End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.98,
|
| 3 |
+
"eval_cer": 5.352801251133512,
|
| 4 |
+
"eval_loss": 0.2793508470058441,
|
| 5 |
+
"eval_runtime": 4201.4681,
|
| 6 |
+
"eval_samples_per_second": 1.096,
|
| 7 |
+
"eval_steps_per_second": 0.069,
|
| 8 |
+
"eval_wer": 8.60641891891892,
|
| 9 |
+
"train_loss": 0.043234722255170346,
|
| 10 |
+
"train_runtime": 85032.8136,
|
| 11 |
+
"train_samples_per_second": 3.763,
|
| 12 |
+
"train_steps_per_second": 0.118
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.98,
|
| 3 |
+
"eval_cer": 5.352801251133512,
|
| 4 |
+
"eval_loss": 0.2793508470058441,
|
| 5 |
+
"eval_runtime": 4201.4681,
|
| 6 |
+
"eval_samples_per_second": 1.096,
|
| 7 |
+
"eval_steps_per_second": 0.069,
|
| 8 |
+
"eval_wer": 8.60641891891892
|
| 9 |
+
}
|
runs/Dec15_18-20-49_129-146-162-127/events.out.tfevents.1671220686.129-146-162-127.75243.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb17a42136204a748e5afd95621cee525d2afa303ee63ef6b93f4dab97d7dce0
|
| 3 |
+
size 405
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.98,
|
| 3 |
+
"train_loss": 0.043234722255170346,
|
| 4 |
+
"train_runtime": 85032.8136,
|
| 5 |
+
"train_samples_per_second": 3.763,
|
| 6 |
+
"train_steps_per_second": 0.118
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,725 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 8.60641891891892,
|
| 3 |
+
"best_model_checkpoint": "./checkpoint-9000",
|
| 4 |
+
"epoch": 10.97694840834248,
|
| 5 |
+
"global_step": 10000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.11,
|
| 12 |
+
"learning_rate": 1.94e-06,
|
| 13 |
+
"loss": 0.8124,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.22,
|
| 18 |
+
"learning_rate": 3.94e-06,
|
| 19 |
+
"loss": 0.3235,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.33,
|
| 24 |
+
"learning_rate": 5.94e-06,
|
| 25 |
+
"loss": 0.2396,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.44,
|
| 30 |
+
"learning_rate": 7.94e-06,
|
| 31 |
+
"loss": 0.2138,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.55,
|
| 36 |
+
"learning_rate": 9.940000000000001e-06,
|
| 37 |
+
"loss": 0.1972,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.66,
|
| 42 |
+
"learning_rate": 9.897894736842107e-06,
|
| 43 |
+
"loss": 0.1885,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.77,
|
| 48 |
+
"learning_rate": 9.79263157894737e-06,
|
| 49 |
+
"loss": 0.1792,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.88,
|
| 54 |
+
"learning_rate": 9.687368421052632e-06,
|
| 55 |
+
"loss": 0.174,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.99,
|
| 60 |
+
"learning_rate": 9.582105263157897e-06,
|
| 61 |
+
"loss": 0.1669,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 1.1,
|
| 66 |
+
"learning_rate": 9.476842105263158e-06,
|
| 67 |
+
"loss": 0.1106,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.1,
|
| 72 |
+
"eval_cer": 6.4784271464430745,
|
| 73 |
+
"eval_loss": 0.18272116780281067,
|
| 74 |
+
"eval_runtime": 4332.708,
|
| 75 |
+
"eval_samples_per_second": 1.063,
|
| 76 |
+
"eval_steps_per_second": 0.066,
|
| 77 |
+
"eval_wer": 10.347972972972972,
|
| 78 |
+
"step": 1000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 1.21,
|
| 82 |
+
"learning_rate": 9.371578947368421e-06,
|
| 83 |
+
"loss": 0.1083,
|
| 84 |
+
"step": 1100
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 1.32,
|
| 88 |
+
"learning_rate": 9.266315789473685e-06,
|
| 89 |
+
"loss": 0.1043,
|
| 90 |
+
"step": 1200
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 1.43,
|
| 94 |
+
"learning_rate": 9.161052631578948e-06,
|
| 95 |
+
"loss": 0.1016,
|
| 96 |
+
"step": 1300
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.54,
|
| 100 |
+
"learning_rate": 9.055789473684211e-06,
|
| 101 |
+
"loss": 0.1023,
|
| 102 |
+
"step": 1400
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.65,
|
| 106 |
+
"learning_rate": 8.950526315789474e-06,
|
| 107 |
+
"loss": 0.101,
|
| 108 |
+
"step": 1500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.76,
|
| 112 |
+
"learning_rate": 8.845263157894738e-06,
|
| 113 |
+
"loss": 0.099,
|
| 114 |
+
"step": 1600
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.87,
|
| 118 |
+
"learning_rate": 8.740000000000001e-06,
|
| 119 |
+
"loss": 0.1005,
|
| 120 |
+
"step": 1700
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.98,
|
| 124 |
+
"learning_rate": 8.634736842105264e-06,
|
| 125 |
+
"loss": 0.1002,
|
| 126 |
+
"step": 1800
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 2.09,
|
| 130 |
+
"learning_rate": 8.529473684210528e-06,
|
| 131 |
+
"loss": 0.0601,
|
| 132 |
+
"step": 1900
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.2,
|
| 136 |
+
"learning_rate": 8.42421052631579e-06,
|
| 137 |
+
"loss": 0.0487,
|
| 138 |
+
"step": 2000
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 2.2,
|
| 142 |
+
"eval_cer": 5.912657213073819,
|
| 143 |
+
"eval_loss": 0.17993855476379395,
|
| 144 |
+
"eval_runtime": 4312.9082,
|
| 145 |
+
"eval_samples_per_second": 1.067,
|
| 146 |
+
"eval_steps_per_second": 0.067,
|
| 147 |
+
"eval_wer": 9.47635135135135,
|
| 148 |
+
"step": 2000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 2.31,
|
| 152 |
+
"learning_rate": 8.318947368421052e-06,
|
| 153 |
+
"loss": 0.0508,
|
| 154 |
+
"step": 2100
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 2.41,
|
| 158 |
+
"learning_rate": 8.213684210526316e-06,
|
| 159 |
+
"loss": 0.0501,
|
| 160 |
+
"step": 2200
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.52,
|
| 164 |
+
"learning_rate": 8.10842105263158e-06,
|
| 165 |
+
"loss": 0.0506,
|
| 166 |
+
"step": 2300
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 2.63,
|
| 170 |
+
"learning_rate": 8.003157894736842e-06,
|
| 171 |
+
"loss": 0.0498,
|
| 172 |
+
"step": 2400
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 2.74,
|
| 176 |
+
"learning_rate": 7.897894736842106e-06,
|
| 177 |
+
"loss": 0.0486,
|
| 178 |
+
"step": 2500
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 2.85,
|
| 182 |
+
"learning_rate": 7.792631578947369e-06,
|
| 183 |
+
"loss": 0.0513,
|
| 184 |
+
"step": 2600
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.96,
|
| 188 |
+
"learning_rate": 7.687368421052632e-06,
|
| 189 |
+
"loss": 0.0504,
|
| 190 |
+
"step": 2700
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.07,
|
| 194 |
+
"learning_rate": 7.582105263157895e-06,
|
| 195 |
+
"loss": 0.0307,
|
| 196 |
+
"step": 2800
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 3.18,
|
| 200 |
+
"learning_rate": 7.476842105263159e-06,
|
| 201 |
+
"loss": 0.0239,
|
| 202 |
+
"step": 2900
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 3.29,
|
| 206 |
+
"learning_rate": 7.371578947368422e-06,
|
| 207 |
+
"loss": 0.0243,
|
| 208 |
+
"step": 3000
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 3.29,
|
| 212 |
+
"eval_cer": 5.806862835289325,
|
| 213 |
+
"eval_loss": 0.19502821564674377,
|
| 214 |
+
"eval_runtime": 4299.7561,
|
| 215 |
+
"eval_samples_per_second": 1.071,
|
| 216 |
+
"eval_steps_per_second": 0.067,
|
| 217 |
+
"eval_wer": 9.21114864864865,
|
| 218 |
+
"step": 3000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 3.4,
|
| 222 |
+
"learning_rate": 7.266315789473684e-06,
|
| 223 |
+
"loss": 0.024,
|
| 224 |
+
"step": 3100
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 3.51,
|
| 228 |
+
"learning_rate": 7.161052631578948e-06,
|
| 229 |
+
"loss": 0.0252,
|
| 230 |
+
"step": 3200
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 3.62,
|
| 234 |
+
"learning_rate": 7.055789473684212e-06,
|
| 235 |
+
"loss": 0.0244,
|
| 236 |
+
"step": 3300
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 3.73,
|
| 240 |
+
"learning_rate": 6.950526315789474e-06,
|
| 241 |
+
"loss": 0.0249,
|
| 242 |
+
"step": 3400
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 3.84,
|
| 246 |
+
"learning_rate": 6.845263157894737e-06,
|
| 247 |
+
"loss": 0.0247,
|
| 248 |
+
"step": 3500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 3.95,
|
| 252 |
+
"learning_rate": 6.740000000000001e-06,
|
| 253 |
+
"loss": 0.0245,
|
| 254 |
+
"step": 3600
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 4.06,
|
| 258 |
+
"learning_rate": 6.634736842105264e-06,
|
| 259 |
+
"loss": 0.0173,
|
| 260 |
+
"step": 3700
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 4.17,
|
| 264 |
+
"learning_rate": 6.529473684210526e-06,
|
| 265 |
+
"loss": 0.0106,
|
| 266 |
+
"step": 3800
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 4.28,
|
| 270 |
+
"learning_rate": 6.42421052631579e-06,
|
| 271 |
+
"loss": 0.0116,
|
| 272 |
+
"step": 3900
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 4.39,
|
| 276 |
+
"learning_rate": 6.318947368421054e-06,
|
| 277 |
+
"loss": 0.0106,
|
| 278 |
+
"step": 4000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 4.39,
|
| 282 |
+
"eval_cer": 5.575560841623846,
|
| 283 |
+
"eval_loss": 0.21126572787761688,
|
| 284 |
+
"eval_runtime": 4218.4134,
|
| 285 |
+
"eval_samples_per_second": 1.091,
|
| 286 |
+
"eval_steps_per_second": 0.068,
|
| 287 |
+
"eval_wer": 8.971283783783784,
|
| 288 |
+
"step": 4000
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 4.5,
|
| 292 |
+
"learning_rate": 6.213684210526316e-06,
|
| 293 |
+
"loss": 0.0115,
|
| 294 |
+
"step": 4100
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 4.61,
|
| 298 |
+
"learning_rate": 6.108421052631579e-06,
|
| 299 |
+
"loss": 0.0105,
|
| 300 |
+
"step": 4200
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 4.72,
|
| 304 |
+
"learning_rate": 6.003157894736843e-06,
|
| 305 |
+
"loss": 0.0104,
|
| 306 |
+
"step": 4300
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 4.83,
|
| 310 |
+
"learning_rate": 5.897894736842106e-06,
|
| 311 |
+
"loss": 0.0123,
|
| 312 |
+
"step": 4400
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 4.94,
|
| 316 |
+
"learning_rate": 5.7926315789473685e-06,
|
| 317 |
+
"loss": 0.0116,
|
| 318 |
+
"step": 4500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 5.05,
|
| 322 |
+
"learning_rate": 5.687368421052633e-06,
|
| 323 |
+
"loss": 0.0079,
|
| 324 |
+
"step": 4600
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 5.16,
|
| 328 |
+
"learning_rate": 5.582105263157895e-06,
|
| 329 |
+
"loss": 0.0053,
|
| 330 |
+
"step": 4700
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 5.27,
|
| 334 |
+
"learning_rate": 5.476842105263158e-06,
|
| 335 |
+
"loss": 0.0052,
|
| 336 |
+
"step": 4800
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 5.38,
|
| 340 |
+
"learning_rate": 5.371578947368421e-06,
|
| 341 |
+
"loss": 0.0053,
|
| 342 |
+
"step": 4900
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 5.49,
|
| 346 |
+
"learning_rate": 5.266315789473685e-06,
|
| 347 |
+
"loss": 0.0054,
|
| 348 |
+
"step": 5000
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 5.49,
|
| 352 |
+
"eval_cer": 5.404055670184384,
|
| 353 |
+
"eval_loss": 0.23248465359210968,
|
| 354 |
+
"eval_runtime": 4237.1053,
|
| 355 |
+
"eval_samples_per_second": 1.087,
|
| 356 |
+
"eval_steps_per_second": 0.068,
|
| 357 |
+
"eval_wer": 8.64695945945946,
|
| 358 |
+
"step": 5000
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 5.6,
|
| 362 |
+
"learning_rate": 5.161052631578948e-06,
|
| 363 |
+
"loss": 0.0053,
|
| 364 |
+
"step": 5100
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 5.71,
|
| 368 |
+
"learning_rate": 5.0557894736842105e-06,
|
| 369 |
+
"loss": 0.0056,
|
| 370 |
+
"step": 5200
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 5.82,
|
| 374 |
+
"learning_rate": 4.950526315789474e-06,
|
| 375 |
+
"loss": 0.005,
|
| 376 |
+
"step": 5300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 5.93,
|
| 380 |
+
"learning_rate": 4.845263157894737e-06,
|
| 381 |
+
"loss": 0.0054,
|
| 382 |
+
"step": 5400
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 6.04,
|
| 386 |
+
"learning_rate": 4.74e-06,
|
| 387 |
+
"loss": 0.0045,
|
| 388 |
+
"step": 5500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 6.15,
|
| 392 |
+
"learning_rate": 4.634736842105264e-06,
|
| 393 |
+
"loss": 0.0029,
|
| 394 |
+
"step": 5600
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 6.26,
|
| 398 |
+
"learning_rate": 4.529473684210527e-06,
|
| 399 |
+
"loss": 0.0024,
|
| 400 |
+
"step": 5700
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 6.37,
|
| 404 |
+
"learning_rate": 4.424210526315789e-06,
|
| 405 |
+
"loss": 0.0025,
|
| 406 |
+
"step": 5800
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 6.48,
|
| 410 |
+
"learning_rate": 4.3189473684210535e-06,
|
| 411 |
+
"loss": 0.0026,
|
| 412 |
+
"step": 5900
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 6.59,
|
| 416 |
+
"learning_rate": 4.213684210526316e-06,
|
| 417 |
+
"loss": 0.0031,
|
| 418 |
+
"step": 6000
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 6.59,
|
| 422 |
+
"eval_cer": 5.440853714631165,
|
| 423 |
+
"eval_loss": 0.24615277349948883,
|
| 424 |
+
"eval_runtime": 4218.0606,
|
| 425 |
+
"eval_samples_per_second": 1.091,
|
| 426 |
+
"eval_steps_per_second": 0.068,
|
| 427 |
+
"eval_wer": 8.70777027027027,
|
| 428 |
+
"step": 6000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 6.7,
|
| 432 |
+
"learning_rate": 4.108421052631579e-06,
|
| 433 |
+
"loss": 0.0028,
|
| 434 |
+
"step": 6100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 6.81,
|
| 438 |
+
"learning_rate": 4.0031578947368424e-06,
|
| 439 |
+
"loss": 0.0026,
|
| 440 |
+
"step": 6200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 6.92,
|
| 444 |
+
"learning_rate": 3.897894736842106e-06,
|
| 445 |
+
"loss": 0.0027,
|
| 446 |
+
"step": 6300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 7.03,
|
| 450 |
+
"learning_rate": 3.792631578947369e-06,
|
| 451 |
+
"loss": 0.0021,
|
| 452 |
+
"step": 6400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 7.14,
|
| 456 |
+
"learning_rate": 3.687368421052632e-06,
|
| 457 |
+
"loss": 0.0014,
|
| 458 |
+
"step": 6500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 7.24,
|
| 462 |
+
"learning_rate": 3.582105263157895e-06,
|
| 463 |
+
"loss": 0.0015,
|
| 464 |
+
"step": 6600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 7.35,
|
| 468 |
+
"learning_rate": 3.476842105263158e-06,
|
| 469 |
+
"loss": 0.001,
|
| 470 |
+
"step": 6700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 7.46,
|
| 474 |
+
"learning_rate": 3.3715789473684212e-06,
|
| 475 |
+
"loss": 0.0015,
|
| 476 |
+
"step": 6800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 7.57,
|
| 480 |
+
"learning_rate": 3.2663157894736845e-06,
|
| 481 |
+
"loss": 0.0013,
|
| 482 |
+
"step": 6900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 7.68,
|
| 486 |
+
"learning_rate": 3.1610526315789474e-06,
|
| 487 |
+
"loss": 0.0014,
|
| 488 |
+
"step": 7000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 7.68,
|
| 492 |
+
"eval_cer": 5.484879946379992,
|
| 493 |
+
"eval_loss": 0.26076802611351013,
|
| 494 |
+
"eval_runtime": 4205.2214,
|
| 495 |
+
"eval_samples_per_second": 1.095,
|
| 496 |
+
"eval_steps_per_second": 0.068,
|
| 497 |
+
"eval_wer": 8.714527027027028,
|
| 498 |
+
"step": 7000
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 7.79,
|
| 502 |
+
"learning_rate": 3.055789473684211e-06,
|
| 503 |
+
"loss": 0.0013,
|
| 504 |
+
"step": 7100
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 7.9,
|
| 508 |
+
"learning_rate": 2.9505263157894735e-06,
|
| 509 |
+
"loss": 0.0013,
|
| 510 |
+
"step": 7200
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 8.01,
|
| 514 |
+
"learning_rate": 2.845263157894737e-06,
|
| 515 |
+
"loss": 0.0018,
|
| 516 |
+
"step": 7300
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 8.12,
|
| 520 |
+
"learning_rate": 2.7400000000000004e-06,
|
| 521 |
+
"loss": 0.0011,
|
| 522 |
+
"step": 7400
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 8.23,
|
| 526 |
+
"learning_rate": 2.6347368421052633e-06,
|
| 527 |
+
"loss": 0.0009,
|
| 528 |
+
"step": 7500
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 8.34,
|
| 532 |
+
"learning_rate": 2.5294736842105266e-06,
|
| 533 |
+
"loss": 0.0008,
|
| 534 |
+
"step": 7600
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 8.45,
|
| 538 |
+
"learning_rate": 2.42421052631579e-06,
|
| 539 |
+
"loss": 0.0008,
|
| 540 |
+
"step": 7700
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 8.56,
|
| 544 |
+
"learning_rate": 2.3189473684210527e-06,
|
| 545 |
+
"loss": 0.0008,
|
| 546 |
+
"step": 7800
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 8.67,
|
| 550 |
+
"learning_rate": 2.213684210526316e-06,
|
| 551 |
+
"loss": 0.0008,
|
| 552 |
+
"step": 7900
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 8.78,
|
| 556 |
+
"learning_rate": 2.1084210526315792e-06,
|
| 557 |
+
"loss": 0.0009,
|
| 558 |
+
"step": 8000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 8.78,
|
| 562 |
+
"eval_cer": 5.387627971770643,
|
| 563 |
+
"eval_loss": 0.269546777009964,
|
| 564 |
+
"eval_runtime": 4205.0269,
|
| 565 |
+
"eval_samples_per_second": 1.095,
|
| 566 |
+
"eval_steps_per_second": 0.068,
|
| 567 |
+
"eval_wer": 8.630067567567568,
|
| 568 |
+
"step": 8000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 8.89,
|
| 572 |
+
"learning_rate": 2.003157894736842e-06,
|
| 573 |
+
"loss": 0.0007,
|
| 574 |
+
"step": 8100
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 9.0,
|
| 578 |
+
"learning_rate": 1.8978947368421056e-06,
|
| 579 |
+
"loss": 0.0009,
|
| 580 |
+
"step": 8200
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 9.11,
|
| 584 |
+
"learning_rate": 1.7926315789473686e-06,
|
| 585 |
+
"loss": 0.0007,
|
| 586 |
+
"step": 8300
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 9.22,
|
| 590 |
+
"learning_rate": 1.6873684210526317e-06,
|
| 591 |
+
"loss": 0.0006,
|
| 592 |
+
"step": 8400
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 9.33,
|
| 596 |
+
"learning_rate": 1.5821052631578948e-06,
|
| 597 |
+
"loss": 0.0005,
|
| 598 |
+
"step": 8500
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 9.44,
|
| 602 |
+
"learning_rate": 1.4768421052631578e-06,
|
| 603 |
+
"loss": 0.0005,
|
| 604 |
+
"step": 8600
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 9.55,
|
| 608 |
+
"learning_rate": 1.3715789473684213e-06,
|
| 609 |
+
"loss": 0.0006,
|
| 610 |
+
"step": 8700
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 9.66,
|
| 614 |
+
"learning_rate": 1.2663157894736844e-06,
|
| 615 |
+
"loss": 0.0005,
|
| 616 |
+
"step": 8800
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 9.77,
|
| 620 |
+
"learning_rate": 1.1610526315789474e-06,
|
| 621 |
+
"loss": 0.0005,
|
| 622 |
+
"step": 8900
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 9.88,
|
| 626 |
+
"learning_rate": 1.0557894736842105e-06,
|
| 627 |
+
"loss": 0.0004,
|
| 628 |
+
"step": 9000
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 9.88,
|
| 632 |
+
"eval_cer": 5.352801251133512,
|
| 633 |
+
"eval_loss": 0.2793508470058441,
|
| 634 |
+
"eval_runtime": 4203.0281,
|
| 635 |
+
"eval_samples_per_second": 1.095,
|
| 636 |
+
"eval_steps_per_second": 0.069,
|
| 637 |
+
"eval_wer": 8.60641891891892,
|
| 638 |
+
"step": 9000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 9.99,
|
| 642 |
+
"learning_rate": 9.505263157894738e-07,
|
| 643 |
+
"loss": 0.0005,
|
| 644 |
+
"step": 9100
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 10.1,
|
| 648 |
+
"learning_rate": 8.452631578947369e-07,
|
| 649 |
+
"loss": 0.0005,
|
| 650 |
+
"step": 9200
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 10.21,
|
| 654 |
+
"learning_rate": 7.4e-07,
|
| 655 |
+
"loss": 0.0003,
|
| 656 |
+
"step": 9300
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 10.32,
|
| 660 |
+
"learning_rate": 6.347368421052633e-07,
|
| 661 |
+
"loss": 0.0004,
|
| 662 |
+
"step": 9400
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 10.43,
|
| 666 |
+
"learning_rate": 5.294736842105263e-07,
|
| 667 |
+
"loss": 0.0003,
|
| 668 |
+
"step": 9500
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 10.54,
|
| 672 |
+
"learning_rate": 4.242105263157895e-07,
|
| 673 |
+
"loss": 0.0004,
|
| 674 |
+
"step": 9600
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 10.65,
|
| 678 |
+
"learning_rate": 3.1894736842105263e-07,
|
| 679 |
+
"loss": 0.0003,
|
| 680 |
+
"step": 9700
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 10.76,
|
| 684 |
+
"learning_rate": 2.136842105263158e-07,
|
| 685 |
+
"loss": 0.0003,
|
| 686 |
+
"step": 9800
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 10.87,
|
| 690 |
+
"learning_rate": 1.0842105263157895e-07,
|
| 691 |
+
"loss": 0.0004,
|
| 692 |
+
"step": 9900
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 10.98,
|
| 696 |
+
"learning_rate": 3.1578947368421054e-09,
|
| 697 |
+
"loss": 0.0003,
|
| 698 |
+
"step": 10000
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 10.98,
|
| 702 |
+
"eval_cer": 5.469766463839351,
|
| 703 |
+
"eval_loss": 0.2836119830608368,
|
| 704 |
+
"eval_runtime": 4263.2576,
|
| 705 |
+
"eval_samples_per_second": 1.08,
|
| 706 |
+
"eval_steps_per_second": 0.068,
|
| 707 |
+
"eval_wer": 8.721283783783784,
|
| 708 |
+
"step": 10000
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 10.98,
|
| 712 |
+
"step": 10000,
|
| 713 |
+
"total_flos": 3.265323341119488e+20,
|
| 714 |
+
"train_loss": 0.043234722255170346,
|
| 715 |
+
"train_runtime": 85032.8136,
|
| 716 |
+
"train_samples_per_second": 3.763,
|
| 717 |
+
"train_steps_per_second": 0.118
|
| 718 |
+
}
|
| 719 |
+
],
|
| 720 |
+
"max_steps": 10000,
|
| 721 |
+
"num_train_epochs": 11,
|
| 722 |
+
"total_flos": 3.265323341119488e+20,
|
| 723 |
+
"trial_name": null,
|
| 724 |
+
"trial_params": null
|
| 725 |
+
}
|