Commit
•
0be4b52
1
Parent(s):
137a2d7
End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +853 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 200.3,
|
| 3 |
+
"eval_cer": 0.10818039039712811,
|
| 4 |
+
"eval_loss": 0.5164415836334229,
|
| 5 |
+
"eval_runtime": 120.4047,
|
| 6 |
+
"eval_samples": 2839,
|
| 7 |
+
"eval_samples_per_second": 23.579,
|
| 8 |
+
"eval_steps_per_second": 0.374,
|
| 9 |
+
"eval_wer": 0.3348876083227027,
|
| 10 |
+
"train_loss": 1.3263086397673363,
|
| 11 |
+
"train_runtime": 56983.3895,
|
| 12 |
+
"train_samples": 6281,
|
| 13 |
+
"train_samples_per_second": 22.047,
|
| 14 |
+
"train_steps_per_second": 0.172
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 200.3,
|
| 3 |
+
"eval_cer": 0.10818039039712811,
|
| 4 |
+
"eval_loss": 0.5164415836334229,
|
| 5 |
+
"eval_runtime": 120.4047,
|
| 6 |
+
"eval_samples": 2839,
|
| 7 |
+
"eval_samples_per_second": 23.579,
|
| 8 |
+
"eval_steps_per_second": 0.374,
|
| 9 |
+
"eval_wer": 0.3348876083227027
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 200.3,
|
| 3 |
+
"train_loss": 1.3263086397673363,
|
| 4 |
+
"train_runtime": 56983.3895,
|
| 5 |
+
"train_samples": 6281,
|
| 6 |
+
"train_samples_per_second": 22.047,
|
| 7 |
+
"train_steps_per_second": 0.172
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,853 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 200.3030303030303,
|
| 5 |
+
"global_step": 9815,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 2.04,
|
| 12 |
+
"learning_rate": 8.16743119266055e-06,
|
| 13 |
+
"loss": 15.5589,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 4.08,
|
| 18 |
+
"learning_rate": 1.573623853211009e-05,
|
| 19 |
+
"loss": 6.7324,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 6.12,
|
| 24 |
+
"learning_rate": 2.330504587155963e-05,
|
| 25 |
+
"loss": 4.9014,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 8.16,
|
| 30 |
+
"learning_rate": 3.087385321100917e-05,
|
| 31 |
+
"loss": 3.9471,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 8.16,
|
| 36 |
+
"eval_cer": 1.0,
|
| 37 |
+
"eval_loss": 3.7109458446502686,
|
| 38 |
+
"eval_runtime": 120.434,
|
| 39 |
+
"eval_samples_per_second": 23.573,
|
| 40 |
+
"eval_steps_per_second": 0.374,
|
| 41 |
+
"eval_wer": 1.0,
|
| 42 |
+
"step": 400
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 10.2,
|
| 46 |
+
"learning_rate": 3.8442660550458714e-05,
|
| 47 |
+
"loss": 3.5201,
|
| 48 |
+
"step": 500
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 12.24,
|
| 52 |
+
"learning_rate": 4.601146788990825e-05,
|
| 53 |
+
"loss": 3.4075,
|
| 54 |
+
"step": 600
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 14.28,
|
| 58 |
+
"learning_rate": 5.35802752293578e-05,
|
| 59 |
+
"loss": 3.3368,
|
| 60 |
+
"step": 700
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 16.32,
|
| 64 |
+
"learning_rate": 6.114908256880733e-05,
|
| 65 |
+
"loss": 3.274,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 16.32,
|
| 70 |
+
"eval_cer": 0.9572537581332735,
|
| 71 |
+
"eval_loss": 3.1581757068634033,
|
| 72 |
+
"eval_runtime": 118.5811,
|
| 73 |
+
"eval_samples_per_second": 23.941,
|
| 74 |
+
"eval_steps_per_second": 0.379,
|
| 75 |
+
"eval_wer": 0.9917300840188272,
|
| 76 |
+
"step": 800
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 18.36,
|
| 80 |
+
"learning_rate": 6.871788990825688e-05,
|
| 81 |
+
"loss": 2.8673,
|
| 82 |
+
"step": 900
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 20.4,
|
| 86 |
+
"learning_rate": 7.5e-05,
|
| 87 |
+
"loss": 2.0372,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 22.44,
|
| 92 |
+
"learning_rate": 7.5e-05,
|
| 93 |
+
"loss": 1.7262,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 24.48,
|
| 98 |
+
"learning_rate": 7.5e-05,
|
| 99 |
+
"loss": 1.5889,
|
| 100 |
+
"step": 1200
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 24.48,
|
| 104 |
+
"eval_cer": 0.19903073816468478,
|
| 105 |
+
"eval_loss": 0.7762717604637146,
|
| 106 |
+
"eval_runtime": 118.6201,
|
| 107 |
+
"eval_samples_per_second": 23.934,
|
| 108 |
+
"eval_steps_per_second": 0.379,
|
| 109 |
+
"eval_wer": 0.6030440329037082,
|
| 110 |
+
"step": 1200
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 26.53,
|
| 114 |
+
"learning_rate": 7.5e-05,
|
| 115 |
+
"loss": 1.5113,
|
| 116 |
+
"step": 1300
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 28.57,
|
| 120 |
+
"learning_rate": 7.5e-05,
|
| 121 |
+
"loss": 1.4529,
|
| 122 |
+
"step": 1400
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 30.61,
|
| 126 |
+
"learning_rate": 7.5e-05,
|
| 127 |
+
"loss": 1.4042,
|
| 128 |
+
"step": 1500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 32.65,
|
| 132 |
+
"learning_rate": 7.5e-05,
|
| 133 |
+
"loss": 1.3647,
|
| 134 |
+
"step": 1600
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 32.65,
|
| 138 |
+
"eval_cer": 0.16868745793134396,
|
| 139 |
+
"eval_loss": 0.605066180229187,
|
| 140 |
+
"eval_runtime": 119.2577,
|
| 141 |
+
"eval_samples_per_second": 23.806,
|
| 142 |
+
"eval_steps_per_second": 0.377,
|
| 143 |
+
"eval_wer": 0.5134826023841992,
|
| 144 |
+
"step": 1600
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 34.69,
|
| 148 |
+
"learning_rate": 7.5e-05,
|
| 149 |
+
"loss": 1.3249,
|
| 150 |
+
"step": 1700
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 36.73,
|
| 154 |
+
"learning_rate": 7.5e-05,
|
| 155 |
+
"loss": 1.3121,
|
| 156 |
+
"step": 1800
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 38.77,
|
| 160 |
+
"learning_rate": 7.5e-05,
|
| 161 |
+
"loss": 1.2899,
|
| 162 |
+
"step": 1900
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 40.81,
|
| 166 |
+
"learning_rate": 7.5e-05,
|
| 167 |
+
"loss": 1.2532,
|
| 168 |
+
"step": 2000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 40.81,
|
| 172 |
+
"eval_cer": 0.15389724029616333,
|
| 173 |
+
"eval_loss": 0.5423398017883301,
|
| 174 |
+
"eval_runtime": 118.7297,
|
| 175 |
+
"eval_samples_per_second": 23.911,
|
| 176 |
+
"eval_steps_per_second": 0.379,
|
| 177 |
+
"eval_wer": 0.4711652663528791,
|
| 178 |
+
"step": 2000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 42.85,
|
| 182 |
+
"learning_rate": 7.5e-05,
|
| 183 |
+
"loss": 1.2254,
|
| 184 |
+
"step": 2100
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 44.89,
|
| 188 |
+
"learning_rate": 7.5e-05,
|
| 189 |
+
"loss": 1.2028,
|
| 190 |
+
"step": 2200
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 46.93,
|
| 194 |
+
"learning_rate": 7.5e-05,
|
| 195 |
+
"loss": 1.1903,
|
| 196 |
+
"step": 2300
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 48.97,
|
| 200 |
+
"learning_rate": 7.5e-05,
|
| 201 |
+
"loss": 1.1905,
|
| 202 |
+
"step": 2400
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 48.97,
|
| 206 |
+
"eval_cer": 0.1490329818263406,
|
| 207 |
+
"eval_loss": 0.5179790258407593,
|
| 208 |
+
"eval_runtime": 119.6696,
|
| 209 |
+
"eval_samples_per_second": 23.724,
|
| 210 |
+
"eval_steps_per_second": 0.376,
|
| 211 |
+
"eval_wer": 0.4532177891171425,
|
| 212 |
+
"step": 2400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 51.02,
|
| 216 |
+
"learning_rate": 7.5e-05,
|
| 217 |
+
"loss": 1.1671,
|
| 218 |
+
"step": 2500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 53.06,
|
| 222 |
+
"learning_rate": 7.5e-05,
|
| 223 |
+
"loss": 1.1381,
|
| 224 |
+
"step": 2600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 55.1,
|
| 228 |
+
"learning_rate": 7.5e-05,
|
| 229 |
+
"loss": 1.131,
|
| 230 |
+
"step": 2700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 57.14,
|
| 234 |
+
"learning_rate": 7.5e-05,
|
| 235 |
+
"loss": 1.1193,
|
| 236 |
+
"step": 2800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 57.14,
|
| 240 |
+
"eval_cer": 0.13925959165357865,
|
| 241 |
+
"eval_loss": 0.4906373620033264,
|
| 242 |
+
"eval_runtime": 119.1486,
|
| 243 |
+
"eval_samples_per_second": 23.827,
|
| 244 |
+
"eval_steps_per_second": 0.378,
|
| 245 |
+
"eval_wer": 0.4248009501605595,
|
| 246 |
+
"step": 2800
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 59.18,
|
| 250 |
+
"learning_rate": 7.5e-05,
|
| 251 |
+
"loss": 1.1008,
|
| 252 |
+
"step": 2900
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 61.22,
|
| 256 |
+
"learning_rate": 7.5e-05,
|
| 257 |
+
"loss": 1.0784,
|
| 258 |
+
"step": 3000
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 63.26,
|
| 262 |
+
"learning_rate": 7.5e-05,
|
| 263 |
+
"loss": 1.0678,
|
| 264 |
+
"step": 3100
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 65.3,
|
| 268 |
+
"learning_rate": 7.5e-05,
|
| 269 |
+
"loss": 1.0584,
|
| 270 |
+
"step": 3200
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 65.3,
|
| 274 |
+
"eval_cer": 0.13315683194974198,
|
| 275 |
+
"eval_loss": 0.48543456196784973,
|
| 276 |
+
"eval_runtime": 119.1946,
|
| 277 |
+
"eval_samples_per_second": 23.818,
|
| 278 |
+
"eval_steps_per_second": 0.378,
|
| 279 |
+
"eval_wer": 0.40685347292482293,
|
| 280 |
+
"step": 3200
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 67.34,
|
| 284 |
+
"learning_rate": 7.5e-05,
|
| 285 |
+
"loss": 1.0434,
|
| 286 |
+
"step": 3300
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 69.38,
|
| 290 |
+
"learning_rate": 7.5e-05,
|
| 291 |
+
"loss": 1.0332,
|
| 292 |
+
"step": 3400
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 71.42,
|
| 296 |
+
"learning_rate": 7.5e-05,
|
| 297 |
+
"loss": 1.0198,
|
| 298 |
+
"step": 3500
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 73.46,
|
| 302 |
+
"learning_rate": 7.5e-05,
|
| 303 |
+
"loss": 1.0095,
|
| 304 |
+
"step": 3600
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 73.46,
|
| 308 |
+
"eval_cer": 0.12868745793134395,
|
| 309 |
+
"eval_loss": 0.4780372381210327,
|
| 310 |
+
"eval_runtime": 118.9725,
|
| 311 |
+
"eval_samples_per_second": 23.863,
|
| 312 |
+
"eval_steps_per_second": 0.378,
|
| 313 |
+
"eval_wer": 0.39255707561694453,
|
| 314 |
+
"step": 3600
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 75.51,
|
| 318 |
+
"learning_rate": 7.5e-05,
|
| 319 |
+
"loss": 0.9989,
|
| 320 |
+
"step": 3700
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 77.55,
|
| 324 |
+
"learning_rate": 7.5e-05,
|
| 325 |
+
"loss": 0.9924,
|
| 326 |
+
"step": 3800
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 79.59,
|
| 330 |
+
"learning_rate": 7.5e-05,
|
| 331 |
+
"loss": 0.9878,
|
| 332 |
+
"step": 3900
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 81.63,
|
| 336 |
+
"learning_rate": 7.5e-05,
|
| 337 |
+
"loss": 0.9759,
|
| 338 |
+
"step": 4000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 81.63,
|
| 342 |
+
"eval_cer": 0.12686560466681623,
|
| 343 |
+
"eval_loss": 0.46656644344329834,
|
| 344 |
+
"eval_runtime": 116.6705,
|
| 345 |
+
"eval_samples_per_second": 24.333,
|
| 346 |
+
"eval_steps_per_second": 0.386,
|
| 347 |
+
"eval_wer": 0.39251308670215107,
|
| 348 |
+
"step": 4000
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 83.67,
|
| 352 |
+
"learning_rate": 7.5e-05,
|
| 353 |
+
"loss": 0.9659,
|
| 354 |
+
"step": 4100
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 85.71,
|
| 358 |
+
"learning_rate": 7.5e-05,
|
| 359 |
+
"loss": 0.9531,
|
| 360 |
+
"step": 4200
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 87.75,
|
| 364 |
+
"learning_rate": 7.5e-05,
|
| 365 |
+
"loss": 0.9452,
|
| 366 |
+
"step": 4300
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 89.79,
|
| 370 |
+
"learning_rate": 7.5e-05,
|
| 371 |
+
"loss": 0.9593,
|
| 372 |
+
"step": 4400
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 89.79,
|
| 376 |
+
"eval_cer": 0.1247386134170967,
|
| 377 |
+
"eval_loss": 0.4808085262775421,
|
| 378 |
+
"eval_runtime": 119.5296,
|
| 379 |
+
"eval_samples_per_second": 23.751,
|
| 380 |
+
"eval_steps_per_second": 0.376,
|
| 381 |
+
"eval_wer": 0.38296749219196763,
|
| 382 |
+
"step": 4400
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 91.83,
|
| 386 |
+
"learning_rate": 7.5e-05,
|
| 387 |
+
"loss": 0.928,
|
| 388 |
+
"step": 4500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 93.87,
|
| 392 |
+
"learning_rate": 7.5e-05,
|
| 393 |
+
"loss": 0.9251,
|
| 394 |
+
"step": 4600
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 95.91,
|
| 398 |
+
"learning_rate": 7.5e-05,
|
| 399 |
+
"loss": 0.9098,
|
| 400 |
+
"step": 4700
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 97.95,
|
| 404 |
+
"learning_rate": 7.5e-05,
|
| 405 |
+
"loss": 0.909,
|
| 406 |
+
"step": 4800
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 97.95,
|
| 410 |
+
"eval_cer": 0.1212385012340139,
|
| 411 |
+
"eval_loss": 0.4797753393650055,
|
| 412 |
+
"eval_runtime": 116.9311,
|
| 413 |
+
"eval_samples_per_second": 24.279,
|
| 414 |
+
"eval_steps_per_second": 0.385,
|
| 415 |
+
"eval_wer": 0.37645713280253373,
|
| 416 |
+
"step": 4800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 99.99,
|
| 420 |
+
"learning_rate": 7.5e-05,
|
| 421 |
+
"loss": 0.8979,
|
| 422 |
+
"step": 4900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 102.04,
|
| 426 |
+
"learning_rate": 7.370797677261614e-05,
|
| 427 |
+
"loss": 0.8993,
|
| 428 |
+
"step": 5000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 104.08,
|
| 432 |
+
"learning_rate": 7.22562652811736e-05,
|
| 433 |
+
"loss": 0.8888,
|
| 434 |
+
"step": 5100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 106.12,
|
| 438 |
+
"learning_rate": 7.080455378973104e-05,
|
| 439 |
+
"loss": 0.8788,
|
| 440 |
+
"step": 5200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 106.12,
|
| 444 |
+
"eval_cer": 0.11615885124523222,
|
| 445 |
+
"eval_loss": 0.4906252324581146,
|
| 446 |
+
"eval_runtime": 119.2237,
|
| 447 |
+
"eval_samples_per_second": 23.812,
|
| 448 |
+
"eval_steps_per_second": 0.377,
|
| 449 |
+
"eval_wer": 0.3608410680508512,
|
| 450 |
+
"step": 5200
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 108.16,
|
| 454 |
+
"learning_rate": 6.93528422982885e-05,
|
| 455 |
+
"loss": 0.8701,
|
| 456 |
+
"step": 5300
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 110.2,
|
| 460 |
+
"learning_rate": 6.790113080684596e-05,
|
| 461 |
+
"loss": 0.8621,
|
| 462 |
+
"step": 5400
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 112.24,
|
| 466 |
+
"learning_rate": 6.644941931540342e-05,
|
| 467 |
+
"loss": 0.8525,
|
| 468 |
+
"step": 5500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 114.28,
|
| 472 |
+
"learning_rate": 6.499770782396088e-05,
|
| 473 |
+
"loss": 0.8471,
|
| 474 |
+
"step": 5600
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 114.28,
|
| 478 |
+
"eval_cer": 0.11655373569665695,
|
| 479 |
+
"eval_loss": 0.4758923053741455,
|
| 480 |
+
"eval_runtime": 117.4104,
|
| 481 |
+
"eval_samples_per_second": 24.18,
|
| 482 |
+
"eval_steps_per_second": 0.383,
|
| 483 |
+
"eval_wer": 0.360357189988123,
|
| 484 |
+
"step": 5600
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 116.32,
|
| 488 |
+
"learning_rate": 6.354599633251834e-05,
|
| 489 |
+
"loss": 0.8442,
|
| 490 |
+
"step": 5700
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 118.36,
|
| 494 |
+
"learning_rate": 6.209428484107578e-05,
|
| 495 |
+
"loss": 0.8244,
|
| 496 |
+
"step": 5800
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 120.4,
|
| 500 |
+
"learning_rate": 6.0642573349633246e-05,
|
| 501 |
+
"loss": 0.8224,
|
| 502 |
+
"step": 5900
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 122.44,
|
| 506 |
+
"learning_rate": 5.9190861858190705e-05,
|
| 507 |
+
"loss": 0.8116,
|
| 508 |
+
"step": 6000
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 122.44,
|
| 512 |
+
"eval_cer": 0.11757684541171191,
|
| 513 |
+
"eval_loss": 0.5079970359802246,
|
| 514 |
+
"eval_runtime": 118.6326,
|
| 515 |
+
"eval_samples_per_second": 23.931,
|
| 516 |
+
"eval_steps_per_second": 0.379,
|
| 517 |
+
"eval_wer": 0.3627325913869705,
|
| 518 |
+
"step": 6000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 124.48,
|
| 522 |
+
"learning_rate": 5.773915036674816e-05,
|
| 523 |
+
"loss": 0.807,
|
| 524 |
+
"step": 6100
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 126.53,
|
| 528 |
+
"learning_rate": 5.6287438875305616e-05,
|
| 529 |
+
"loss": 0.8011,
|
| 530 |
+
"step": 6200
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 128.57,
|
| 534 |
+
"learning_rate": 5.4835727383863075e-05,
|
| 535 |
+
"loss": 0.799,
|
| 536 |
+
"step": 6300
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 130.61,
|
| 540 |
+
"learning_rate": 5.3384015892420534e-05,
|
| 541 |
+
"loss": 0.7881,
|
| 542 |
+
"step": 6400
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 130.61,
|
| 546 |
+
"eval_cer": 0.11347543190486875,
|
| 547 |
+
"eval_loss": 0.4867645502090454,
|
| 548 |
+
"eval_runtime": 118.7534,
|
| 549 |
+
"eval_samples_per_second": 23.907,
|
| 550 |
+
"eval_steps_per_second": 0.379,
|
| 551 |
+
"eval_wer": 0.34892007214182025,
|
| 552 |
+
"step": 6400
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 132.65,
|
| 556 |
+
"learning_rate": 5.193230440097799e-05,
|
| 557 |
+
"loss": 0.7781,
|
| 558 |
+
"step": 6500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 134.69,
|
| 562 |
+
"learning_rate": 5.048059290953545e-05,
|
| 563 |
+
"loss": 0.7685,
|
| 564 |
+
"step": 6600
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 136.73,
|
| 568 |
+
"learning_rate": 4.902888141809291e-05,
|
| 569 |
+
"loss": 0.7654,
|
| 570 |
+
"step": 6700
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 138.77,
|
| 574 |
+
"learning_rate": 4.757716992665037e-05,
|
| 575 |
+
"loss": 0.766,
|
| 576 |
+
"step": 6800
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 138.77,
|
| 580 |
+
"eval_cer": 0.11357415301772493,
|
| 581 |
+
"eval_loss": 0.49552106857299805,
|
| 582 |
+
"eval_runtime": 117.4435,
|
| 583 |
+
"eval_samples_per_second": 24.173,
|
| 584 |
+
"eval_steps_per_second": 0.383,
|
| 585 |
+
"eval_wer": 0.34922799454537456,
|
| 586 |
+
"step": 6800
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 140.81,
|
| 590 |
+
"learning_rate": 4.6139975550122245e-05,
|
| 591 |
+
"loss": 0.752,
|
| 592 |
+
"step": 6900
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 142.85,
|
| 596 |
+
"learning_rate": 4.4688264058679704e-05,
|
| 597 |
+
"loss": 0.7451,
|
| 598 |
+
"step": 7000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 144.89,
|
| 602 |
+
"learning_rate": 4.323655256723716e-05,
|
| 603 |
+
"loss": 0.7398,
|
| 604 |
+
"step": 7100
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 146.93,
|
| 608 |
+
"learning_rate": 4.178484107579462e-05,
|
| 609 |
+
"loss": 0.7333,
|
| 610 |
+
"step": 7200
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 146.93,
|
| 614 |
+
"eval_cer": 0.11247027148306035,
|
| 615 |
+
"eval_loss": 0.5019353032112122,
|
| 616 |
+
"eval_runtime": 118.907,
|
| 617 |
+
"eval_samples_per_second": 23.876,
|
| 618 |
+
"eval_steps_per_second": 0.378,
|
| 619 |
+
"eval_wer": 0.3461487705098315,
|
| 620 |
+
"step": 7200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 148.97,
|
| 624 |
+
"learning_rate": 4.033312958435208e-05,
|
| 625 |
+
"loss": 0.7255,
|
| 626 |
+
"step": 7300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 151.02,
|
| 630 |
+
"learning_rate": 3.888141809290954e-05,
|
| 631 |
+
"loss": 0.7283,
|
| 632 |
+
"step": 7400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 153.06,
|
| 636 |
+
"learning_rate": 3.742970660146699e-05,
|
| 637 |
+
"loss": 0.7149,
|
| 638 |
+
"step": 7500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 155.1,
|
| 642 |
+
"learning_rate": 3.597799511002445e-05,
|
| 643 |
+
"loss": 0.709,
|
| 644 |
+
"step": 7600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 155.1,
|
| 648 |
+
"eval_cer": 0.11168050258021091,
|
| 649 |
+
"eval_loss": 0.5083914995193481,
|
| 650 |
+
"eval_runtime": 120.383,
|
| 651 |
+
"eval_samples_per_second": 23.583,
|
| 652 |
+
"eval_steps_per_second": 0.374,
|
| 653 |
+
"eval_wer": 0.3468086042317336,
|
| 654 |
+
"step": 7600
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 157.14,
|
| 658 |
+
"learning_rate": 3.452628361858191e-05,
|
| 659 |
+
"loss": 0.7084,
|
| 660 |
+
"step": 7700
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 159.18,
|
| 664 |
+
"learning_rate": 3.307457212713937e-05,
|
| 665 |
+
"loss": 0.704,
|
| 666 |
+
"step": 7800
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 161.22,
|
| 670 |
+
"learning_rate": 3.1622860635696827e-05,
|
| 671 |
+
"loss": 0.698,
|
| 672 |
+
"step": 7900
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 163.26,
|
| 676 |
+
"learning_rate": 3.0171149144254275e-05,
|
| 677 |
+
"loss": 0.6911,
|
| 678 |
+
"step": 8000
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 163.26,
|
| 682 |
+
"eval_cer": 0.11064841821853265,
|
| 683 |
+
"eval_loss": 0.5144167542457581,
|
| 684 |
+
"eval_runtime": 120.6954,
|
| 685 |
+
"eval_samples_per_second": 23.522,
|
| 686 |
+
"eval_steps_per_second": 0.373,
|
| 687 |
+
"eval_wer": 0.34122201205296265,
|
| 688 |
+
"step": 8000
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 165.3,
|
| 692 |
+
"learning_rate": 2.8719437652811734e-05,
|
| 693 |
+
"loss": 0.6889,
|
| 694 |
+
"step": 8100
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 167.34,
|
| 698 |
+
"learning_rate": 2.726772616136919e-05,
|
| 699 |
+
"loss": 0.6863,
|
| 700 |
+
"step": 8200
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 169.38,
|
| 704 |
+
"learning_rate": 2.581601466992665e-05,
|
| 705 |
+
"loss": 0.6825,
|
| 706 |
+
"step": 8300
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 171.42,
|
| 710 |
+
"learning_rate": 2.4364303178484107e-05,
|
| 711 |
+
"loss": 0.6683,
|
| 712 |
+
"step": 8400
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 171.42,
|
| 716 |
+
"eval_cer": 0.11165357864034103,
|
| 717 |
+
"eval_loss": 0.5218909382820129,
|
| 718 |
+
"eval_runtime": 119.5764,
|
| 719 |
+
"eval_samples_per_second": 23.742,
|
| 720 |
+
"eval_steps_per_second": 0.376,
|
| 721 |
+
"eval_wer": 0.34091408964940834,
|
| 722 |
+
"step": 8400
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 173.46,
|
| 726 |
+
"learning_rate": 2.2912591687041566e-05,
|
| 727 |
+
"loss": 0.6627,
|
| 728 |
+
"step": 8500
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 175.51,
|
| 732 |
+
"learning_rate": 2.1460880195599022e-05,
|
| 733 |
+
"loss": 0.6655,
|
| 734 |
+
"step": 8600
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 177.55,
|
| 738 |
+
"learning_rate": 2.000916870415648e-05,
|
| 739 |
+
"loss": 0.6662,
|
| 740 |
+
"step": 8700
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 179.59,
|
| 744 |
+
"learning_rate": 1.855745721271394e-05,
|
| 745 |
+
"loss": 0.659,
|
| 746 |
+
"step": 8800
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 179.59,
|
| 750 |
+
"eval_cer": 0.1095983845636078,
|
| 751 |
+
"eval_loss": 0.5229596495628357,
|
| 752 |
+
"eval_runtime": 119.167,
|
| 753 |
+
"eval_samples_per_second": 23.824,
|
| 754 |
+
"eval_steps_per_second": 0.378,
|
| 755 |
+
"eval_wer": 0.3375709321251045,
|
| 756 |
+
"step": 8800
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 181.63,
|
| 760 |
+
"learning_rate": 1.7105745721271395e-05,
|
| 761 |
+
"loss": 0.6526,
|
| 762 |
+
"step": 8900
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 183.67,
|
| 766 |
+
"learning_rate": 1.5654034229828854e-05,
|
| 767 |
+
"loss": 0.654,
|
| 768 |
+
"step": 9000
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 185.71,
|
| 772 |
+
"learning_rate": 1.4202322738386313e-05,
|
| 773 |
+
"loss": 0.6481,
|
| 774 |
+
"step": 9100
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 187.75,
|
| 778 |
+
"learning_rate": 1.275061124694377e-05,
|
| 779 |
+
"loss": 0.6475,
|
| 780 |
+
"step": 9200
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 187.75,
|
| 784 |
+
"eval_cer": 0.10967915638321742,
|
| 785 |
+
"eval_loss": 0.5228886604309082,
|
| 786 |
+
"eval_runtime": 118.4312,
|
| 787 |
+
"eval_samples_per_second": 23.972,
|
| 788 |
+
"eval_steps_per_second": 0.38,
|
| 789 |
+
"eval_wer": 0.3397703778647781,
|
| 790 |
+
"step": 9200
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 189.79,
|
| 794 |
+
"learning_rate": 1.1298899755501229e-05,
|
| 795 |
+
"loss": 0.6476,
|
| 796 |
+
"step": 9300
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 191.83,
|
| 800 |
+
"learning_rate": 9.847188264058686e-06,
|
| 801 |
+
"loss": 0.6485,
|
| 802 |
+
"step": 9400
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 193.87,
|
| 806 |
+
"learning_rate": 8.395476772616145e-06,
|
| 807 |
+
"loss": 0.6449,
|
| 808 |
+
"step": 9500
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 195.91,
|
| 812 |
+
"learning_rate": 6.943765281173602e-06,
|
| 813 |
+
"loss": 0.6419,
|
| 814 |
+
"step": 9600
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 195.91,
|
| 818 |
+
"eval_cer": 0.10843168050258022,
|
| 819 |
+
"eval_loss": 0.520015299320221,
|
| 820 |
+
"eval_runtime": 118.8519,
|
| 821 |
+
"eval_samples_per_second": 23.887,
|
| 822 |
+
"eval_steps_per_second": 0.379,
|
| 823 |
+
"eval_wer": 0.33369990762327895,
|
| 824 |
+
"step": 9600
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 197.95,
|
| 828 |
+
"learning_rate": 5.4920537897310516e-06,
|
| 829 |
+
"loss": 0.627,
|
| 830 |
+
"step": 9700
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 199.99,
|
| 834 |
+
"learning_rate": 4.054859413202941e-06,
|
| 835 |
+
"loss": 0.6301,
|
| 836 |
+
"step": 9800
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 200.3,
|
| 840 |
+
"step": 9815,
|
| 841 |
+
"total_flos": 1.72528931544575e+20,
|
| 842 |
+
"train_loss": 1.3263086397673363,
|
| 843 |
+
"train_runtime": 56983.3895,
|
| 844 |
+
"train_samples_per_second": 22.047,
|
| 845 |
+
"train_steps_per_second": 0.172
|
| 846 |
+
}
|
| 847 |
+
],
|
| 848 |
+
"max_steps": 9815,
|
| 849 |
+
"num_train_epochs": 201,
|
| 850 |
+
"total_flos": 1.72528931544575e+20,
|
| 851 |
+
"trial_name": null,
|
| 852 |
+
"trial_params": null
|
| 853 |
+
}
|