marinone94 commited on
Commit
bb89fa3
β€’
1 Parent(s): 7d31b92

Model save

Browse files
Files changed (30) hide show
  1. checkpoint-300/scaler.pt +0 -3
  2. checkpoint-300/trainer_state.json +0 -133
  3. {checkpoint-300 β†’ checkpoint-3600}/config.json +0 -0
  4. {checkpoint-300 β†’ checkpoint-3600}/optimizer.pt +1 -1
  5. {checkpoint-300 β†’ checkpoint-3600}/preprocessor_config.json +0 -0
  6. {checkpoint-400 β†’ checkpoint-3600}/pytorch_model.bin +1 -1
  7. {checkpoint-500 β†’ checkpoint-3600}/rng_state.pth +1 -1
  8. {checkpoint-500 β†’ checkpoint-3600}/scaler.pt +1 -1
  9. {checkpoint-300 β†’ checkpoint-3600}/scheduler.pt +1 -1
  10. checkpoint-3600/trainer_state.json +1420 -0
  11. {checkpoint-300 β†’ checkpoint-3600}/training_args.bin +0 -0
  12. {checkpoint-400 β†’ checkpoint-3700}/config.json +0 -0
  13. {checkpoint-400 β†’ checkpoint-3700}/optimizer.pt +1 -1
  14. {checkpoint-400 β†’ checkpoint-3700}/preprocessor_config.json +0 -0
  15. {checkpoint-300 β†’ checkpoint-3700}/pytorch_model.bin +1 -1
  16. {checkpoint-300 β†’ checkpoint-3700}/rng_state.pth +1 -1
  17. {checkpoint-400 β†’ checkpoint-3700}/scaler.pt +1 -1
  18. {checkpoint-500 β†’ checkpoint-3700}/scheduler.pt +1 -1
  19. checkpoint-3700/trainer_state.json +1459 -0
  20. {checkpoint-400 β†’ checkpoint-3700}/training_args.bin +0 -0
  21. checkpoint-400/rng_state.pth +0 -3
  22. checkpoint-400/scheduler.pt +0 -3
  23. checkpoint-400/trainer_state.json +0 -172
  24. checkpoint-500/config.json +0 -115
  25. checkpoint-500/optimizer.pt +0 -3
  26. checkpoint-500/preprocessor_config.json +0 -9
  27. checkpoint-500/pytorch_model.bin +0 -3
  28. checkpoint-500/trainer_state.json +0 -211
  29. checkpoint-500/training_args.bin +0 -3
  30. pytorch_model.bin +1 -1
checkpoint-300/scaler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:7369b8dc53151ba1e5e48a8b502e6cab11dd93c5bc6715f57c74cbb1e95c4f54
3
- size 559
 
 
 
 
checkpoint-300/trainer_state.json DELETED
@@ -1,133 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 0.16014947284131856,
5
- "global_step": 300,
6
- "is_hyper_param_search": false,
7
- "is_local_process_zero": true,
8
- "is_world_process_zero": true,
9
- "log_history": [
10
- {
11
- "epoch": 0.01,
12
- "learning_rate": 0.0002,
13
- "loss": 10.6012,
14
- "step": 20
15
- },
16
- {
17
- "epoch": 0.02,
18
- "learning_rate": 0.0004,
19
- "loss": 3.651,
20
- "step": 40
21
- },
22
- {
23
- "epoch": 0.03,
24
- "learning_rate": 0.00059,
25
- "loss": 3.3911,
26
- "step": 60
27
- },
28
- {
29
- "epoch": 0.04,
30
- "learning_rate": 0.0007491827839825661,
31
- "loss": 3.0481,
32
- "step": 80
33
- },
34
- {
35
- "epoch": 0.05,
36
- "learning_rate": 0.0007453010078997548,
37
- "loss": 3.4039,
38
- "step": 100
39
- },
40
- {
41
- "epoch": 0.05,
42
- "eval_loss": Infinity,
43
- "eval_runtime": 1260.1628,
44
- "eval_samples_per_second": 21.123,
45
- "eval_steps_per_second": 0.66,
46
- "eval_wer": 1.0,
47
- "step": 100
48
- },
49
- {
50
- "epoch": 0.06,
51
- "learning_rate": 0.0007412149278125851,
52
- "loss": 3.0553,
53
- "step": 120
54
- },
55
- {
56
- "epoch": 0.07,
57
- "learning_rate": 0.0007371288477254155,
58
- "loss": 3.2191,
59
- "step": 140
60
- },
61
- {
62
- "epoch": 0.09,
63
- "learning_rate": 0.0007332470716426042,
64
- "loss": 3.4094,
65
- "step": 160
66
- },
67
- {
68
- "epoch": 0.1,
69
- "learning_rate": 0.0007291609915554345,
70
- "loss": 3.0607,
71
- "step": 180
72
- },
73
- {
74
- "epoch": 0.11,
75
- "learning_rate": 0.0007252792154726233,
76
- "loss": 3.4396,
77
- "step": 200
78
- },
79
- {
80
- "epoch": 0.11,
81
- "eval_loss": Infinity,
82
- "eval_runtime": 1235.1686,
83
- "eval_samples_per_second": 21.55,
84
- "eval_steps_per_second": 0.674,
85
- "eval_wer": 1.0,
86
- "step": 200
87
- },
88
- {
89
- "epoch": 0.12,
90
- "learning_rate": 0.0007211931353854536,
91
- "loss": 3.1198,
92
- "step": 220
93
- },
94
- {
95
- "epoch": 0.13,
96
- "learning_rate": 0.0007171070552982838,
97
- "loss": 3.1801,
98
- "step": 240
99
- },
100
- {
101
- "epoch": 0.14,
102
- "learning_rate": 0.0007132252792154726,
103
- "loss": 3.3954,
104
- "step": 260
105
- },
106
- {
107
- "epoch": 0.15,
108
- "learning_rate": 0.000709139199128303,
109
- "loss": 3.0566,
110
- "step": 280
111
- },
112
- {
113
- "epoch": 0.16,
114
- "learning_rate": 0.0007052574230454917,
115
- "loss": 3.483,
116
- "step": 300
117
- },
118
- {
119
- "epoch": 0.16,
120
- "eval_loss": Infinity,
121
- "eval_runtime": 1262.2253,
122
- "eval_samples_per_second": 21.088,
123
- "eval_steps_per_second": 0.659,
124
- "eval_wer": 1.0,
125
- "step": 300
126
- }
127
- ],
128
- "max_steps": 3746,
129
- "num_train_epochs": 2,
130
- "total_flos": 5.78379652841472e+18,
131
- "trial_name": null,
132
- "trial_params": null
133
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
{checkpoint-300 β†’ checkpoint-3600}/config.json RENAMED
File without changes
{checkpoint-300 β†’ checkpoint-3600}/optimizer.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:88bc51cc836c3951a88666eb2d2f418c2f696cb44d10c8300bca59c23334fcef
3
  size 2490337809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:23c5f7bb4f84d734d3656c47571c43ea944a272c13718be4a13e253457771f9d
3
  size 2490337809
{checkpoint-300 β†’ checkpoint-3600}/preprocessor_config.json RENAMED
File without changes
{checkpoint-400 β†’ checkpoint-3600}/pytorch_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:194566e34d57488248b1919886c65f8deca88913eab75e722c68d42a68701e31
3
  size 1262063089
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e95c67d08a1a0d2f34fb48babadb769336f0b21eeae723abbb0714176642fdf
3
  size 1262063089
{checkpoint-500 β†’ checkpoint-3600}/rng_state.pth RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:15485b7eb0d45ca11031ac084d04c8498c568293a5191feb9da6f84ff0bed287
3
  size 14567
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e58a6e31480cc85142d7450734a3e86737cc05e8a891ca912b4f46401282f9b
3
  size 14567
{checkpoint-500 β†’ checkpoint-3600}/scaler.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ed08b6959607e1ceba4db69e588fe44dbde3a530951e2ddbd01ef86b2a48da44
3
  size 559
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ce746b1527f3fc7f37eb070e5591bb9dc30314d4af77a55668c323725bd3836
3
  size 559
{checkpoint-300 β†’ checkpoint-3600}/scheduler.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fecb8f8e10eeacd02e5cba6adfff5885f11cbd3916997c6e20535764844df005
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:729533b728ab49b2fc48188304555e6065671b6ebd68bbb26d0f9a3555e530ec
3
  size 623
checkpoint-3600/trainer_state.json ADDED
@@ -0,0 +1,1420 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.9219271319898572,
5
+ "global_step": 3600,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 0.0002,
13
+ "loss": 10.6012,
14
+ "step": 20
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 0.0004,
19
+ "loss": 3.651,
20
+ "step": 40
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 0.00059,
25
+ "loss": 3.3911,
26
+ "step": 60
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 0.0007491827839825661,
31
+ "loss": 3.0481,
32
+ "step": 80
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 0.0007453010078997548,
37
+ "loss": 3.4039,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.05,
42
+ "eval_loss": Infinity,
43
+ "eval_runtime": 1260.1628,
44
+ "eval_samples_per_second": 21.123,
45
+ "eval_steps_per_second": 0.66,
46
+ "eval_wer": 1.0,
47
+ "step": 100
48
+ },
49
+ {
50
+ "epoch": 0.06,
51
+ "learning_rate": 0.0007412149278125851,
52
+ "loss": 3.0553,
53
+ "step": 120
54
+ },
55
+ {
56
+ "epoch": 0.07,
57
+ "learning_rate": 0.0007371288477254155,
58
+ "loss": 3.2191,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.09,
63
+ "learning_rate": 0.0007332470716426042,
64
+ "loss": 3.4094,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.1,
69
+ "learning_rate": 0.0007291609915554345,
70
+ "loss": 3.0607,
71
+ "step": 180
72
+ },
73
+ {
74
+ "epoch": 0.11,
75
+ "learning_rate": 0.0007252792154726233,
76
+ "loss": 3.4396,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 0.11,
81
+ "eval_loss": Infinity,
82
+ "eval_runtime": 1235.1686,
83
+ "eval_samples_per_second": 21.55,
84
+ "eval_steps_per_second": 0.674,
85
+ "eval_wer": 1.0,
86
+ "step": 200
87
+ },
88
+ {
89
+ "epoch": 0.12,
90
+ "learning_rate": 0.0007211931353854536,
91
+ "loss": 3.1198,
92
+ "step": 220
93
+ },
94
+ {
95
+ "epoch": 0.13,
96
+ "learning_rate": 0.0007171070552982838,
97
+ "loss": 3.1801,
98
+ "step": 240
99
+ },
100
+ {
101
+ "epoch": 0.14,
102
+ "learning_rate": 0.0007132252792154726,
103
+ "loss": 3.3954,
104
+ "step": 260
105
+ },
106
+ {
107
+ "epoch": 0.15,
108
+ "learning_rate": 0.000709139199128303,
109
+ "loss": 3.0566,
110
+ "step": 280
111
+ },
112
+ {
113
+ "epoch": 0.16,
114
+ "learning_rate": 0.0007052574230454917,
115
+ "loss": 3.483,
116
+ "step": 300
117
+ },
118
+ {
119
+ "epoch": 0.16,
120
+ "eval_loss": Infinity,
121
+ "eval_runtime": 1262.2253,
122
+ "eval_samples_per_second": 21.088,
123
+ "eval_steps_per_second": 0.659,
124
+ "eval_wer": 1.0,
125
+ "step": 300
126
+ },
127
+ {
128
+ "epoch": 0.17,
129
+ "learning_rate": 0.000701171342958322,
130
+ "loss": 3.1616,
131
+ "step": 320
132
+ },
133
+ {
134
+ "epoch": 0.18,
135
+ "learning_rate": 0.0006970852628711522,
136
+ "loss": 3.2017,
137
+ "step": 340
138
+ },
139
+ {
140
+ "epoch": 0.19,
141
+ "learning_rate": 0.0006932034867883411,
142
+ "loss": 3.4474,
143
+ "step": 360
144
+ },
145
+ {
146
+ "epoch": 0.2,
147
+ "learning_rate": 0.0006891174067011714,
148
+ "loss": 3.0656,
149
+ "step": 380
150
+ },
151
+ {
152
+ "epoch": 0.21,
153
+ "learning_rate": 0.0006852356306183601,
154
+ "loss": 3.5014,
155
+ "step": 400
156
+ },
157
+ {
158
+ "epoch": 0.21,
159
+ "eval_loss": Infinity,
160
+ "eval_runtime": 1255.7751,
161
+ "eval_samples_per_second": 21.196,
162
+ "eval_steps_per_second": 0.663,
163
+ "eval_wer": 1.0,
164
+ "step": 400
165
+ },
166
+ {
167
+ "epoch": 0.22,
168
+ "learning_rate": 0.0006811495505311904,
169
+ "loss": 3.1757,
170
+ "step": 420
171
+ },
172
+ {
173
+ "epoch": 0.23,
174
+ "learning_rate": 0.0006770634704440206,
175
+ "loss": 3.2038,
176
+ "step": 440
177
+ },
178
+ {
179
+ "epoch": 0.25,
180
+ "learning_rate": 0.0006731816943612095,
181
+ "loss": 3.4072,
182
+ "step": 460
183
+ },
184
+ {
185
+ "epoch": 0.26,
186
+ "learning_rate": 0.0006690956142740398,
187
+ "loss": 3.0498,
188
+ "step": 480
189
+ },
190
+ {
191
+ "epoch": 0.27,
192
+ "learning_rate": 0.00066500953418687,
193
+ "loss": 3.331,
194
+ "step": 500
195
+ },
196
+ {
197
+ "epoch": 0.27,
198
+ "eval_loss": Infinity,
199
+ "eval_runtime": 1239.2922,
200
+ "eval_samples_per_second": 21.478,
201
+ "eval_steps_per_second": 0.671,
202
+ "eval_wer": 1.0,
203
+ "step": 500
204
+ },
205
+ {
206
+ "epoch": 0.28,
207
+ "learning_rate": 0.0006609234540997004,
208
+ "loss": 3.0833,
209
+ "step": 520
210
+ },
211
+ {
212
+ "epoch": 0.29,
213
+ "learning_rate": 0.0006568373740125306,
214
+ "loss": 3.1989,
215
+ "step": 540
216
+ },
217
+ {
218
+ "epoch": 0.3,
219
+ "learning_rate": 0.0006531599019340779,
220
+ "loss": 3.5001,
221
+ "step": 560
222
+ },
223
+ {
224
+ "epoch": 0.31,
225
+ "learning_rate": 0.0006490738218469082,
226
+ "loss": 3.0685,
227
+ "step": 580
228
+ },
229
+ {
230
+ "epoch": 0.32,
231
+ "learning_rate": 0.000645192045764097,
232
+ "loss": 3.4809,
233
+ "step": 600
234
+ },
235
+ {
236
+ "epoch": 0.32,
237
+ "eval_loss": Infinity,
238
+ "eval_runtime": 1232.3119,
239
+ "eval_samples_per_second": 21.6,
240
+ "eval_steps_per_second": 0.675,
241
+ "eval_wer": 1.0,
242
+ "step": 600
243
+ },
244
+ {
245
+ "epoch": 0.33,
246
+ "learning_rate": 0.0006411059656769273,
247
+ "loss": 3.1877,
248
+ "step": 620
249
+ },
250
+ {
251
+ "epoch": 0.34,
252
+ "learning_rate": 0.0006370198855897576,
253
+ "loss": 3.2365,
254
+ "step": 640
255
+ },
256
+ {
257
+ "epoch": 0.35,
258
+ "learning_rate": 0.0006333424135113049,
259
+ "loss": 3.67,
260
+ "step": 660
261
+ },
262
+ {
263
+ "epoch": 0.36,
264
+ "learning_rate": 0.0006292563334241351,
265
+ "loss": 3.086,
266
+ "step": 680
267
+ },
268
+ {
269
+ "epoch": 0.37,
270
+ "learning_rate": 0.0006253745573413239,
271
+ "loss": 3.4678,
272
+ "step": 700
273
+ },
274
+ {
275
+ "epoch": 0.37,
276
+ "eval_loss": Infinity,
277
+ "eval_runtime": 1236.8236,
278
+ "eval_samples_per_second": 21.521,
279
+ "eval_steps_per_second": 0.673,
280
+ "eval_wer": 1.0,
281
+ "step": 700
282
+ },
283
+ {
284
+ "epoch": 0.38,
285
+ "learning_rate": 0.0006212884772541542,
286
+ "loss": 3.2344,
287
+ "step": 720
288
+ },
289
+ {
290
+ "epoch": 0.4,
291
+ "learning_rate": 0.0006172023971669845,
292
+ "loss": 3.2507,
293
+ "step": 740
294
+ },
295
+ {
296
+ "epoch": 0.41,
297
+ "learning_rate": 0.0006131163170798147,
298
+ "loss": 3.4283,
299
+ "step": 760
300
+ },
301
+ {
302
+ "epoch": 0.42,
303
+ "learning_rate": 0.0006090302369926451,
304
+ "loss": 3.0802,
305
+ "step": 780
306
+ },
307
+ {
308
+ "epoch": 0.43,
309
+ "learning_rate": 0.0006051484609098339,
310
+ "loss": 3.4596,
311
+ "step": 800
312
+ },
313
+ {
314
+ "epoch": 0.43,
315
+ "eval_loss": Infinity,
316
+ "eval_runtime": 1233.785,
317
+ "eval_samples_per_second": 21.574,
318
+ "eval_steps_per_second": 0.674,
319
+ "eval_wer": 1.0,
320
+ "step": 800
321
+ },
322
+ {
323
+ "epoch": 0.44,
324
+ "learning_rate": 0.0006010623808226641,
325
+ "loss": 3.2227,
326
+ "step": 820
327
+ },
328
+ {
329
+ "epoch": 0.45,
330
+ "learning_rate": 0.0005969763007354945,
331
+ "loss": 3.2357,
332
+ "step": 840
333
+ },
334
+ {
335
+ "epoch": 0.46,
336
+ "learning_rate": 0.0005930945246526832,
337
+ "loss": 3.5563,
338
+ "step": 860
339
+ },
340
+ {
341
+ "epoch": 0.47,
342
+ "learning_rate": 0.0005890084445655135,
343
+ "loss": 3.0845,
344
+ "step": 880
345
+ },
346
+ {
347
+ "epoch": 0.48,
348
+ "learning_rate": 0.0005851266684827023,
349
+ "loss": 3.4644,
350
+ "step": 900
351
+ },
352
+ {
353
+ "epoch": 0.48,
354
+ "eval_loss": Infinity,
355
+ "eval_runtime": 1235.1488,
356
+ "eval_samples_per_second": 21.55,
357
+ "eval_steps_per_second": 0.674,
358
+ "eval_wer": 1.0,
359
+ "step": 900
360
+ },
361
+ {
362
+ "epoch": 0.49,
363
+ "learning_rate": 0.0005810405883955325,
364
+ "loss": 3.2092,
365
+ "step": 920
366
+ },
367
+ {
368
+ "epoch": 0.5,
369
+ "learning_rate": 0.0005769545083083629,
370
+ "loss": 3.2157,
371
+ "step": 940
372
+ },
373
+ {
374
+ "epoch": 0.51,
375
+ "learning_rate": 0.0005730727322255517,
376
+ "loss": 3.4679,
377
+ "step": 960
378
+ },
379
+ {
380
+ "epoch": 0.52,
381
+ "learning_rate": 0.0005689866521383819,
382
+ "loss": 3.0662,
383
+ "step": 980
384
+ },
385
+ {
386
+ "epoch": 0.53,
387
+ "learning_rate": 0.0005651048760555707,
388
+ "loss": 3.4671,
389
+ "step": 1000
390
+ },
391
+ {
392
+ "epoch": 0.53,
393
+ "eval_loss": Infinity,
394
+ "eval_runtime": 1241.2249,
395
+ "eval_samples_per_second": 21.445,
396
+ "eval_steps_per_second": 0.67,
397
+ "eval_wer": 1.0,
398
+ "step": 1000
399
+ },
400
+ {
401
+ "epoch": 0.54,
402
+ "learning_rate": 0.0005610187959684009,
403
+ "loss": 3.2428,
404
+ "step": 1020
405
+ },
406
+ {
407
+ "epoch": 0.56,
408
+ "learning_rate": 0.0005569327158812314,
409
+ "loss": 3.2165,
410
+ "step": 1040
411
+ },
412
+ {
413
+ "epoch": 0.57,
414
+ "learning_rate": 0.0005530509397984201,
415
+ "loss": 3.5106,
416
+ "step": 1060
417
+ },
418
+ {
419
+ "epoch": 0.58,
420
+ "learning_rate": 0.0005489648597112503,
421
+ "loss": 3.1137,
422
+ "step": 1080
423
+ },
424
+ {
425
+ "epoch": 0.59,
426
+ "learning_rate": 0.0005452873876327976,
427
+ "loss": 3.6005,
428
+ "step": 1100
429
+ },
430
+ {
431
+ "epoch": 0.59,
432
+ "eval_loss": Infinity,
433
+ "eval_runtime": 1236.2955,
434
+ "eval_samples_per_second": 21.53,
435
+ "eval_steps_per_second": 0.673,
436
+ "eval_wer": 1.0,
437
+ "step": 1100
438
+ },
439
+ {
440
+ "epoch": 0.6,
441
+ "learning_rate": 0.0005412013075456279,
442
+ "loss": 3.2702,
443
+ "step": 1120
444
+ },
445
+ {
446
+ "epoch": 0.61,
447
+ "learning_rate": 0.0005371152274584582,
448
+ "loss": 3.6768,
449
+ "step": 1140
450
+ },
451
+ {
452
+ "epoch": 0.62,
453
+ "learning_rate": 0.000533233451375647,
454
+ "loss": 3.6313,
455
+ "step": 1160
456
+ },
457
+ {
458
+ "epoch": 0.63,
459
+ "learning_rate": 0.0005291473712884773,
460
+ "loss": 3.2456,
461
+ "step": 1180
462
+ },
463
+ {
464
+ "epoch": 0.64,
465
+ "learning_rate": 0.000525265595205666,
466
+ "loss": 3.9182,
467
+ "step": 1200
468
+ },
469
+ {
470
+ "epoch": 0.64,
471
+ "eval_loss": Infinity,
472
+ "eval_runtime": 1229.4316,
473
+ "eval_samples_per_second": 21.651,
474
+ "eval_steps_per_second": 0.677,
475
+ "eval_wer": 1.0,
476
+ "step": 1200
477
+ },
478
+ {
479
+ "epoch": 0.65,
480
+ "learning_rate": 0.0005211795151184963,
481
+ "loss": 3.3805,
482
+ "step": 1220
483
+ },
484
+ {
485
+ "epoch": 0.66,
486
+ "learning_rate": 0.0005170934350313267,
487
+ "loss": 3.4687,
488
+ "step": 1240
489
+ },
490
+ {
491
+ "epoch": 0.67,
492
+ "learning_rate": 0.000513007354944157,
493
+ "loss": 3.4473,
494
+ "step": 1260
495
+ },
496
+ {
497
+ "epoch": 0.68,
498
+ "learning_rate": 0.0005089212748569872,
499
+ "loss": 3.2232,
500
+ "step": 1280
501
+ },
502
+ {
503
+ "epoch": 0.69,
504
+ "learning_rate": 0.000505039498774176,
505
+ "loss": 3.6466,
506
+ "step": 1300
507
+ },
508
+ {
509
+ "epoch": 0.69,
510
+ "eval_loss": Infinity,
511
+ "eval_runtime": 1237.3739,
512
+ "eval_samples_per_second": 21.512,
513
+ "eval_steps_per_second": 0.672,
514
+ "eval_wer": 1.0,
515
+ "step": 1300
516
+ },
517
+ {
518
+ "epoch": 0.7,
519
+ "learning_rate": 0.0005009534186870062,
520
+ "loss": 3.3496,
521
+ "step": 1320
522
+ },
523
+ {
524
+ "epoch": 0.72,
525
+ "learning_rate": 0.0004968673385998365,
526
+ "loss": 3.3918,
527
+ "step": 1340
528
+ },
529
+ {
530
+ "epoch": 0.73,
531
+ "learning_rate": 0.0004927812585126669,
532
+ "loss": 3.3232,
533
+ "step": 1360
534
+ },
535
+ {
536
+ "epoch": 0.74,
537
+ "learning_rate": 0.0004886951784254972,
538
+ "loss": 3.288,
539
+ "step": 1380
540
+ },
541
+ {
542
+ "epoch": 0.75,
543
+ "learning_rate": 0.00048481340234268587,
544
+ "loss": 3.6932,
545
+ "step": 1400
546
+ },
547
+ {
548
+ "epoch": 0.75,
549
+ "eval_loss": Infinity,
550
+ "eval_runtime": 1236.694,
551
+ "eval_samples_per_second": 21.524,
552
+ "eval_steps_per_second": 0.673,
553
+ "eval_wer": 1.0,
554
+ "step": 1400
555
+ },
556
+ {
557
+ "epoch": 0.76,
558
+ "learning_rate": 0.00048072732225551624,
559
+ "loss": 3.5411,
560
+ "step": 1420
561
+ },
562
+ {
563
+ "epoch": 0.77,
564
+ "learning_rate": 0.0004766412421683465,
565
+ "loss": 3.5658,
566
+ "step": 1440
567
+ },
568
+ {
569
+ "epoch": 0.78,
570
+ "learning_rate": 0.00047275946608553524,
571
+ "loss": 3.7666,
572
+ "step": 1460
573
+ },
574
+ {
575
+ "epoch": 0.79,
576
+ "learning_rate": 0.0004686733859983656,
577
+ "loss": 3.7564,
578
+ "step": 1480
579
+ },
580
+ {
581
+ "epoch": 0.8,
582
+ "learning_rate": 0.00046479160991555436,
583
+ "loss": 3.7939,
584
+ "step": 1500
585
+ },
586
+ {
587
+ "epoch": 0.8,
588
+ "eval_loss": Infinity,
589
+ "eval_runtime": 1235.7295,
590
+ "eval_samples_per_second": 21.54,
591
+ "eval_steps_per_second": 0.673,
592
+ "eval_wer": 1.0,
593
+ "step": 1500
594
+ },
595
+ {
596
+ "epoch": 0.81,
597
+ "learning_rate": 0.00046070552982838467,
598
+ "loss": 3.7471,
599
+ "step": 1520
600
+ },
601
+ {
602
+ "epoch": 0.82,
603
+ "learning_rate": 0.00045661944974121493,
604
+ "loss": 3.7457,
605
+ "step": 1540
606
+ },
607
+ {
608
+ "epoch": 0.83,
609
+ "learning_rate": 0.0004527376736584037,
610
+ "loss": 3.798,
611
+ "step": 1560
612
+ },
613
+ {
614
+ "epoch": 0.84,
615
+ "learning_rate": 0.00044865159357123405,
616
+ "loss": 3.7611,
617
+ "step": 1580
618
+ },
619
+ {
620
+ "epoch": 0.85,
621
+ "learning_rate": 0.0004447698174884228,
622
+ "loss": 3.9284,
623
+ "step": 1600
624
+ },
625
+ {
626
+ "epoch": 0.85,
627
+ "eval_loss": Infinity,
628
+ "eval_runtime": 1251.8158,
629
+ "eval_samples_per_second": 21.264,
630
+ "eval_steps_per_second": 0.665,
631
+ "eval_wer": 1.0,
632
+ "step": 1600
633
+ },
634
+ {
635
+ "epoch": 0.86,
636
+ "learning_rate": 0.0004406837374012531,
637
+ "loss": 3.7416,
638
+ "step": 1620
639
+ },
640
+ {
641
+ "epoch": 0.88,
642
+ "learning_rate": 0.00043659765731408337,
643
+ "loss": 3.7416,
644
+ "step": 1640
645
+ },
646
+ {
647
+ "epoch": 0.89,
648
+ "learning_rate": 0.0004327158812312721,
649
+ "loss": 3.8024,
650
+ "step": 1660
651
+ },
652
+ {
653
+ "epoch": 0.9,
654
+ "learning_rate": 0.0004286298011441025,
655
+ "loss": 3.7578,
656
+ "step": 1680
657
+ },
658
+ {
659
+ "epoch": 0.91,
660
+ "learning_rate": 0.0004247480250612912,
661
+ "loss": 3.7859,
662
+ "step": 1700
663
+ },
664
+ {
665
+ "epoch": 0.91,
666
+ "eval_loss": Infinity,
667
+ "eval_runtime": 1235.734,
668
+ "eval_samples_per_second": 21.54,
669
+ "eval_steps_per_second": 0.673,
670
+ "eval_wer": 1.0,
671
+ "step": 1700
672
+ },
673
+ {
674
+ "epoch": 0.92,
675
+ "learning_rate": 0.0004206619449741215,
676
+ "loss": 3.7427,
677
+ "step": 1720
678
+ },
679
+ {
680
+ "epoch": 0.93,
681
+ "learning_rate": 0.0004165758648869518,
682
+ "loss": 3.751,
683
+ "step": 1740
684
+ },
685
+ {
686
+ "epoch": 0.94,
687
+ "learning_rate": 0.00041248978479978206,
688
+ "loss": 3.6701,
689
+ "step": 1760
690
+ },
691
+ {
692
+ "epoch": 0.95,
693
+ "learning_rate": 0.0004084037047126123,
694
+ "loss": 3.7613,
695
+ "step": 1780
696
+ },
697
+ {
698
+ "epoch": 0.96,
699
+ "learning_rate": 0.00040472623263415966,
700
+ "loss": 3.9363,
701
+ "step": 1800
702
+ },
703
+ {
704
+ "epoch": 0.96,
705
+ "eval_loss": Infinity,
706
+ "eval_runtime": 1236.9158,
707
+ "eval_samples_per_second": 21.52,
708
+ "eval_steps_per_second": 0.673,
709
+ "eval_wer": 1.0,
710
+ "step": 1800
711
+ },
712
+ {
713
+ "epoch": 0.97,
714
+ "learning_rate": 0.0004006401525469899,
715
+ "loss": 3.7464,
716
+ "step": 1820
717
+ },
718
+ {
719
+ "epoch": 0.98,
720
+ "learning_rate": 0.00039655407245982023,
721
+ "loss": 3.7414,
722
+ "step": 1840
723
+ },
724
+ {
725
+ "epoch": 0.99,
726
+ "learning_rate": 0.000392672296377009,
727
+ "loss": 3.8036,
728
+ "step": 1860
729
+ },
730
+ {
731
+ "epoch": 1.0,
732
+ "learning_rate": 0.00038879052029419783,
733
+ "loss": 3.8768,
734
+ "step": 1880
735
+ },
736
+ {
737
+ "epoch": 1.01,
738
+ "learning_rate": 0.0003847044402070281,
739
+ "loss": 3.7573,
740
+ "step": 1900
741
+ },
742
+ {
743
+ "epoch": 1.01,
744
+ "eval_loss": Infinity,
745
+ "eval_runtime": 1232.3535,
746
+ "eval_samples_per_second": 21.599,
747
+ "eval_steps_per_second": 0.675,
748
+ "eval_wer": 1.0,
749
+ "step": 1900
750
+ },
751
+ {
752
+ "epoch": 1.03,
753
+ "learning_rate": 0.00038061836011985835,
754
+ "loss": 3.6959,
755
+ "step": 1920
756
+ },
757
+ {
758
+ "epoch": 1.04,
759
+ "learning_rate": 0.00037653228003268867,
760
+ "loss": 3.7313,
761
+ "step": 1940
762
+ },
763
+ {
764
+ "epoch": 1.05,
765
+ "learning_rate": 0.00037244619994551893,
766
+ "loss": 3.7676,
767
+ "step": 1960
768
+ },
769
+ {
770
+ "epoch": 1.06,
771
+ "learning_rate": 0.0003685644238627077,
772
+ "loss": 3.7872,
773
+ "step": 1980
774
+ },
775
+ {
776
+ "epoch": 1.07,
777
+ "learning_rate": 0.000364478343775538,
778
+ "loss": 3.7553,
779
+ "step": 2000
780
+ },
781
+ {
782
+ "epoch": 1.07,
783
+ "eval_loss": Infinity,
784
+ "eval_runtime": 1239.2247,
785
+ "eval_samples_per_second": 21.48,
786
+ "eval_steps_per_second": 0.671,
787
+ "eval_wer": 1.0,
788
+ "step": 2000
789
+ },
790
+ {
791
+ "epoch": 1.08,
792
+ "learning_rate": 0.0003603922636883683,
793
+ "loss": 3.694,
794
+ "step": 2020
795
+ },
796
+ {
797
+ "epoch": 1.09,
798
+ "learning_rate": 0.0003565104876055571,
799
+ "loss": 3.8389,
800
+ "step": 2040
801
+ },
802
+ {
803
+ "epoch": 1.1,
804
+ "learning_rate": 0.00035242440751838736,
805
+ "loss": 3.7573,
806
+ "step": 2060
807
+ },
808
+ {
809
+ "epoch": 1.11,
810
+ "learning_rate": 0.0003483383274312177,
811
+ "loss": 3.6641,
812
+ "step": 2080
813
+ },
814
+ {
815
+ "epoch": 1.12,
816
+ "learning_rate": 0.00034425224734404794,
817
+ "loss": 3.7606,
818
+ "step": 2100
819
+ },
820
+ {
821
+ "epoch": 1.12,
822
+ "eval_loss": Infinity,
823
+ "eval_runtime": 1237.7997,
824
+ "eval_samples_per_second": 21.504,
825
+ "eval_steps_per_second": 0.672,
826
+ "eval_wer": 1.0,
827
+ "step": 2100
828
+ },
829
+ {
830
+ "epoch": 1.13,
831
+ "learning_rate": 0.00034016616725687825,
832
+ "loss": 3.6827,
833
+ "step": 2120
834
+ },
835
+ {
836
+ "epoch": 1.14,
837
+ "learning_rate": 0.000336284391174067,
838
+ "loss": 3.8375,
839
+ "step": 2140
840
+ },
841
+ {
842
+ "epoch": 1.15,
843
+ "learning_rate": 0.0003321983110868973,
844
+ "loss": 3.7527,
845
+ "step": 2160
846
+ },
847
+ {
848
+ "epoch": 1.16,
849
+ "learning_rate": 0.0003283165350040861,
850
+ "loss": 3.7881,
851
+ "step": 2180
852
+ },
853
+ {
854
+ "epoch": 1.17,
855
+ "learning_rate": 0.0003242304549169164,
856
+ "loss": 3.7514,
857
+ "step": 2200
858
+ },
859
+ {
860
+ "epoch": 1.17,
861
+ "eval_loss": Infinity,
862
+ "eval_runtime": 1237.076,
863
+ "eval_samples_per_second": 21.517,
864
+ "eval_steps_per_second": 0.673,
865
+ "eval_wer": 1.0,
866
+ "step": 2200
867
+ },
868
+ {
869
+ "epoch": 1.19,
870
+ "learning_rate": 0.0003201443748297467,
871
+ "loss": 3.6727,
872
+ "step": 2220
873
+ },
874
+ {
875
+ "epoch": 1.2,
876
+ "learning_rate": 0.00031626259874693543,
877
+ "loss": 3.8424,
878
+ "step": 2240
879
+ },
880
+ {
881
+ "epoch": 1.21,
882
+ "learning_rate": 0.00031217651865976574,
883
+ "loss": 3.7594,
884
+ "step": 2260
885
+ },
886
+ {
887
+ "epoch": 1.22,
888
+ "learning_rate": 0.00030829474257695454,
889
+ "loss": 3.7862,
890
+ "step": 2280
891
+ },
892
+ {
893
+ "epoch": 1.23,
894
+ "learning_rate": 0.0003042086624897848,
895
+ "loss": 3.7472,
896
+ "step": 2300
897
+ },
898
+ {
899
+ "epoch": 1.23,
900
+ "eval_loss": Infinity,
901
+ "eval_runtime": 1236.6319,
902
+ "eval_samples_per_second": 21.525,
903
+ "eval_steps_per_second": 0.673,
904
+ "eval_wer": 1.0,
905
+ "step": 2300
906
+ },
907
+ {
908
+ "epoch": 1.24,
909
+ "learning_rate": 0.0003001225824026151,
910
+ "loss": 3.6956,
911
+ "step": 2320
912
+ },
913
+ {
914
+ "epoch": 1.25,
915
+ "learning_rate": 0.0002960365023154454,
916
+ "loss": 3.7248,
917
+ "step": 2340
918
+ },
919
+ {
920
+ "epoch": 1.26,
921
+ "learning_rate": 0.0002919504222282757,
922
+ "loss": 3.7531,
923
+ "step": 2360
924
+ },
925
+ {
926
+ "epoch": 1.27,
927
+ "learning_rate": 0.0002882729501498229,
928
+ "loss": 3.9245,
929
+ "step": 2380
930
+ },
931
+ {
932
+ "epoch": 1.28,
933
+ "learning_rate": 0.00028418687006265324,
934
+ "loss": 3.7478,
935
+ "step": 2400
936
+ },
937
+ {
938
+ "epoch": 1.28,
939
+ "eval_loss": Infinity,
940
+ "eval_runtime": 1236.7725,
941
+ "eval_samples_per_second": 21.522,
942
+ "eval_steps_per_second": 0.673,
943
+ "eval_wer": 1.0,
944
+ "step": 2400
945
+ },
946
+ {
947
+ "epoch": 1.29,
948
+ "learning_rate": 0.00028010078997548355,
949
+ "loss": 3.6826,
950
+ "step": 2420
951
+ },
952
+ {
953
+ "epoch": 1.3,
954
+ "learning_rate": 0.0002760147098883138,
955
+ "loss": 3.7236,
956
+ "step": 2440
957
+ },
958
+ {
959
+ "epoch": 1.31,
960
+ "learning_rate": 0.00027192862980114413,
961
+ "loss": 3.7609,
962
+ "step": 2460
963
+ },
964
+ {
965
+ "epoch": 1.32,
966
+ "learning_rate": 0.00026804685371833287,
967
+ "loss": 3.7846,
968
+ "step": 2480
969
+ },
970
+ {
971
+ "epoch": 1.33,
972
+ "learning_rate": 0.0002639607736311632,
973
+ "loss": 3.7496,
974
+ "step": 2500
975
+ },
976
+ {
977
+ "epoch": 1.33,
978
+ "eval_loss": Infinity,
979
+ "eval_runtime": 1233.5496,
980
+ "eval_samples_per_second": 21.578,
981
+ "eval_steps_per_second": 0.674,
982
+ "eval_wer": 1.0,
983
+ "step": 2500
984
+ },
985
+ {
986
+ "epoch": 1.35,
987
+ "learning_rate": 0.00025987469354399345,
988
+ "loss": 3.6785,
989
+ "step": 2520
990
+ },
991
+ {
992
+ "epoch": 1.36,
993
+ "learning_rate": 0.00025578861345682376,
994
+ "loss": 3.7212,
995
+ "step": 2540
996
+ },
997
+ {
998
+ "epoch": 1.37,
999
+ "learning_rate": 0.0002517025333696541,
1000
+ "loss": 3.7637,
1001
+ "step": 2560
1002
+ },
1003
+ {
1004
+ "epoch": 1.38,
1005
+ "learning_rate": 0.00024761645328248434,
1006
+ "loss": 3.6513,
1007
+ "step": 2580
1008
+ },
1009
+ {
1010
+ "epoch": 1.39,
1011
+ "learning_rate": 0.00024353037319531465,
1012
+ "loss": 3.7513,
1013
+ "step": 2600
1014
+ },
1015
+ {
1016
+ "epoch": 1.39,
1017
+ "eval_loss": Infinity,
1018
+ "eval_runtime": 1231.8132,
1019
+ "eval_samples_per_second": 21.609,
1020
+ "eval_steps_per_second": 0.675,
1021
+ "eval_wer": 1.0,
1022
+ "step": 2600
1023
+ },
1024
+ {
1025
+ "epoch": 1.4,
1026
+ "learning_rate": 0.00023944429310814494,
1027
+ "loss": 3.6818,
1028
+ "step": 2620
1029
+ },
1030
+ {
1031
+ "epoch": 1.41,
1032
+ "learning_rate": 0.00023556251702533368,
1033
+ "loss": 3.8417,
1034
+ "step": 2640
1035
+ },
1036
+ {
1037
+ "epoch": 1.42,
1038
+ "learning_rate": 0.000231476436938164,
1039
+ "loss": 3.7639,
1040
+ "step": 2660
1041
+ },
1042
+ {
1043
+ "epoch": 1.43,
1044
+ "learning_rate": 0.00022759466085535277,
1045
+ "loss": 3.7842,
1046
+ "step": 2680
1047
+ },
1048
+ {
1049
+ "epoch": 1.44,
1050
+ "learning_rate": 0.00022350858076818309,
1051
+ "loss": 3.7497,
1052
+ "step": 2700
1053
+ },
1054
+ {
1055
+ "epoch": 1.44,
1056
+ "eval_loss": Infinity,
1057
+ "eval_runtime": 1234.8437,
1058
+ "eval_samples_per_second": 21.556,
1059
+ "eval_steps_per_second": 0.674,
1060
+ "eval_wer": 1.0,
1061
+ "step": 2700
1062
+ },
1063
+ {
1064
+ "epoch": 1.45,
1065
+ "learning_rate": 0.00021942250068101335,
1066
+ "loss": 3.6795,
1067
+ "step": 2720
1068
+ },
1069
+ {
1070
+ "epoch": 1.46,
1071
+ "learning_rate": 0.00021554072459820212,
1072
+ "loss": 3.8421,
1073
+ "step": 2740
1074
+ },
1075
+ {
1076
+ "epoch": 1.47,
1077
+ "learning_rate": 0.00021145464451103243,
1078
+ "loss": 3.7614,
1079
+ "step": 2760
1080
+ },
1081
+ {
1082
+ "epoch": 1.48,
1083
+ "learning_rate": 0.0002075728684282212,
1084
+ "loss": 3.9259,
1085
+ "step": 2780
1086
+ },
1087
+ {
1088
+ "epoch": 1.49,
1089
+ "learning_rate": 0.00020348678834105147,
1090
+ "loss": 3.7539,
1091
+ "step": 2800
1092
+ },
1093
+ {
1094
+ "epoch": 1.49,
1095
+ "eval_loss": Infinity,
1096
+ "eval_runtime": 1240.1103,
1097
+ "eval_samples_per_second": 21.464,
1098
+ "eval_steps_per_second": 0.671,
1099
+ "eval_wer": 1.0,
1100
+ "step": 2800
1101
+ },
1102
+ {
1103
+ "epoch": 1.51,
1104
+ "learning_rate": 0.00019940070825388178,
1105
+ "loss": 3.6889,
1106
+ "step": 2820
1107
+ },
1108
+ {
1109
+ "epoch": 1.52,
1110
+ "learning_rate": 0.00019551893217107055,
1111
+ "loss": 3.843,
1112
+ "step": 2840
1113
+ },
1114
+ {
1115
+ "epoch": 1.53,
1116
+ "learning_rate": 0.00019143285208390087,
1117
+ "loss": 3.7666,
1118
+ "step": 2860
1119
+ },
1120
+ {
1121
+ "epoch": 1.54,
1122
+ "learning_rate": 0.00018755107600108964,
1123
+ "loss": 3.786,
1124
+ "step": 2880
1125
+ },
1126
+ {
1127
+ "epoch": 1.55,
1128
+ "learning_rate": 0.0001834649959139199,
1129
+ "loss": 3.7581,
1130
+ "step": 2900
1131
+ },
1132
+ {
1133
+ "epoch": 1.55,
1134
+ "eval_loss": Infinity,
1135
+ "eval_runtime": 1231.1699,
1136
+ "eval_samples_per_second": 21.62,
1137
+ "eval_steps_per_second": 0.676,
1138
+ "eval_wer": 1.0,
1139
+ "step": 2900
1140
+ },
1141
+ {
1142
+ "epoch": 1.56,
1143
+ "learning_rate": 0.00017937891582675021,
1144
+ "loss": 3.672,
1145
+ "step": 2920
1146
+ },
1147
+ {
1148
+ "epoch": 1.57,
1149
+ "learning_rate": 0.00017549713974393899,
1150
+ "loss": 3.8539,
1151
+ "step": 2940
1152
+ },
1153
+ {
1154
+ "epoch": 1.58,
1155
+ "learning_rate": 0.00017141105965676927,
1156
+ "loss": 3.7652,
1157
+ "step": 2960
1158
+ },
1159
+ {
1160
+ "epoch": 1.59,
1161
+ "learning_rate": 0.00016752928357395807,
1162
+ "loss": 3.7825,
1163
+ "step": 2980
1164
+ },
1165
+ {
1166
+ "epoch": 1.6,
1167
+ "learning_rate": 0.00016344320348678833,
1168
+ "loss": 3.7572,
1169
+ "step": 3000
1170
+ },
1171
+ {
1172
+ "epoch": 1.6,
1173
+ "eval_loss": Infinity,
1174
+ "eval_runtime": 1241.8499,
1175
+ "eval_samples_per_second": 21.434,
1176
+ "eval_steps_per_second": 0.67,
1177
+ "eval_wer": 1.0,
1178
+ "step": 3000
1179
+ },
1180
+ {
1181
+ "epoch": 1.61,
1182
+ "learning_rate": 0.00015935712339961862,
1183
+ "loss": 3.6906,
1184
+ "step": 3020
1185
+ },
1186
+ {
1187
+ "epoch": 1.62,
1188
+ "learning_rate": 0.00015547534731680742,
1189
+ "loss": 3.845,
1190
+ "step": 3040
1191
+ },
1192
+ {
1193
+ "epoch": 1.63,
1194
+ "learning_rate": 0.0001513892672296377,
1195
+ "loss": 3.763,
1196
+ "step": 3060
1197
+ },
1198
+ {
1199
+ "epoch": 1.64,
1200
+ "learning_rate": 0.00014750749114682648,
1201
+ "loss": 3.779,
1202
+ "step": 3080
1203
+ },
1204
+ {
1205
+ "epoch": 1.66,
1206
+ "learning_rate": 0.0001434214110596568,
1207
+ "loss": 3.7589,
1208
+ "step": 3100
1209
+ },
1210
+ {
1211
+ "epoch": 1.66,
1212
+ "eval_loss": Infinity,
1213
+ "eval_runtime": 1236.916,
1214
+ "eval_samples_per_second": 21.52,
1215
+ "eval_steps_per_second": 0.673,
1216
+ "eval_wer": 1.0,
1217
+ "step": 3100
1218
+ },
1219
+ {
1220
+ "epoch": 1.67,
1221
+ "learning_rate": 0.00013933533097248705,
1222
+ "loss": 3.6948,
1223
+ "step": 3120
1224
+ },
1225
+ {
1226
+ "epoch": 1.68,
1227
+ "learning_rate": 0.00013524925088531734,
1228
+ "loss": 3.7211,
1229
+ "step": 3140
1230
+ },
1231
+ {
1232
+ "epoch": 1.69,
1233
+ "learning_rate": 0.00013116317079814763,
1234
+ "loss": 3.7588,
1235
+ "step": 3160
1236
+ },
1237
+ {
1238
+ "epoch": 1.7,
1239
+ "learning_rate": 0.00012728139471533643,
1240
+ "loss": 3.7898,
1241
+ "step": 3180
1242
+ },
1243
+ {
1244
+ "epoch": 1.71,
1245
+ "learning_rate": 0.00012319531462816672,
1246
+ "loss": 3.7592,
1247
+ "step": 3200
1248
+ },
1249
+ {
1250
+ "epoch": 1.71,
1251
+ "eval_loss": Infinity,
1252
+ "eval_runtime": 1232.1567,
1253
+ "eval_samples_per_second": 21.603,
1254
+ "eval_steps_per_second": 0.675,
1255
+ "eval_wer": 1.0,
1256
+ "step": 3200
1257
+ },
1258
+ {
1259
+ "epoch": 1.72,
1260
+ "learning_rate": 0.000119109234540997,
1261
+ "loss": 3.688,
1262
+ "step": 3220
1263
+ },
1264
+ {
1265
+ "epoch": 1.73,
1266
+ "learning_rate": 0.00011522745845818579,
1267
+ "loss": 3.8419,
1268
+ "step": 3240
1269
+ },
1270
+ {
1271
+ "epoch": 1.74,
1272
+ "learning_rate": 0.00011114137837101608,
1273
+ "loss": 3.7578,
1274
+ "step": 3260
1275
+ },
1276
+ {
1277
+ "epoch": 1.75,
1278
+ "learning_rate": 0.00010705529828384636,
1279
+ "loss": 3.6534,
1280
+ "step": 3280
1281
+ },
1282
+ {
1283
+ "epoch": 1.76,
1284
+ "learning_rate": 0.00010296921819667667,
1285
+ "loss": 3.7531,
1286
+ "step": 3300
1287
+ },
1288
+ {
1289
+ "epoch": 1.76,
1290
+ "eval_loss": Infinity,
1291
+ "eval_runtime": 1237.4955,
1292
+ "eval_samples_per_second": 21.51,
1293
+ "eval_steps_per_second": 0.672,
1294
+ "eval_wer": 1.0,
1295
+ "step": 3300
1296
+ },
1297
+ {
1298
+ "epoch": 1.77,
1299
+ "learning_rate": 9.888313810950695e-05,
1300
+ "loss": 3.691,
1301
+ "step": 3320
1302
+ },
1303
+ {
1304
+ "epoch": 1.78,
1305
+ "learning_rate": 9.500136202669572e-05,
1306
+ "loss": 3.8424,
1307
+ "step": 3340
1308
+ },
1309
+ {
1310
+ "epoch": 1.79,
1311
+ "learning_rate": 9.091528193952601e-05,
1312
+ "loss": 3.7698,
1313
+ "step": 3360
1314
+ },
1315
+ {
1316
+ "epoch": 1.8,
1317
+ "learning_rate": 8.70335058567148e-05,
1318
+ "loss": 3.7836,
1319
+ "step": 3380
1320
+ },
1321
+ {
1322
+ "epoch": 1.82,
1323
+ "learning_rate": 8.294742576954509e-05,
1324
+ "loss": 3.7567,
1325
+ "step": 3400
1326
+ },
1327
+ {
1328
+ "epoch": 1.82,
1329
+ "eval_loss": Infinity,
1330
+ "eval_runtime": 1242.7888,
1331
+ "eval_samples_per_second": 21.418,
1332
+ "eval_steps_per_second": 0.669,
1333
+ "eval_wer": 1.0,
1334
+ "step": 3400
1335
+ },
1336
+ {
1337
+ "epoch": 1.83,
1338
+ "learning_rate": 7.886134568237537e-05,
1339
+ "loss": 3.6741,
1340
+ "step": 3420
1341
+ },
1342
+ {
1343
+ "epoch": 1.84,
1344
+ "learning_rate": 7.497956959956416e-05,
1345
+ "loss": 3.8442,
1346
+ "step": 3440
1347
+ },
1348
+ {
1349
+ "epoch": 1.85,
1350
+ "learning_rate": 7.089348951239445e-05,
1351
+ "loss": 3.7663,
1352
+ "step": 3460
1353
+ },
1354
+ {
1355
+ "epoch": 1.86,
1356
+ "learning_rate": 6.701171342958322e-05,
1357
+ "loss": 3.7789,
1358
+ "step": 3480
1359
+ },
1360
+ {
1361
+ "epoch": 1.87,
1362
+ "learning_rate": 6.292563334241352e-05,
1363
+ "loss": 3.7613,
1364
+ "step": 3500
1365
+ },
1366
+ {
1367
+ "epoch": 1.87,
1368
+ "eval_loss": Infinity,
1369
+ "eval_runtime": 1244.1733,
1370
+ "eval_samples_per_second": 21.394,
1371
+ "eval_steps_per_second": 0.669,
1372
+ "eval_wer": 1.0,
1373
+ "step": 3500
1374
+ },
1375
+ {
1376
+ "epoch": 1.88,
1377
+ "learning_rate": 5.88395532552438e-05,
1378
+ "loss": 3.6884,
1379
+ "step": 3520
1380
+ },
1381
+ {
1382
+ "epoch": 1.89,
1383
+ "learning_rate": 5.4957777172432585e-05,
1384
+ "loss": 3.8268,
1385
+ "step": 3540
1386
+ },
1387
+ {
1388
+ "epoch": 1.9,
1389
+ "learning_rate": 5.0871697085262866e-05,
1390
+ "loss": 3.7517,
1391
+ "step": 3560
1392
+ },
1393
+ {
1394
+ "epoch": 1.91,
1395
+ "learning_rate": 4.6989921002451645e-05,
1396
+ "loss": 3.7802,
1397
+ "step": 3580
1398
+ },
1399
+ {
1400
+ "epoch": 1.92,
1401
+ "learning_rate": 4.290384091528194e-05,
1402
+ "loss": 3.7516,
1403
+ "step": 3600
1404
+ },
1405
+ {
1406
+ "epoch": 1.92,
1407
+ "eval_loss": Infinity,
1408
+ "eval_runtime": 1253.9236,
1409
+ "eval_samples_per_second": 21.228,
1410
+ "eval_steps_per_second": 0.664,
1411
+ "eval_wer": 1.0,
1412
+ "step": 3600
1413
+ }
1414
+ ],
1415
+ "max_steps": 3746,
1416
+ "num_train_epochs": 2,
1417
+ "total_flos": 6.928820217293665e+19,
1418
+ "trial_name": null,
1419
+ "trial_params": null
1420
+ }
{checkpoint-300 β†’ checkpoint-3600}/training_args.bin RENAMED
File without changes
{checkpoint-400 β†’ checkpoint-3700}/config.json RENAMED
File without changes
{checkpoint-400 β†’ checkpoint-3700}/optimizer.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:32cea0be94924847e882a0462eee1db9f48388af0a8fc31b2591d926a51e783f
3
  size 2490337809
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:619caf3673344ab2a56d8920b454d1a102940191430a61b3330186dad60d1659
3
  size 2490337809
{checkpoint-400 β†’ checkpoint-3700}/preprocessor_config.json RENAMED
File without changes
{checkpoint-300 β†’ checkpoint-3700}/pytorch_model.bin RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3d10776e6a5adb1fe7db9d5f3268aa456e14c076f50d16eba046ef8c902080ec
3
  size 1262063089
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e95c67d08a1a0d2f34fb48babadb769336f0b21eeae723abbb0714176642fdf
3
  size 1262063089
{checkpoint-300 β†’ checkpoint-3700}/rng_state.pth RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9c7d04bb10253125453a6502f466a85d2f22da84c84fb7b619a7b640d0736e84
3
  size 14503
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff21b04ab59e592ca04a33825328c1a76f9615b11b66591c24235296cebfad21
3
  size 14503
{checkpoint-400 β†’ checkpoint-3700}/scaler.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:c1e02160a95d49a675cbb9a046de5c535a710ba6e764888e1228126b79ea4baa
3
  size 559
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d78f89f1dff85262347440e00b108cf0d3f1eb1be85989db2947bdafade5f248
3
  size 559
{checkpoint-500 β†’ checkpoint-3700}/scheduler.pt RENAMED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:150a6cef4019e3f97102978f0cff6d65ca1b19f0ead0397a080c28ea9c5a296b
3
  size 623
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e5a5afc631f3556ad61fe3b8190b2405aece04ce6c9c704fbd2ff1d06026a9b
3
  size 623
checkpoint-3700/trainer_state.json ADDED
@@ -0,0 +1,1459 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 1.97531028960363,
5
+ "global_step": 3700,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.01,
12
+ "learning_rate": 0.0002,
13
+ "loss": 10.6012,
14
+ "step": 20
15
+ },
16
+ {
17
+ "epoch": 0.02,
18
+ "learning_rate": 0.0004,
19
+ "loss": 3.651,
20
+ "step": 40
21
+ },
22
+ {
23
+ "epoch": 0.03,
24
+ "learning_rate": 0.00059,
25
+ "loss": 3.3911,
26
+ "step": 60
27
+ },
28
+ {
29
+ "epoch": 0.04,
30
+ "learning_rate": 0.0007491827839825661,
31
+ "loss": 3.0481,
32
+ "step": 80
33
+ },
34
+ {
35
+ "epoch": 0.05,
36
+ "learning_rate": 0.0007453010078997548,
37
+ "loss": 3.4039,
38
+ "step": 100
39
+ },
40
+ {
41
+ "epoch": 0.05,
42
+ "eval_loss": Infinity,
43
+ "eval_runtime": 1260.1628,
44
+ "eval_samples_per_second": 21.123,
45
+ "eval_steps_per_second": 0.66,
46
+ "eval_wer": 1.0,
47
+ "step": 100
48
+ },
49
+ {
50
+ "epoch": 0.06,
51
+ "learning_rate": 0.0007412149278125851,
52
+ "loss": 3.0553,
53
+ "step": 120
54
+ },
55
+ {
56
+ "epoch": 0.07,
57
+ "learning_rate": 0.0007371288477254155,
58
+ "loss": 3.2191,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.09,
63
+ "learning_rate": 0.0007332470716426042,
64
+ "loss": 3.4094,
65
+ "step": 160
66
+ },
67
+ {
68
+ "epoch": 0.1,
69
+ "learning_rate": 0.0007291609915554345,
70
+ "loss": 3.0607,
71
+ "step": 180
72
+ },
73
+ {
74
+ "epoch": 0.11,
75
+ "learning_rate": 0.0007252792154726233,
76
+ "loss": 3.4396,
77
+ "step": 200
78
+ },
79
+ {
80
+ "epoch": 0.11,
81
+ "eval_loss": Infinity,
82
+ "eval_runtime": 1235.1686,
83
+ "eval_samples_per_second": 21.55,
84
+ "eval_steps_per_second": 0.674,
85
+ "eval_wer": 1.0,
86
+ "step": 200
87
+ },
88
+ {
89
+ "epoch": 0.12,
90
+ "learning_rate": 0.0007211931353854536,
91
+ "loss": 3.1198,
92
+ "step": 220
93
+ },
94
+ {
95
+ "epoch": 0.13,
96
+ "learning_rate": 0.0007171070552982838,
97
+ "loss": 3.1801,
98
+ "step": 240
99
+ },
100
+ {
101
+ "epoch": 0.14,
102
+ "learning_rate": 0.0007132252792154726,
103
+ "loss": 3.3954,
104
+ "step": 260
105
+ },
106
+ {
107
+ "epoch": 0.15,
108
+ "learning_rate": 0.000709139199128303,
109
+ "loss": 3.0566,
110
+ "step": 280
111
+ },
112
+ {
113
+ "epoch": 0.16,
114
+ "learning_rate": 0.0007052574230454917,
115
+ "loss": 3.483,
116
+ "step": 300
117
+ },
118
+ {
119
+ "epoch": 0.16,
120
+ "eval_loss": Infinity,
121
+ "eval_runtime": 1262.2253,
122
+ "eval_samples_per_second": 21.088,
123
+ "eval_steps_per_second": 0.659,
124
+ "eval_wer": 1.0,
125
+ "step": 300
126
+ },
127
+ {
128
+ "epoch": 0.17,
129
+ "learning_rate": 0.000701171342958322,
130
+ "loss": 3.1616,
131
+ "step": 320
132
+ },
133
+ {
134
+ "epoch": 0.18,
135
+ "learning_rate": 0.0006970852628711522,
136
+ "loss": 3.2017,
137
+ "step": 340
138
+ },
139
+ {
140
+ "epoch": 0.19,
141
+ "learning_rate": 0.0006932034867883411,
142
+ "loss": 3.4474,
143
+ "step": 360
144
+ },
145
+ {
146
+ "epoch": 0.2,
147
+ "learning_rate": 0.0006891174067011714,
148
+ "loss": 3.0656,
149
+ "step": 380
150
+ },
151
+ {
152
+ "epoch": 0.21,
153
+ "learning_rate": 0.0006852356306183601,
154
+ "loss": 3.5014,
155
+ "step": 400
156
+ },
157
+ {
158
+ "epoch": 0.21,
159
+ "eval_loss": Infinity,
160
+ "eval_runtime": 1255.7751,
161
+ "eval_samples_per_second": 21.196,
162
+ "eval_steps_per_second": 0.663,
163
+ "eval_wer": 1.0,
164
+ "step": 400
165
+ },
166
+ {
167
+ "epoch": 0.22,
168
+ "learning_rate": 0.0006811495505311904,
169
+ "loss": 3.1757,
170
+ "step": 420
171
+ },
172
+ {
173
+ "epoch": 0.23,
174
+ "learning_rate": 0.0006770634704440206,
175
+ "loss": 3.2038,
176
+ "step": 440
177
+ },
178
+ {
179
+ "epoch": 0.25,
180
+ "learning_rate": 0.0006731816943612095,
181
+ "loss": 3.4072,
182
+ "step": 460
183
+ },
184
+ {
185
+ "epoch": 0.26,
186
+ "learning_rate": 0.0006690956142740398,
187
+ "loss": 3.0498,
188
+ "step": 480
189
+ },
190
+ {
191
+ "epoch": 0.27,
192
+ "learning_rate": 0.00066500953418687,
193
+ "loss": 3.331,
194
+ "step": 500
195
+ },
196
+ {
197
+ "epoch": 0.27,
198
+ "eval_loss": Infinity,
199
+ "eval_runtime": 1239.2922,
200
+ "eval_samples_per_second": 21.478,
201
+ "eval_steps_per_second": 0.671,
202
+ "eval_wer": 1.0,
203
+ "step": 500
204
+ },
205
+ {
206
+ "epoch": 0.28,
207
+ "learning_rate": 0.0006609234540997004,
208
+ "loss": 3.0833,
209
+ "step": 520
210
+ },
211
+ {
212
+ "epoch": 0.29,
213
+ "learning_rate": 0.0006568373740125306,
214
+ "loss": 3.1989,
215
+ "step": 540
216
+ },
217
+ {
218
+ "epoch": 0.3,
219
+ "learning_rate": 0.0006531599019340779,
220
+ "loss": 3.5001,
221
+ "step": 560
222
+ },
223
+ {
224
+ "epoch": 0.31,
225
+ "learning_rate": 0.0006490738218469082,
226
+ "loss": 3.0685,
227
+ "step": 580
228
+ },
229
+ {
230
+ "epoch": 0.32,
231
+ "learning_rate": 0.000645192045764097,
232
+ "loss": 3.4809,
233
+ "step": 600
234
+ },
235
+ {
236
+ "epoch": 0.32,
237
+ "eval_loss": Infinity,
238
+ "eval_runtime": 1232.3119,
239
+ "eval_samples_per_second": 21.6,
240
+ "eval_steps_per_second": 0.675,
241
+ "eval_wer": 1.0,
242
+ "step": 600
243
+ },
244
+ {
245
+ "epoch": 0.33,
246
+ "learning_rate": 0.0006411059656769273,
247
+ "loss": 3.1877,
248
+ "step": 620
249
+ },
250
+ {
251
+ "epoch": 0.34,
252
+ "learning_rate": 0.0006370198855897576,
253
+ "loss": 3.2365,
254
+ "step": 640
255
+ },
256
+ {
257
+ "epoch": 0.35,
258
+ "learning_rate": 0.0006333424135113049,
259
+ "loss": 3.67,
260
+ "step": 660
261
+ },
262
+ {
263
+ "epoch": 0.36,
264
+ "learning_rate": 0.0006292563334241351,
265
+ "loss": 3.086,
266
+ "step": 680
267
+ },
268
+ {
269
+ "epoch": 0.37,
270
+ "learning_rate": 0.0006253745573413239,
271
+ "loss": 3.4678,
272
+ "step": 700
273
+ },
274
+ {
275
+ "epoch": 0.37,
276
+ "eval_loss": Infinity,
277
+ "eval_runtime": 1236.8236,
278
+ "eval_samples_per_second": 21.521,
279
+ "eval_steps_per_second": 0.673,
280
+ "eval_wer": 1.0,
281
+ "step": 700
282
+ },
283
+ {
284
+ "epoch": 0.38,
285
+ "learning_rate": 0.0006212884772541542,
286
+ "loss": 3.2344,
287
+ "step": 720
288
+ },
289
+ {
290
+ "epoch": 0.4,
291
+ "learning_rate": 0.0006172023971669845,
292
+ "loss": 3.2507,
293
+ "step": 740
294
+ },
295
+ {
296
+ "epoch": 0.41,
297
+ "learning_rate": 0.0006131163170798147,
298
+ "loss": 3.4283,
299
+ "step": 760
300
+ },
301
+ {
302
+ "epoch": 0.42,
303
+ "learning_rate": 0.0006090302369926451,
304
+ "loss": 3.0802,
305
+ "step": 780
306
+ },
307
+ {
308
+ "epoch": 0.43,
309
+ "learning_rate": 0.0006051484609098339,
310
+ "loss": 3.4596,
311
+ "step": 800
312
+ },
313
+ {
314
+ "epoch": 0.43,
315
+ "eval_loss": Infinity,
316
+ "eval_runtime": 1233.785,
317
+ "eval_samples_per_second": 21.574,
318
+ "eval_steps_per_second": 0.674,
319
+ "eval_wer": 1.0,
320
+ "step": 800
321
+ },
322
+ {
323
+ "epoch": 0.44,
324
+ "learning_rate": 0.0006010623808226641,
325
+ "loss": 3.2227,
326
+ "step": 820
327
+ },
328
+ {
329
+ "epoch": 0.45,
330
+ "learning_rate": 0.0005969763007354945,
331
+ "loss": 3.2357,
332
+ "step": 840
333
+ },
334
+ {
335
+ "epoch": 0.46,
336
+ "learning_rate": 0.0005930945246526832,
337
+ "loss": 3.5563,
338
+ "step": 860
339
+ },
340
+ {
341
+ "epoch": 0.47,
342
+ "learning_rate": 0.0005890084445655135,
343
+ "loss": 3.0845,
344
+ "step": 880
345
+ },
346
+ {
347
+ "epoch": 0.48,
348
+ "learning_rate": 0.0005851266684827023,
349
+ "loss": 3.4644,
350
+ "step": 900
351
+ },
352
+ {
353
+ "epoch": 0.48,
354
+ "eval_loss": Infinity,
355
+ "eval_runtime": 1235.1488,
356
+ "eval_samples_per_second": 21.55,
357
+ "eval_steps_per_second": 0.674,
358
+ "eval_wer": 1.0,
359
+ "step": 900
360
+ },
361
+ {
362
+ "epoch": 0.49,
363
+ "learning_rate": 0.0005810405883955325,
364
+ "loss": 3.2092,
365
+ "step": 920
366
+ },
367
+ {
368
+ "epoch": 0.5,
369
+ "learning_rate": 0.0005769545083083629,
370
+ "loss": 3.2157,
371
+ "step": 940
372
+ },
373
+ {
374
+ "epoch": 0.51,
375
+ "learning_rate": 0.0005730727322255517,
376
+ "loss": 3.4679,
377
+ "step": 960
378
+ },
379
+ {
380
+ "epoch": 0.52,
381
+ "learning_rate": 0.0005689866521383819,
382
+ "loss": 3.0662,
383
+ "step": 980
384
+ },
385
+ {
386
+ "epoch": 0.53,
387
+ "learning_rate": 0.0005651048760555707,
388
+ "loss": 3.4671,
389
+ "step": 1000
390
+ },
391
+ {
392
+ "epoch": 0.53,
393
+ "eval_loss": Infinity,
394
+ "eval_runtime": 1241.2249,
395
+ "eval_samples_per_second": 21.445,
396
+ "eval_steps_per_second": 0.67,
397
+ "eval_wer": 1.0,
398
+ "step": 1000
399
+ },
400
+ {
401
+ "epoch": 0.54,
402
+ "learning_rate": 0.0005610187959684009,
403
+ "loss": 3.2428,
404
+ "step": 1020
405
+ },
406
+ {
407
+ "epoch": 0.56,
408
+ "learning_rate": 0.0005569327158812314,
409
+ "loss": 3.2165,
410
+ "step": 1040
411
+ },
412
+ {
413
+ "epoch": 0.57,
414
+ "learning_rate": 0.0005530509397984201,
415
+ "loss": 3.5106,
416
+ "step": 1060
417
+ },
418
+ {
419
+ "epoch": 0.58,
420
+ "learning_rate": 0.0005489648597112503,
421
+ "loss": 3.1137,
422
+ "step": 1080
423
+ },
424
+ {
425
+ "epoch": 0.59,
426
+ "learning_rate": 0.0005452873876327976,
427
+ "loss": 3.6005,
428
+ "step": 1100
429
+ },
430
+ {
431
+ "epoch": 0.59,
432
+ "eval_loss": Infinity,
433
+ "eval_runtime": 1236.2955,
434
+ "eval_samples_per_second": 21.53,
435
+ "eval_steps_per_second": 0.673,
436
+ "eval_wer": 1.0,
437
+ "step": 1100
438
+ },
439
+ {
440
+ "epoch": 0.6,
441
+ "learning_rate": 0.0005412013075456279,
442
+ "loss": 3.2702,
443
+ "step": 1120
444
+ },
445
+ {
446
+ "epoch": 0.61,
447
+ "learning_rate": 0.0005371152274584582,
448
+ "loss": 3.6768,
449
+ "step": 1140
450
+ },
451
+ {
452
+ "epoch": 0.62,
453
+ "learning_rate": 0.000533233451375647,
454
+ "loss": 3.6313,
455
+ "step": 1160
456
+ },
457
+ {
458
+ "epoch": 0.63,
459
+ "learning_rate": 0.0005291473712884773,
460
+ "loss": 3.2456,
461
+ "step": 1180
462
+ },
463
+ {
464
+ "epoch": 0.64,
465
+ "learning_rate": 0.000525265595205666,
466
+ "loss": 3.9182,
467
+ "step": 1200
468
+ },
469
+ {
470
+ "epoch": 0.64,
471
+ "eval_loss": Infinity,
472
+ "eval_runtime": 1229.4316,
473
+ "eval_samples_per_second": 21.651,
474
+ "eval_steps_per_second": 0.677,
475
+ "eval_wer": 1.0,
476
+ "step": 1200
477
+ },
478
+ {
479
+ "epoch": 0.65,
480
+ "learning_rate": 0.0005211795151184963,
481
+ "loss": 3.3805,
482
+ "step": 1220
483
+ },
484
+ {
485
+ "epoch": 0.66,
486
+ "learning_rate": 0.0005170934350313267,
487
+ "loss": 3.4687,
488
+ "step": 1240
489
+ },
490
+ {
491
+ "epoch": 0.67,
492
+ "learning_rate": 0.000513007354944157,
493
+ "loss": 3.4473,
494
+ "step": 1260
495
+ },
496
+ {
497
+ "epoch": 0.68,
498
+ "learning_rate": 0.0005089212748569872,
499
+ "loss": 3.2232,
500
+ "step": 1280
501
+ },
502
+ {
503
+ "epoch": 0.69,
504
+ "learning_rate": 0.000505039498774176,
505
+ "loss": 3.6466,
506
+ "step": 1300
507
+ },
508
+ {
509
+ "epoch": 0.69,
510
+ "eval_loss": Infinity,
511
+ "eval_runtime": 1237.3739,
512
+ "eval_samples_per_second": 21.512,
513
+ "eval_steps_per_second": 0.672,
514
+ "eval_wer": 1.0,
515
+ "step": 1300
516
+ },
517
+ {
518
+ "epoch": 0.7,
519
+ "learning_rate": 0.0005009534186870062,
520
+ "loss": 3.3496,
521
+ "step": 1320
522
+ },
523
+ {
524
+ "epoch": 0.72,
525
+ "learning_rate": 0.0004968673385998365,
526
+ "loss": 3.3918,
527
+ "step": 1340
528
+ },
529
+ {
530
+ "epoch": 0.73,
531
+ "learning_rate": 0.0004927812585126669,
532
+ "loss": 3.3232,
533
+ "step": 1360
534
+ },
535
+ {
536
+ "epoch": 0.74,
537
+ "learning_rate": 0.0004886951784254972,
538
+ "loss": 3.288,
539
+ "step": 1380
540
+ },
541
+ {
542
+ "epoch": 0.75,
543
+ "learning_rate": 0.00048481340234268587,
544
+ "loss": 3.6932,
545
+ "step": 1400
546
+ },
547
+ {
548
+ "epoch": 0.75,
549
+ "eval_loss": Infinity,
550
+ "eval_runtime": 1236.694,
551
+ "eval_samples_per_second": 21.524,
552
+ "eval_steps_per_second": 0.673,
553
+ "eval_wer": 1.0,
554
+ "step": 1400
555
+ },
556
+ {
557
+ "epoch": 0.76,
558
+ "learning_rate": 0.00048072732225551624,
559
+ "loss": 3.5411,
560
+ "step": 1420
561
+ },
562
+ {
563
+ "epoch": 0.77,
564
+ "learning_rate": 0.0004766412421683465,
565
+ "loss": 3.5658,
566
+ "step": 1440
567
+ },
568
+ {
569
+ "epoch": 0.78,
570
+ "learning_rate": 0.00047275946608553524,
571
+ "loss": 3.7666,
572
+ "step": 1460
573
+ },
574
+ {
575
+ "epoch": 0.79,
576
+ "learning_rate": 0.0004686733859983656,
577
+ "loss": 3.7564,
578
+ "step": 1480
579
+ },
580
+ {
581
+ "epoch": 0.8,
582
+ "learning_rate": 0.00046479160991555436,
583
+ "loss": 3.7939,
584
+ "step": 1500
585
+ },
586
+ {
587
+ "epoch": 0.8,
588
+ "eval_loss": Infinity,
589
+ "eval_runtime": 1235.7295,
590
+ "eval_samples_per_second": 21.54,
591
+ "eval_steps_per_second": 0.673,
592
+ "eval_wer": 1.0,
593
+ "step": 1500
594
+ },
595
+ {
596
+ "epoch": 0.81,
597
+ "learning_rate": 0.00046070552982838467,
598
+ "loss": 3.7471,
599
+ "step": 1520
600
+ },
601
+ {
602
+ "epoch": 0.82,
603
+ "learning_rate": 0.00045661944974121493,
604
+ "loss": 3.7457,
605
+ "step": 1540
606
+ },
607
+ {
608
+ "epoch": 0.83,
609
+ "learning_rate": 0.0004527376736584037,
610
+ "loss": 3.798,
611
+ "step": 1560
612
+ },
613
+ {
614
+ "epoch": 0.84,
615
+ "learning_rate": 0.00044865159357123405,
616
+ "loss": 3.7611,
617
+ "step": 1580
618
+ },
619
+ {
620
+ "epoch": 0.85,
621
+ "learning_rate": 0.0004447698174884228,
622
+ "loss": 3.9284,
623
+ "step": 1600
624
+ },
625
+ {
626
+ "epoch": 0.85,
627
+ "eval_loss": Infinity,
628
+ "eval_runtime": 1251.8158,
629
+ "eval_samples_per_second": 21.264,
630
+ "eval_steps_per_second": 0.665,
631
+ "eval_wer": 1.0,
632
+ "step": 1600
633
+ },
634
+ {
635
+ "epoch": 0.86,
636
+ "learning_rate": 0.0004406837374012531,
637
+ "loss": 3.7416,
638
+ "step": 1620
639
+ },
640
+ {
641
+ "epoch": 0.88,
642
+ "learning_rate": 0.00043659765731408337,
643
+ "loss": 3.7416,
644
+ "step": 1640
645
+ },
646
+ {
647
+ "epoch": 0.89,
648
+ "learning_rate": 0.0004327158812312721,
649
+ "loss": 3.8024,
650
+ "step": 1660
651
+ },
652
+ {
653
+ "epoch": 0.9,
654
+ "learning_rate": 0.0004286298011441025,
655
+ "loss": 3.7578,
656
+ "step": 1680
657
+ },
658
+ {
659
+ "epoch": 0.91,
660
+ "learning_rate": 0.0004247480250612912,
661
+ "loss": 3.7859,
662
+ "step": 1700
663
+ },
664
+ {
665
+ "epoch": 0.91,
666
+ "eval_loss": Infinity,
667
+ "eval_runtime": 1235.734,
668
+ "eval_samples_per_second": 21.54,
669
+ "eval_steps_per_second": 0.673,
670
+ "eval_wer": 1.0,
671
+ "step": 1700
672
+ },
673
+ {
674
+ "epoch": 0.92,
675
+ "learning_rate": 0.0004206619449741215,
676
+ "loss": 3.7427,
677
+ "step": 1720
678
+ },
679
+ {
680
+ "epoch": 0.93,
681
+ "learning_rate": 0.0004165758648869518,
682
+ "loss": 3.751,
683
+ "step": 1740
684
+ },
685
+ {
686
+ "epoch": 0.94,
687
+ "learning_rate": 0.00041248978479978206,
688
+ "loss": 3.6701,
689
+ "step": 1760
690
+ },
691
+ {
692
+ "epoch": 0.95,
693
+ "learning_rate": 0.0004084037047126123,
694
+ "loss": 3.7613,
695
+ "step": 1780
696
+ },
697
+ {
698
+ "epoch": 0.96,
699
+ "learning_rate": 0.00040472623263415966,
700
+ "loss": 3.9363,
701
+ "step": 1800
702
+ },
703
+ {
704
+ "epoch": 0.96,
705
+ "eval_loss": Infinity,
706
+ "eval_runtime": 1236.9158,
707
+ "eval_samples_per_second": 21.52,
708
+ "eval_steps_per_second": 0.673,
709
+ "eval_wer": 1.0,
710
+ "step": 1800
711
+ },
712
+ {
713
+ "epoch": 0.97,
714
+ "learning_rate": 0.0004006401525469899,
715
+ "loss": 3.7464,
716
+ "step": 1820
717
+ },
718
+ {
719
+ "epoch": 0.98,
720
+ "learning_rate": 0.00039655407245982023,
721
+ "loss": 3.7414,
722
+ "step": 1840
723
+ },
724
+ {
725
+ "epoch": 0.99,
726
+ "learning_rate": 0.000392672296377009,
727
+ "loss": 3.8036,
728
+ "step": 1860
729
+ },
730
+ {
731
+ "epoch": 1.0,
732
+ "learning_rate": 0.00038879052029419783,
733
+ "loss": 3.8768,
734
+ "step": 1880
735
+ },
736
+ {
737
+ "epoch": 1.01,
738
+ "learning_rate": 0.0003847044402070281,
739
+ "loss": 3.7573,
740
+ "step": 1900
741
+ },
742
+ {
743
+ "epoch": 1.01,
744
+ "eval_loss": Infinity,
745
+ "eval_runtime": 1232.3535,
746
+ "eval_samples_per_second": 21.599,
747
+ "eval_steps_per_second": 0.675,
748
+ "eval_wer": 1.0,
749
+ "step": 1900
750
+ },
751
+ {
752
+ "epoch": 1.03,
753
+ "learning_rate": 0.00038061836011985835,
754
+ "loss": 3.6959,
755
+ "step": 1920
756
+ },
757
+ {
758
+ "epoch": 1.04,
759
+ "learning_rate": 0.00037653228003268867,
760
+ "loss": 3.7313,
761
+ "step": 1940
762
+ },
763
+ {
764
+ "epoch": 1.05,
765
+ "learning_rate": 0.00037244619994551893,
766
+ "loss": 3.7676,
767
+ "step": 1960
768
+ },
769
+ {
770
+ "epoch": 1.06,
771
+ "learning_rate": 0.0003685644238627077,
772
+ "loss": 3.7872,
773
+ "step": 1980
774
+ },
775
+ {
776
+ "epoch": 1.07,
777
+ "learning_rate": 0.000364478343775538,
778
+ "loss": 3.7553,
779
+ "step": 2000
780
+ },
781
+ {
782
+ "epoch": 1.07,
783
+ "eval_loss": Infinity,
784
+ "eval_runtime": 1239.2247,
785
+ "eval_samples_per_second": 21.48,
786
+ "eval_steps_per_second": 0.671,
787
+ "eval_wer": 1.0,
788
+ "step": 2000
789
+ },
790
+ {
791
+ "epoch": 1.08,
792
+ "learning_rate": 0.0003603922636883683,
793
+ "loss": 3.694,
794
+ "step": 2020
795
+ },
796
+ {
797
+ "epoch": 1.09,
798
+ "learning_rate": 0.0003565104876055571,
799
+ "loss": 3.8389,
800
+ "step": 2040
801
+ },
802
+ {
803
+ "epoch": 1.1,
804
+ "learning_rate": 0.00035242440751838736,
805
+ "loss": 3.7573,
806
+ "step": 2060
807
+ },
808
+ {
809
+ "epoch": 1.11,
810
+ "learning_rate": 0.0003483383274312177,
811
+ "loss": 3.6641,
812
+ "step": 2080
813
+ },
814
+ {
815
+ "epoch": 1.12,
816
+ "learning_rate": 0.00034425224734404794,
817
+ "loss": 3.7606,
818
+ "step": 2100
819
+ },
820
+ {
821
+ "epoch": 1.12,
822
+ "eval_loss": Infinity,
823
+ "eval_runtime": 1237.7997,
824
+ "eval_samples_per_second": 21.504,
825
+ "eval_steps_per_second": 0.672,
826
+ "eval_wer": 1.0,
827
+ "step": 2100
828
+ },
829
+ {
830
+ "epoch": 1.13,
831
+ "learning_rate": 0.00034016616725687825,
832
+ "loss": 3.6827,
833
+ "step": 2120
834
+ },
835
+ {
836
+ "epoch": 1.14,
837
+ "learning_rate": 0.000336284391174067,
838
+ "loss": 3.8375,
839
+ "step": 2140
840
+ },
841
+ {
842
+ "epoch": 1.15,
843
+ "learning_rate": 0.0003321983110868973,
844
+ "loss": 3.7527,
845
+ "step": 2160
846
+ },
847
+ {
848
+ "epoch": 1.16,
849
+ "learning_rate": 0.0003283165350040861,
850
+ "loss": 3.7881,
851
+ "step": 2180
852
+ },
853
+ {
854
+ "epoch": 1.17,
855
+ "learning_rate": 0.0003242304549169164,
856
+ "loss": 3.7514,
857
+ "step": 2200
858
+ },
859
+ {
860
+ "epoch": 1.17,
861
+ "eval_loss": Infinity,
862
+ "eval_runtime": 1237.076,
863
+ "eval_samples_per_second": 21.517,
864
+ "eval_steps_per_second": 0.673,
865
+ "eval_wer": 1.0,
866
+ "step": 2200
867
+ },
868
+ {
869
+ "epoch": 1.19,
870
+ "learning_rate": 0.0003201443748297467,
871
+ "loss": 3.6727,
872
+ "step": 2220
873
+ },
874
+ {
875
+ "epoch": 1.2,
876
+ "learning_rate": 0.00031626259874693543,
877
+ "loss": 3.8424,
878
+ "step": 2240
879
+ },
880
+ {
881
+ "epoch": 1.21,
882
+ "learning_rate": 0.00031217651865976574,
883
+ "loss": 3.7594,
884
+ "step": 2260
885
+ },
886
+ {
887
+ "epoch": 1.22,
888
+ "learning_rate": 0.00030829474257695454,
889
+ "loss": 3.7862,
890
+ "step": 2280
891
+ },
892
+ {
893
+ "epoch": 1.23,
894
+ "learning_rate": 0.0003042086624897848,
895
+ "loss": 3.7472,
896
+ "step": 2300
897
+ },
898
+ {
899
+ "epoch": 1.23,
900
+ "eval_loss": Infinity,
901
+ "eval_runtime": 1236.6319,
902
+ "eval_samples_per_second": 21.525,
903
+ "eval_steps_per_second": 0.673,
904
+ "eval_wer": 1.0,
905
+ "step": 2300
906
+ },
907
+ {
908
+ "epoch": 1.24,
909
+ "learning_rate": 0.0003001225824026151,
910
+ "loss": 3.6956,
911
+ "step": 2320
912
+ },
913
+ {
914
+ "epoch": 1.25,
915
+ "learning_rate": 0.0002960365023154454,
916
+ "loss": 3.7248,
917
+ "step": 2340
918
+ },
919
+ {
920
+ "epoch": 1.26,
921
+ "learning_rate": 0.0002919504222282757,
922
+ "loss": 3.7531,
923
+ "step": 2360
924
+ },
925
+ {
926
+ "epoch": 1.27,
927
+ "learning_rate": 0.0002882729501498229,
928
+ "loss": 3.9245,
929
+ "step": 2380
930
+ },
931
+ {
932
+ "epoch": 1.28,
933
+ "learning_rate": 0.00028418687006265324,
934
+ "loss": 3.7478,
935
+ "step": 2400
936
+ },
937
+ {
938
+ "epoch": 1.28,
939
+ "eval_loss": Infinity,
940
+ "eval_runtime": 1236.7725,
941
+ "eval_samples_per_second": 21.522,
942
+ "eval_steps_per_second": 0.673,
943
+ "eval_wer": 1.0,
944
+ "step": 2400
945
+ },
946
+ {
947
+ "epoch": 1.29,
948
+ "learning_rate": 0.00028010078997548355,
949
+ "loss": 3.6826,
950
+ "step": 2420
951
+ },
952
+ {
953
+ "epoch": 1.3,
954
+ "learning_rate": 0.0002760147098883138,
955
+ "loss": 3.7236,
956
+ "step": 2440
957
+ },
958
+ {
959
+ "epoch": 1.31,
960
+ "learning_rate": 0.00027192862980114413,
961
+ "loss": 3.7609,
962
+ "step": 2460
963
+ },
964
+ {
965
+ "epoch": 1.32,
966
+ "learning_rate": 0.00026804685371833287,
967
+ "loss": 3.7846,
968
+ "step": 2480
969
+ },
970
+ {
971
+ "epoch": 1.33,
972
+ "learning_rate": 0.0002639607736311632,
973
+ "loss": 3.7496,
974
+ "step": 2500
975
+ },
976
+ {
977
+ "epoch": 1.33,
978
+ "eval_loss": Infinity,
979
+ "eval_runtime": 1233.5496,
980
+ "eval_samples_per_second": 21.578,
981
+ "eval_steps_per_second": 0.674,
982
+ "eval_wer": 1.0,
983
+ "step": 2500
984
+ },
985
+ {
986
+ "epoch": 1.35,
987
+ "learning_rate": 0.00025987469354399345,
988
+ "loss": 3.6785,
989
+ "step": 2520
990
+ },
991
+ {
992
+ "epoch": 1.36,
993
+ "learning_rate": 0.00025578861345682376,
994
+ "loss": 3.7212,
995
+ "step": 2540
996
+ },
997
+ {
998
+ "epoch": 1.37,
999
+ "learning_rate": 0.0002517025333696541,
1000
+ "loss": 3.7637,
1001
+ "step": 2560
1002
+ },
1003
+ {
1004
+ "epoch": 1.38,
1005
+ "learning_rate": 0.00024761645328248434,
1006
+ "loss": 3.6513,
1007
+ "step": 2580
1008
+ },
1009
+ {
1010
+ "epoch": 1.39,
1011
+ "learning_rate": 0.00024353037319531465,
1012
+ "loss": 3.7513,
1013
+ "step": 2600
1014
+ },
1015
+ {
1016
+ "epoch": 1.39,
1017
+ "eval_loss": Infinity,
1018
+ "eval_runtime": 1231.8132,
1019
+ "eval_samples_per_second": 21.609,
1020
+ "eval_steps_per_second": 0.675,
1021
+ "eval_wer": 1.0,
1022
+ "step": 2600
1023
+ },
1024
+ {
1025
+ "epoch": 1.4,
1026
+ "learning_rate": 0.00023944429310814494,
1027
+ "loss": 3.6818,
1028
+ "step": 2620
1029
+ },
1030
+ {
1031
+ "epoch": 1.41,
1032
+ "learning_rate": 0.00023556251702533368,
1033
+ "loss": 3.8417,
1034
+ "step": 2640
1035
+ },
1036
+ {
1037
+ "epoch": 1.42,
1038
+ "learning_rate": 0.000231476436938164,
1039
+ "loss": 3.7639,
1040
+ "step": 2660
1041
+ },
1042
+ {
1043
+ "epoch": 1.43,
1044
+ "learning_rate": 0.00022759466085535277,
1045
+ "loss": 3.7842,
1046
+ "step": 2680
1047
+ },
1048
+ {
1049
+ "epoch": 1.44,
1050
+ "learning_rate": 0.00022350858076818309,
1051
+ "loss": 3.7497,
1052
+ "step": 2700
1053
+ },
1054
+ {
1055
+ "epoch": 1.44,
1056
+ "eval_loss": Infinity,
1057
+ "eval_runtime": 1234.8437,
1058
+ "eval_samples_per_second": 21.556,
1059
+ "eval_steps_per_second": 0.674,
1060
+ "eval_wer": 1.0,
1061
+ "step": 2700
1062
+ },
1063
+ {
1064
+ "epoch": 1.45,
1065
+ "learning_rate": 0.00021942250068101335,
1066
+ "loss": 3.6795,
1067
+ "step": 2720
1068
+ },
1069
+ {
1070
+ "epoch": 1.46,
1071
+ "learning_rate": 0.00021554072459820212,
1072
+ "loss": 3.8421,
1073
+ "step": 2740
1074
+ },
1075
+ {
1076
+ "epoch": 1.47,
1077
+ "learning_rate": 0.00021145464451103243,
1078
+ "loss": 3.7614,
1079
+ "step": 2760
1080
+ },
1081
+ {
1082
+ "epoch": 1.48,
1083
+ "learning_rate": 0.0002075728684282212,
1084
+ "loss": 3.9259,
1085
+ "step": 2780
1086
+ },
1087
+ {
1088
+ "epoch": 1.49,
1089
+ "learning_rate": 0.00020348678834105147,
1090
+ "loss": 3.7539,
1091
+ "step": 2800
1092
+ },
1093
+ {
1094
+ "epoch": 1.49,
1095
+ "eval_loss": Infinity,
1096
+ "eval_runtime": 1240.1103,
1097
+ "eval_samples_per_second": 21.464,
1098
+ "eval_steps_per_second": 0.671,
1099
+ "eval_wer": 1.0,
1100
+ "step": 2800
1101
+ },
1102
+ {
1103
+ "epoch": 1.51,
1104
+ "learning_rate": 0.00019940070825388178,
1105
+ "loss": 3.6889,
1106
+ "step": 2820
1107
+ },
1108
+ {
1109
+ "epoch": 1.52,
1110
+ "learning_rate": 0.00019551893217107055,
1111
+ "loss": 3.843,
1112
+ "step": 2840
1113
+ },
1114
+ {
1115
+ "epoch": 1.53,
1116
+ "learning_rate": 0.00019143285208390087,
1117
+ "loss": 3.7666,
1118
+ "step": 2860
1119
+ },
1120
+ {
1121
+ "epoch": 1.54,
1122
+ "learning_rate": 0.00018755107600108964,
1123
+ "loss": 3.786,
1124
+ "step": 2880
1125
+ },
1126
+ {
1127
+ "epoch": 1.55,
1128
+ "learning_rate": 0.0001834649959139199,
1129
+ "loss": 3.7581,
1130
+ "step": 2900
1131
+ },
1132
+ {
1133
+ "epoch": 1.55,
1134
+ "eval_loss": Infinity,
1135
+ "eval_runtime": 1231.1699,
1136
+ "eval_samples_per_second": 21.62,
1137
+ "eval_steps_per_second": 0.676,
1138
+ "eval_wer": 1.0,
1139
+ "step": 2900
1140
+ },
1141
+ {
1142
+ "epoch": 1.56,
1143
+ "learning_rate": 0.00017937891582675021,
1144
+ "loss": 3.672,
1145
+ "step": 2920
1146
+ },
1147
+ {
1148
+ "epoch": 1.57,
1149
+ "learning_rate": 0.00017549713974393899,
1150
+ "loss": 3.8539,
1151
+ "step": 2940
1152
+ },
1153
+ {
1154
+ "epoch": 1.58,
1155
+ "learning_rate": 0.00017141105965676927,
1156
+ "loss": 3.7652,
1157
+ "step": 2960
1158
+ },
1159
+ {
1160
+ "epoch": 1.59,
1161
+ "learning_rate": 0.00016752928357395807,
1162
+ "loss": 3.7825,
1163
+ "step": 2980
1164
+ },
1165
+ {
1166
+ "epoch": 1.6,
1167
+ "learning_rate": 0.00016344320348678833,
1168
+ "loss": 3.7572,
1169
+ "step": 3000
1170
+ },
1171
+ {
1172
+ "epoch": 1.6,
1173
+ "eval_loss": Infinity,
1174
+ "eval_runtime": 1241.8499,
1175
+ "eval_samples_per_second": 21.434,
1176
+ "eval_steps_per_second": 0.67,
1177
+ "eval_wer": 1.0,
1178
+ "step": 3000
1179
+ },
1180
+ {
1181
+ "epoch": 1.61,
1182
+ "learning_rate": 0.00015935712339961862,
1183
+ "loss": 3.6906,
1184
+ "step": 3020
1185
+ },
1186
+ {
1187
+ "epoch": 1.62,
1188
+ "learning_rate": 0.00015547534731680742,
1189
+ "loss": 3.845,
1190
+ "step": 3040
1191
+ },
1192
+ {
1193
+ "epoch": 1.63,
1194
+ "learning_rate": 0.0001513892672296377,
1195
+ "loss": 3.763,
1196
+ "step": 3060
1197
+ },
1198
+ {
1199
+ "epoch": 1.64,
1200
+ "learning_rate": 0.00014750749114682648,
1201
+ "loss": 3.779,
1202
+ "step": 3080
1203
+ },
1204
+ {
1205
+ "epoch": 1.66,
1206
+ "learning_rate": 0.0001434214110596568,
1207
+ "loss": 3.7589,
1208
+ "step": 3100
1209
+ },
1210
+ {
1211
+ "epoch": 1.66,
1212
+ "eval_loss": Infinity,
1213
+ "eval_runtime": 1236.916,
1214
+ "eval_samples_per_second": 21.52,
1215
+ "eval_steps_per_second": 0.673,
1216
+ "eval_wer": 1.0,
1217
+ "step": 3100
1218
+ },
1219
+ {
1220
+ "epoch": 1.67,
1221
+ "learning_rate": 0.00013933533097248705,
1222
+ "loss": 3.6948,
1223
+ "step": 3120
1224
+ },
1225
+ {
1226
+ "epoch": 1.68,
1227
+ "learning_rate": 0.00013524925088531734,
1228
+ "loss": 3.7211,
1229
+ "step": 3140
1230
+ },
1231
+ {
1232
+ "epoch": 1.69,
1233
+ "learning_rate": 0.00013116317079814763,
1234
+ "loss": 3.7588,
1235
+ "step": 3160
1236
+ },
1237
+ {
1238
+ "epoch": 1.7,
1239
+ "learning_rate": 0.00012728139471533643,
1240
+ "loss": 3.7898,
1241
+ "step": 3180
1242
+ },
1243
+ {
1244
+ "epoch": 1.71,
1245
+ "learning_rate": 0.00012319531462816672,
1246
+ "loss": 3.7592,
1247
+ "step": 3200
1248
+ },
1249
+ {
1250
+ "epoch": 1.71,
1251
+ "eval_loss": Infinity,
1252
+ "eval_runtime": 1232.1567,
1253
+ "eval_samples_per_second": 21.603,
1254
+ "eval_steps_per_second": 0.675,
1255
+ "eval_wer": 1.0,
1256
+ "step": 3200
1257
+ },
1258
+ {
1259
+ "epoch": 1.72,
1260
+ "learning_rate": 0.000119109234540997,
1261
+ "loss": 3.688,
1262
+ "step": 3220
1263
+ },
1264
+ {
1265
+ "epoch": 1.73,
1266
+ "learning_rate": 0.00011522745845818579,
1267
+ "loss": 3.8419,
1268
+ "step": 3240
1269
+ },
1270
+ {
1271
+ "epoch": 1.74,
1272
+ "learning_rate": 0.00011114137837101608,
1273
+ "loss": 3.7578,
1274
+ "step": 3260
1275
+ },
1276
+ {
1277
+ "epoch": 1.75,
1278
+ "learning_rate": 0.00010705529828384636,
1279
+ "loss": 3.6534,
1280
+ "step": 3280
1281
+ },
1282
+ {
1283
+ "epoch": 1.76,
1284
+ "learning_rate": 0.00010296921819667667,
1285
+ "loss": 3.7531,
1286
+ "step": 3300
1287
+ },
1288
+ {
1289
+ "epoch": 1.76,
1290
+ "eval_loss": Infinity,
1291
+ "eval_runtime": 1237.4955,
1292
+ "eval_samples_per_second": 21.51,
1293
+ "eval_steps_per_second": 0.672,
1294
+ "eval_wer": 1.0,
1295
+ "step": 3300
1296
+ },
1297
+ {
1298
+ "epoch": 1.77,
1299
+ "learning_rate": 9.888313810950695e-05,
1300
+ "loss": 3.691,
1301
+ "step": 3320
1302
+ },
1303
+ {
1304
+ "epoch": 1.78,
1305
+ "learning_rate": 9.500136202669572e-05,
1306
+ "loss": 3.8424,
1307
+ "step": 3340
1308
+ },
1309
+ {
1310
+ "epoch": 1.79,
1311
+ "learning_rate": 9.091528193952601e-05,
1312
+ "loss": 3.7698,
1313
+ "step": 3360
1314
+ },
1315
+ {
1316
+ "epoch": 1.8,
1317
+ "learning_rate": 8.70335058567148e-05,
1318
+ "loss": 3.7836,
1319
+ "step": 3380
1320
+ },
1321
+ {
1322
+ "epoch": 1.82,
1323
+ "learning_rate": 8.294742576954509e-05,
1324
+ "loss": 3.7567,
1325
+ "step": 3400
1326
+ },
1327
+ {
1328
+ "epoch": 1.82,
1329
+ "eval_loss": Infinity,
1330
+ "eval_runtime": 1242.7888,
1331
+ "eval_samples_per_second": 21.418,
1332
+ "eval_steps_per_second": 0.669,
1333
+ "eval_wer": 1.0,
1334
+ "step": 3400
1335
+ },
1336
+ {
1337
+ "epoch": 1.83,
1338
+ "learning_rate": 7.886134568237537e-05,
1339
+ "loss": 3.6741,
1340
+ "step": 3420
1341
+ },
1342
+ {
1343
+ "epoch": 1.84,
1344
+ "learning_rate": 7.497956959956416e-05,
1345
+ "loss": 3.8442,
1346
+ "step": 3440
1347
+ },
1348
+ {
1349
+ "epoch": 1.85,
1350
+ "learning_rate": 7.089348951239445e-05,
1351
+ "loss": 3.7663,
1352
+ "step": 3460
1353
+ },
1354
+ {
1355
+ "epoch": 1.86,
1356
+ "learning_rate": 6.701171342958322e-05,
1357
+ "loss": 3.7789,
1358
+ "step": 3480
1359
+ },
1360
+ {
1361
+ "epoch": 1.87,
1362
+ "learning_rate": 6.292563334241352e-05,
1363
+ "loss": 3.7613,
1364
+ "step": 3500
1365
+ },
1366
+ {
1367
+ "epoch": 1.87,
1368
+ "eval_loss": Infinity,
1369
+ "eval_runtime": 1244.1733,
1370
+ "eval_samples_per_second": 21.394,
1371
+ "eval_steps_per_second": 0.669,
1372
+ "eval_wer": 1.0,
1373
+ "step": 3500
1374
+ },
1375
+ {
1376
+ "epoch": 1.88,
1377
+ "learning_rate": 5.88395532552438e-05,
1378
+ "loss": 3.6884,
1379
+ "step": 3520
1380
+ },
1381
+ {
1382
+ "epoch": 1.89,
1383
+ "learning_rate": 5.4957777172432585e-05,
1384
+ "loss": 3.8268,
1385
+ "step": 3540
1386
+ },
1387
+ {
1388
+ "epoch": 1.9,
1389
+ "learning_rate": 5.0871697085262866e-05,
1390
+ "loss": 3.7517,
1391
+ "step": 3560
1392
+ },
1393
+ {
1394
+ "epoch": 1.91,
1395
+ "learning_rate": 4.6989921002451645e-05,
1396
+ "loss": 3.7802,
1397
+ "step": 3580
1398
+ },
1399
+ {
1400
+ "epoch": 1.92,
1401
+ "learning_rate": 4.290384091528194e-05,
1402
+ "loss": 3.7516,
1403
+ "step": 3600
1404
+ },
1405
+ {
1406
+ "epoch": 1.92,
1407
+ "eval_loss": Infinity,
1408
+ "eval_runtime": 1253.9236,
1409
+ "eval_samples_per_second": 21.228,
1410
+ "eval_steps_per_second": 0.664,
1411
+ "eval_wer": 1.0,
1412
+ "step": 3600
1413
+ },
1414
+ {
1415
+ "epoch": 1.93,
1416
+ "learning_rate": 3.8817760828112234e-05,
1417
+ "loss": 3.6719,
1418
+ "step": 3620
1419
+ },
1420
+ {
1421
+ "epoch": 1.94,
1422
+ "learning_rate": 3.4935984745301006e-05,
1423
+ "loss": 3.8471,
1424
+ "step": 3640
1425
+ },
1426
+ {
1427
+ "epoch": 1.95,
1428
+ "learning_rate": 3.08499046581313e-05,
1429
+ "loss": 3.7568,
1430
+ "step": 3660
1431
+ },
1432
+ {
1433
+ "epoch": 1.96,
1434
+ "learning_rate": 2.6968128575320075e-05,
1435
+ "loss": 3.7887,
1436
+ "step": 3680
1437
+ },
1438
+ {
1439
+ "epoch": 1.98,
1440
+ "learning_rate": 2.2882048488150366e-05,
1441
+ "loss": 3.7581,
1442
+ "step": 3700
1443
+ },
1444
+ {
1445
+ "epoch": 1.98,
1446
+ "eval_loss": Infinity,
1447
+ "eval_runtime": 1242.0033,
1448
+ "eval_samples_per_second": 21.432,
1449
+ "eval_steps_per_second": 0.67,
1450
+ "eval_wer": 1.0,
1451
+ "step": 3700
1452
+ }
1453
+ ],
1454
+ "max_steps": 3746,
1455
+ "num_train_epochs": 2,
1456
+ "total_flos": 7.120999649529697e+19,
1457
+ "trial_name": null,
1458
+ "trial_params": null
1459
+ }
{checkpoint-400 β†’ checkpoint-3700}/training_args.bin RENAMED
File without changes
checkpoint-400/rng_state.pth DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:5557a0280e45c20fd6c512a9059337e2e4df99cce21cc1085d3606bd7275bbaf
3
- size 14567
 
 
 
 
checkpoint-400/scheduler.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:8d50355695b5459a6f341aeea7980dab38b7e889fa5ae575cb66ccb34c54e1a8
3
- size 623
 
 
 
 
checkpoint-400/trainer_state.json DELETED
@@ -1,172 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 0.2135326304550914,
5
- "global_step": 400,
6
- "is_hyper_param_search": false,
7
- "is_local_process_zero": true,
8
- "is_world_process_zero": true,
9
- "log_history": [
10
- {
11
- "epoch": 0.01,
12
- "learning_rate": 0.0002,
13
- "loss": 10.6012,
14
- "step": 20
15
- },
16
- {
17
- "epoch": 0.02,
18
- "learning_rate": 0.0004,
19
- "loss": 3.651,
20
- "step": 40
21
- },
22
- {
23
- "epoch": 0.03,
24
- "learning_rate": 0.00059,
25
- "loss": 3.3911,
26
- "step": 60
27
- },
28
- {
29
- "epoch": 0.04,
30
- "learning_rate": 0.0007491827839825661,
31
- "loss": 3.0481,
32
- "step": 80
33
- },
34
- {
35
- "epoch": 0.05,
36
- "learning_rate": 0.0007453010078997548,
37
- "loss": 3.4039,
38
- "step": 100
39
- },
40
- {
41
- "epoch": 0.05,
42
- "eval_loss": Infinity,
43
- "eval_runtime": 1260.1628,
44
- "eval_samples_per_second": 21.123,
45
- "eval_steps_per_second": 0.66,
46
- "eval_wer": 1.0,
47
- "step": 100
48
- },
49
- {
50
- "epoch": 0.06,
51
- "learning_rate": 0.0007412149278125851,
52
- "loss": 3.0553,
53
- "step": 120
54
- },
55
- {
56
- "epoch": 0.07,
57
- "learning_rate": 0.0007371288477254155,
58
- "loss": 3.2191,
59
- "step": 140
60
- },
61
- {
62
- "epoch": 0.09,
63
- "learning_rate": 0.0007332470716426042,
64
- "loss": 3.4094,
65
- "step": 160
66
- },
67
- {
68
- "epoch": 0.1,
69
- "learning_rate": 0.0007291609915554345,
70
- "loss": 3.0607,
71
- "step": 180
72
- },
73
- {
74
- "epoch": 0.11,
75
- "learning_rate": 0.0007252792154726233,
76
- "loss": 3.4396,
77
- "step": 200
78
- },
79
- {
80
- "epoch": 0.11,
81
- "eval_loss": Infinity,
82
- "eval_runtime": 1235.1686,
83
- "eval_samples_per_second": 21.55,
84
- "eval_steps_per_second": 0.674,
85
- "eval_wer": 1.0,
86
- "step": 200
87
- },
88
- {
89
- "epoch": 0.12,
90
- "learning_rate": 0.0007211931353854536,
91
- "loss": 3.1198,
92
- "step": 220
93
- },
94
- {
95
- "epoch": 0.13,
96
- "learning_rate": 0.0007171070552982838,
97
- "loss": 3.1801,
98
- "step": 240
99
- },
100
- {
101
- "epoch": 0.14,
102
- "learning_rate": 0.0007132252792154726,
103
- "loss": 3.3954,
104
- "step": 260
105
- },
106
- {
107
- "epoch": 0.15,
108
- "learning_rate": 0.000709139199128303,
109
- "loss": 3.0566,
110
- "step": 280
111
- },
112
- {
113
- "epoch": 0.16,
114
- "learning_rate": 0.0007052574230454917,
115
- "loss": 3.483,
116
- "step": 300
117
- },
118
- {
119
- "epoch": 0.16,
120
- "eval_loss": Infinity,
121
- "eval_runtime": 1262.2253,
122
- "eval_samples_per_second": 21.088,
123
- "eval_steps_per_second": 0.659,
124
- "eval_wer": 1.0,
125
- "step": 300
126
- },
127
- {
128
- "epoch": 0.17,
129
- "learning_rate": 0.000701171342958322,
130
- "loss": 3.1616,
131
- "step": 320
132
- },
133
- {
134
- "epoch": 0.18,
135
- "learning_rate": 0.0006970852628711522,
136
- "loss": 3.2017,
137
- "step": 340
138
- },
139
- {
140
- "epoch": 0.19,
141
- "learning_rate": 0.0006932034867883411,
142
- "loss": 3.4474,
143
- "step": 360
144
- },
145
- {
146
- "epoch": 0.2,
147
- "learning_rate": 0.0006891174067011714,
148
- "loss": 3.0656,
149
- "step": 380
150
- },
151
- {
152
- "epoch": 0.21,
153
- "learning_rate": 0.0006852356306183601,
154
- "loss": 3.5014,
155
- "step": 400
156
- },
157
- {
158
- "epoch": 0.21,
159
- "eval_loss": Infinity,
160
- "eval_runtime": 1255.7751,
161
- "eval_samples_per_second": 21.196,
162
- "eval_steps_per_second": 0.663,
163
- "eval_wer": 1.0,
164
- "step": 400
165
- }
166
- ],
167
- "max_steps": 3746,
168
- "num_train_epochs": 2,
169
- "total_flos": 7.69711092121344e+18,
170
- "trial_name": null,
171
- "trial_params": null
172
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoint-500/config.json DELETED
@@ -1,115 +0,0 @@
1
- {
2
- "_name_or_path": "KBLab/wav2vec2-large-voxrex",
3
- "activation_dropout": 0.1,
4
- "adapter_kernel_size": 3,
5
- "adapter_stride": 2,
6
- "add_adapter": false,
7
- "apply_spec_augment": true,
8
- "architectures": [
9
- "Wav2Vec2ForCTC"
10
- ],
11
- "attention_dropout": 0.0,
12
- "bos_token_id": 1,
13
- "classifier_proj_size": 256,
14
- "codevector_dim": 768,
15
- "contrastive_logits_temperature": 0.1,
16
- "conv_bias": true,
17
- "conv_dim": [
18
- 512,
19
- 512,
20
- 512,
21
- 512,
22
- 512,
23
- 512,
24
- 512
25
- ],
26
- "conv_kernel": [
27
- 10,
28
- 3,
29
- 3,
30
- 3,
31
- 3,
32
- 2,
33
- 2
34
- ],
35
- "conv_stride": [
36
- 5,
37
- 2,
38
- 2,
39
- 2,
40
- 2,
41
- 2,
42
- 2
43
- ],
44
- "ctc_loss_reduction": "mean",
45
- "ctc_zero_infinity": false,
46
- "diversity_loss_weight": 0.1,
47
- "do_stable_layer_norm": true,
48
- "eos_token_id": 2,
49
- "feat_extract_activation": "gelu",
50
- "feat_extract_dropout": 0.0,
51
- "feat_extract_norm": "layer",
52
- "feat_proj_dropout": 0.0,
53
- "feat_quantizer_dropout": 0.0,
54
- "final_dropout": 0.0,
55
- "hidden_act": "gelu",
56
- "hidden_dropout": 0.0,
57
- "hidden_size": 1024,
58
- "initializer_range": 0.02,
59
- "intermediate_size": 4096,
60
- "layer_norm_eps": 1e-05,
61
- "layerdrop": 0.0,
62
- "mask_channel_length": 10,
63
- "mask_channel_min_space": 1,
64
- "mask_channel_other": 0.0,
65
- "mask_channel_prob": 0.0,
66
- "mask_channel_selection": "static",
67
- "mask_feature_length": 64,
68
- "mask_feature_min_masks": 0,
69
- "mask_feature_prob": 0.25,
70
- "mask_time_length": 10,
71
- "mask_time_min_masks": 2,
72
- "mask_time_min_space": 1,
73
- "mask_time_other": 0.0,
74
- "mask_time_prob": 0.75,
75
- "mask_time_selection": "static",
76
- "model_type": "wav2vec2",
77
- "num_adapter_layers": 3,
78
- "num_attention_heads": 16,
79
- "num_codevector_groups": 2,
80
- "num_codevectors_per_group": 320,
81
- "num_conv_pos_embedding_groups": 16,
82
- "num_conv_pos_embeddings": 128,
83
- "num_feat_extract_layers": 7,
84
- "num_hidden_layers": 24,
85
- "num_negatives": 100,
86
- "output_hidden_size": 1024,
87
- "pad_token_id": 31,
88
- "proj_codevector_dim": 768,
89
- "tdnn_dilation": [
90
- 1,
91
- 2,
92
- 3,
93
- 1,
94
- 1
95
- ],
96
- "tdnn_dim": [
97
- 512,
98
- 512,
99
- 512,
100
- 512,
101
- 1500
102
- ],
103
- "tdnn_kernel": [
104
- 5,
105
- 3,
106
- 3,
107
- 1,
108
- 1
109
- ],
110
- "torch_dtype": "float32",
111
- "transformers_version": "4.17.0.dev0",
112
- "use_weighted_layer_sum": false,
113
- "vocab_size": 34,
114
- "xvector_output_dim": 512
115
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoint-500/optimizer.pt DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:f4def1802c330fe822f41ab8f7520445711008f9b615893c7a5bd924412fe3c2
3
- size 2490337809
 
 
 
 
checkpoint-500/preprocessor_config.json DELETED
@@ -1,9 +0,0 @@
1
- {
2
- "do_normalize": true,
3
- "feature_extractor_type": "Wav2Vec2FeatureExtractor",
4
- "feature_size": 1,
5
- "padding_side": "right",
6
- "padding_value": 0,
7
- "return_attention_mask": true,
8
- "sampling_rate": 16000
9
- }
 
 
 
 
 
 
 
 
 
 
checkpoint-500/pytorch_model.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:6f729ded3a8999cd78435e177e0a9a34b5e0461dcb32430b1135abbcf84e13f6
3
- size 1262063089
 
 
 
 
checkpoint-500/trainer_state.json DELETED
@@ -1,211 +0,0 @@
1
- {
2
- "best_metric": null,
3
- "best_model_checkpoint": null,
4
- "epoch": 0.2669157880688643,
5
- "global_step": 500,
6
- "is_hyper_param_search": false,
7
- "is_local_process_zero": true,
8
- "is_world_process_zero": true,
9
- "log_history": [
10
- {
11
- "epoch": 0.01,
12
- "learning_rate": 0.0002,
13
- "loss": 10.6012,
14
- "step": 20
15
- },
16
- {
17
- "epoch": 0.02,
18
- "learning_rate": 0.0004,
19
- "loss": 3.651,
20
- "step": 40
21
- },
22
- {
23
- "epoch": 0.03,
24
- "learning_rate": 0.00059,
25
- "loss": 3.3911,
26
- "step": 60
27
- },
28
- {
29
- "epoch": 0.04,
30
- "learning_rate": 0.0007491827839825661,
31
- "loss": 3.0481,
32
- "step": 80
33
- },
34
- {
35
- "epoch": 0.05,
36
- "learning_rate": 0.0007453010078997548,
37
- "loss": 3.4039,
38
- "step": 100
39
- },
40
- {
41
- "epoch": 0.05,
42
- "eval_loss": Infinity,
43
- "eval_runtime": 1260.1628,
44
- "eval_samples_per_second": 21.123,
45
- "eval_steps_per_second": 0.66,
46
- "eval_wer": 1.0,
47
- "step": 100
48
- },
49
- {
50
- "epoch": 0.06,
51
- "learning_rate": 0.0007412149278125851,
52
- "loss": 3.0553,
53
- "step": 120
54
- },
55
- {
56
- "epoch": 0.07,
57
- "learning_rate": 0.0007371288477254155,
58
- "loss": 3.2191,
59
- "step": 140
60
- },
61
- {
62
- "epoch": 0.09,
63
- "learning_rate": 0.0007332470716426042,
64
- "loss": 3.4094,
65
- "step": 160
66
- },
67
- {
68
- "epoch": 0.1,
69
- "learning_rate": 0.0007291609915554345,
70
- "loss": 3.0607,
71
- "step": 180
72
- },
73
- {
74
- "epoch": 0.11,
75
- "learning_rate": 0.0007252792154726233,
76
- "loss": 3.4396,
77
- "step": 200
78
- },
79
- {
80
- "epoch": 0.11,
81
- "eval_loss": Infinity,
82
- "eval_runtime": 1235.1686,
83
- "eval_samples_per_second": 21.55,
84
- "eval_steps_per_second": 0.674,
85
- "eval_wer": 1.0,
86
- "step": 200
87
- },
88
- {
89
- "epoch": 0.12,
90
- "learning_rate": 0.0007211931353854536,
91
- "loss": 3.1198,
92
- "step": 220
93
- },
94
- {
95
- "epoch": 0.13,
96
- "learning_rate": 0.0007171070552982838,
97
- "loss": 3.1801,
98
- "step": 240
99
- },
100
- {
101
- "epoch": 0.14,
102
- "learning_rate": 0.0007132252792154726,
103
- "loss": 3.3954,
104
- "step": 260
105
- },
106
- {
107
- "epoch": 0.15,
108
- "learning_rate": 0.000709139199128303,
109
- "loss": 3.0566,
110
- "step": 280
111
- },
112
- {
113
- "epoch": 0.16,
114
- "learning_rate": 0.0007052574230454917,
115
- "loss": 3.483,
116
- "step": 300
117
- },
118
- {
119
- "epoch": 0.16,
120
- "eval_loss": Infinity,
121
- "eval_runtime": 1262.2253,
122
- "eval_samples_per_second": 21.088,
123
- "eval_steps_per_second": 0.659,
124
- "eval_wer": 1.0,
125
- "step": 300
126
- },
127
- {
128
- "epoch": 0.17,
129
- "learning_rate": 0.000701171342958322,
130
- "loss": 3.1616,
131
- "step": 320
132
- },
133
- {
134
- "epoch": 0.18,
135
- "learning_rate": 0.0006970852628711522,
136
- "loss": 3.2017,
137
- "step": 340
138
- },
139
- {
140
- "epoch": 0.19,
141
- "learning_rate": 0.0006932034867883411,
142
- "loss": 3.4474,
143
- "step": 360
144
- },
145
- {
146
- "epoch": 0.2,
147
- "learning_rate": 0.0006891174067011714,
148
- "loss": 3.0656,
149
- "step": 380
150
- },
151
- {
152
- "epoch": 0.21,
153
- "learning_rate": 0.0006852356306183601,
154
- "loss": 3.5014,
155
- "step": 400
156
- },
157
- {
158
- "epoch": 0.21,
159
- "eval_loss": Infinity,
160
- "eval_runtime": 1255.7751,
161
- "eval_samples_per_second": 21.196,
162
- "eval_steps_per_second": 0.663,
163
- "eval_wer": 1.0,
164
- "step": 400
165
- },
166
- {
167
- "epoch": 0.22,
168
- "learning_rate": 0.0006811495505311904,
169
- "loss": 3.1757,
170
- "step": 420
171
- },
172
- {
173
- "epoch": 0.23,
174
- "learning_rate": 0.0006770634704440206,
175
- "loss": 3.2038,
176
- "step": 440
177
- },
178
- {
179
- "epoch": 0.25,
180
- "learning_rate": 0.0006731816943612095,
181
- "loss": 3.4072,
182
- "step": 460
183
- },
184
- {
185
- "epoch": 0.26,
186
- "learning_rate": 0.0006690956142740398,
187
- "loss": 3.0498,
188
- "step": 480
189
- },
190
- {
191
- "epoch": 0.27,
192
- "learning_rate": 0.00066500953418687,
193
- "loss": 3.331,
194
- "step": 500
195
- },
196
- {
197
- "epoch": 0.27,
198
- "eval_loss": Infinity,
199
- "eval_runtime": 1239.2922,
200
- "eval_samples_per_second": 21.478,
201
- "eval_steps_per_second": 0.671,
202
- "eval_wer": 1.0,
203
- "step": 500
204
- }
205
- ],
206
- "max_steps": 3746,
207
- "num_train_epochs": 2,
208
- "total_flos": 9.61357500213504e+18,
209
- "trial_name": null,
210
- "trial_params": null
211
- }
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
checkpoint-500/training_args.bin DELETED
@@ -1,3 +0,0 @@
1
- version https://git-lfs.github.com/spec/v1
2
- oid sha256:fcaa687be5ba49f76fb6ca2521b979e7539ac5b1f99f37017cc7c1ba8c1c2387
3
- size 3055
 
 
 
 
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6f729ded3a8999cd78435e177e0a9a34b5e0461dcb32430b1135abbcf84e13f6
3
  size 1262063089
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e95c67d08a1a0d2f34fb48babadb769336f0b21eeae723abbb0714176642fdf
3
  size 1262063089