Hugging Face's logo

GENG
/
hubert_ls_4500

Automatic Speech Recognition
Transformers
PyTorch
hubert
Inference Endpoints
Model card Files Community
GENG commited on
Commit
b05cc26
•
1 Parent(s): 5d40bb4

Upload trainer_state.json

Browse files
Files changed (1) hide show
  1. trainer_state.json +151 -0
trainer_state.json ADDED
@@ -0,0 +1,151 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 31.46853146853147,
5
+ "global_step": 4500,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 3.5,
12
+ "learning_rate": 4.9800000000000004e-05,
13
+ "loss": 4.6313,
14
+ "step": 500
15
+ },
16
+ {
17
+ "epoch": 3.5,
18
+ "eval_cer": 0.9993038029762423,
19
+ "eval_loss": 3.357034921646118,
20
+ "eval_runtime": 144.6064,
21
+ "eval_samples_per_second": 7.897,
22
+ "eval_steps_per_second": 0.989,
23
+ "step": 500
24
+ },
25
+ {
26
+ "epoch": 6.99,
27
+ "learning_rate": 9.98e-05,
28
+ "loss": 2.8906,
29
+ "step": 1000
30
+ },
31
+ {
32
+ "epoch": 6.99,
33
+ "eval_cer": 0.1851720912018101,
34
+ "eval_loss": 0.7587499022483826,
35
+ "eval_runtime": 144.5652,
36
+ "eval_samples_per_second": 7.9,
37
+ "eval_steps_per_second": 0.989,
38
+ "step": 1000
39
+ },
40
+ {
41
+ "epoch": 10.49,
42
+ "learning_rate": 9.190243902439025e-05,
43
+ "loss": 0.4635,
44
+ "step": 1500
45
+ },
46
+ {
47
+ "epoch": 10.49,
48
+ "eval_cer": 0.04839657122965799,
49
+ "eval_loss": 0.18757499754428864,
50
+ "eval_runtime": 146.4828,
51
+ "eval_samples_per_second": 7.796,
52
+ "eval_steps_per_second": 0.976,
53
+ "step": 1500
54
+ },
55
+ {
56
+ "epoch": 13.99,
57
+ "learning_rate": 8.377235772357724e-05,
58
+ "loss": 0.218,
59
+ "step": 2000
60
+ },
61
+ {
62
+ "epoch": 13.99,
63
+ "eval_cer": 0.03471738752066835,
64
+ "eval_loss": 0.14170190691947937,
65
+ "eval_runtime": 145.5208,
66
+ "eval_samples_per_second": 7.848,
67
+ "eval_steps_per_second": 0.983,
68
+ "step": 2000
69
+ },
70
+ {
71
+ "epoch": 17.48,
72
+ "learning_rate": 7.564227642276424e-05,
73
+ "loss": 0.1514,
74
+ "step": 2500
75
+ },
76
+ {
77
+ "epoch": 17.48,
78
+ "eval_cer": 0.032465625271951966,
79
+ "eval_loss": 0.13693825900554657,
80
+ "eval_runtime": 145.4673,
81
+ "eval_samples_per_second": 7.851,
82
+ "eval_steps_per_second": 0.983,
83
+ "step": 2500
84
+ },
85
+ {
86
+ "epoch": 20.98,
87
+ "learning_rate": 6.751219512195123e-05,
88
+ "loss": 0.1149,
89
+ "step": 3000
90
+ },
91
+ {
92
+ "epoch": 20.98,
93
+ "eval_cer": 0.02969715429466539,
94
+ "eval_loss": 0.13597196340560913,
95
+ "eval_runtime": 145.3558,
96
+ "eval_samples_per_second": 7.857,
97
+ "eval_steps_per_second": 0.984,
98
+ "step": 3000
99
+ },
100
+ {
101
+ "epoch": 24.48,
102
+ "learning_rate": 5.938211382113822e-05,
103
+ "loss": 0.1253,
104
+ "step": 3500
105
+ },
106
+ {
107
+ "epoch": 24.48,
108
+ "eval_cer": 0.029827691236619963,
109
+ "eval_loss": 0.1414954513311386,
110
+ "eval_runtime": 146.3702,
111
+ "eval_samples_per_second": 7.802,
112
+ "eval_steps_per_second": 0.977,
113
+ "step": 3500
114
+ },
115
+ {
116
+ "epoch": 27.97,
117
+ "learning_rate": 5.125203252032521e-05,
118
+ "loss": 0.1019,
119
+ "step": 4000
120
+ },
121
+ {
122
+ "epoch": 27.97,
123
+ "eval_cer": 0.028647419719780697,
124
+ "eval_loss": 0.1363374888896942,
125
+ "eval_runtime": 145.0727,
126
+ "eval_samples_per_second": 7.872,
127
+ "eval_steps_per_second": 0.986,
128
+ "step": 4000
129
+ },
130
+ {
131
+ "epoch": 31.47,
132
+ "learning_rate": 4.31219512195122e-05,
133
+ "loss": 0.0806,
134
+ "step": 4500
135
+ },
136
+ {
137
+ "epoch": 31.47,
138
+ "eval_cer": 0.02749434339918197,
139
+ "eval_loss": 0.13236147165298462,
140
+ "eval_runtime": 145.7007,
141
+ "eval_samples_per_second": 7.838,
142
+ "eval_steps_per_second": 0.981,
143
+ "step": 4500
144
+ }
145
+ ],
146
+ "max_steps": 7150,
147
+ "num_train_epochs": 50,
148
+ "total_flos": 1.6733035055952587e+19,
149
+ "trial_name": null,
150
+ "trial_params": null
151
+ }