16000
141
+ },
142
+ {
143
+ "epoch": 41.06,
144
+ "learning_rate": 1.555508982035928e-05,
145
+ "loss": 0.778,
146
+ "step": 17000
147
+ },
148
+ {
149
+ "epoch": 43.48,
150
+ "learning_rate": 1.1367664670658683e-05,
151
+ "loss": 0.7607,
152
+ "step": 18000
153
+ },
154
+ {
155
+ "epoch": 45.89,
156
+ "learning_rate": 7.180239520958083e-06,
157
+ "loss": 0.7481,
158
+ "step": 19000
159
+ },
160
+ {
161
+ "epoch": 48.31,
162
+ "learning_rate": 2.988622754491018e-06,
163
+ "loss": 0.7365,
164
+ "step": 20000
165
+ },
166
+ {
167
+ "epoch": 48.31,
168
+ "eval_loss": 0.1967712938785553,
169
+ "eval_runtime": 167.9953,
170
+ "eval_samples_per_second": 27.894,
171
+ "eval_steps_per_second": 1.744,
172
+ "eval_wer": 0.24489241047152957,
173
+ "step": 20000
174
+ },
175
+ {
176
+ "epoch": 50.0,
177
+ "step": 20700,
178
+ "total_flos": 7.668153416081444e+19,
179
+ "train_loss": 1.2943066022937424,
180
+ "train_runtime": 28854.2781,
181
+ "train_samples_per_second": 22.955,
182
+ "train_steps_per_second": 0.717
183
+ }
184
+ ],
185
+ "max_steps": 20700,
186
+ "num_train_epochs": 50,
187
+ "total_flos": 7.668153416081444e+19,
188
+ "trial_name": null,
189
+ "trial_params": null
190
+ }