End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_cer": 0.06809827094571802,
|
| 4 |
+
"eval_loss": 0.31785446405410767,
|
| 5 |
+
"eval_runtime": 348.931,
|
| 6 |
+
"eval_samples": 10143,
|
| 7 |
+
"eval_samples_per_second": 29.069,
|
| 8 |
+
"eval_steps_per_second": 3.634,
|
| 9 |
+
"eval_wer": 0.2862633203284225,
|
| 10 |
+
"train_loss": 0.2775966154790558,
|
| 11 |
+
"train_runtime": 58895.0057,
|
| 12 |
+
"train_samples": 36125,
|
| 13 |
+
"train_samples_per_second": 18.401,
|
| 14 |
+
"train_steps_per_second": 0.288
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_cer": 0.06809827094571802,
|
| 4 |
+
"eval_loss": 0.31785446405410767,
|
| 5 |
+
"eval_runtime": 348.931,
|
| 6 |
+
"eval_samples": 10143,
|
| 7 |
+
"eval_samples_per_second": 29.069,
|
| 8 |
+
"eval_steps_per_second": 3.634,
|
| 9 |
+
"eval_wer": 0.2862633203284225
|
| 10 |
+
}
|
runs/Jan18_15-30-29_129-159-37-182/events.out.tfevents.1674115180.129-159-37-182.1516213.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:07ead6f53c4f4049a05afd88778f324bbe8afc30ca568b6ad6f4f541791bd0d5
|
| 3 |
+
size 412
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"train_loss": 0.2775966154790558,
|
| 4 |
+
"train_runtime": 58895.0057,
|
| 5 |
+
"train_samples": 36125,
|
| 6 |
+
"train_samples_per_second": 18.401,
|
| 7 |
+
"train_steps_per_second": 0.288
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,643 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
+
"global_step": 16950,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.71,
|
| 12 |
+
"eval_cer": 0.479669124113056,
|
| 13 |
+
"eval_loss": 1.7289525270462036,
|
| 14 |
+
"eval_runtime": 352.9433,
|
| 15 |
+
"eval_samples_per_second": 28.738,
|
| 16 |
+
"eval_steps_per_second": 3.593,
|
| 17 |
+
"eval_wer": 0.9804149925270289,
|
| 18 |
+
"step": 400
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 0.88,
|
| 22 |
+
"learning_rate": 0.0002982,
|
| 23 |
+
"loss": 4.5435,
|
| 24 |
+
"step": 500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.42,
|
| 28 |
+
"eval_cer": 0.1449701156451013,
|
| 29 |
+
"eval_loss": 0.48100030422210693,
|
| 30 |
+
"eval_runtime": 355.0295,
|
| 31 |
+
"eval_samples_per_second": 28.569,
|
| 32 |
+
"eval_steps_per_second": 3.572,
|
| 33 |
+
"eval_wer": 0.5774374502610687,
|
| 34 |
+
"step": 800
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 1.77,
|
| 38 |
+
"learning_rate": 0.0002909361702127659,
|
| 39 |
+
"loss": 0.523,
|
| 40 |
+
"step": 1000
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 2.12,
|
| 44 |
+
"eval_cer": 0.11564221810245158,
|
| 45 |
+
"eval_loss": 0.3859069347381592,
|
| 46 |
+
"eval_runtime": 358.2561,
|
| 47 |
+
"eval_samples_per_second": 28.312,
|
| 48 |
+
"eval_steps_per_second": 3.539,
|
| 49 |
+
"eval_wer": 0.48120111026999746,
|
| 50 |
+
"step": 1200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 2.65,
|
| 54 |
+
"learning_rate": 0.0002818176291793313,
|
| 55 |
+
"loss": 0.3449,
|
| 56 |
+
"step": 1500
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 2.83,
|
| 60 |
+
"eval_cer": 0.1094981215960235,
|
| 61 |
+
"eval_loss": 0.34923675656318665,
|
| 62 |
+
"eval_runtime": 353.6657,
|
| 63 |
+
"eval_samples_per_second": 28.68,
|
| 64 |
+
"eval_steps_per_second": 3.585,
|
| 65 |
+
"eval_wer": 0.44977581086589413,
|
| 66 |
+
"step": 1600
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.54,
|
| 70 |
+
"learning_rate": 0.0002726990881458966,
|
| 71 |
+
"loss": 0.2814,
|
| 72 |
+
"step": 2000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 3.54,
|
| 76 |
+
"eval_cer": 0.10994501740010437,
|
| 77 |
+
"eval_loss": 0.36604171991348267,
|
| 78 |
+
"eval_runtime": 352.2579,
|
| 79 |
+
"eval_samples_per_second": 28.794,
|
| 80 |
+
"eval_steps_per_second": 3.6,
|
| 81 |
+
"eval_wer": 0.44657310895009605,
|
| 82 |
+
"step": 2000
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 4.25,
|
| 86 |
+
"eval_cer": 0.10429683107626926,
|
| 87 |
+
"eval_loss": 0.37658488750457764,
|
| 88 |
+
"eval_runtime": 352.034,
|
| 89 |
+
"eval_samples_per_second": 28.813,
|
| 90 |
+
"eval_steps_per_second": 3.602,
|
| 91 |
+
"eval_wer": 0.4235330654709913,
|
| 92 |
+
"step": 2400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.42,
|
| 96 |
+
"learning_rate": 0.000263580547112462,
|
| 97 |
+
"loss": 0.2463,
|
| 98 |
+
"step": 2500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 4.96,
|
| 102 |
+
"eval_cer": 0.10097250293365471,
|
| 103 |
+
"eval_loss": 0.34164857864379883,
|
| 104 |
+
"eval_runtime": 355.8322,
|
| 105 |
+
"eval_samples_per_second": 28.505,
|
| 106 |
+
"eval_steps_per_second": 3.563,
|
| 107 |
+
"eval_wer": 0.4119256973155535,
|
| 108 |
+
"step": 2800
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 5.31,
|
| 112 |
+
"learning_rate": 0.0002544620060790273,
|
| 113 |
+
"loss": 0.2296,
|
| 114 |
+
"step": 3000
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 5.66,
|
| 118 |
+
"eval_cer": 0.09793072826716873,
|
| 119 |
+
"eval_loss": 0.3322136402130127,
|
| 120 |
+
"eval_runtime": 355.102,
|
| 121 |
+
"eval_samples_per_second": 28.564,
|
| 122 |
+
"eval_steps_per_second": 3.571,
|
| 123 |
+
"eval_wer": 0.4012500242628933,
|
| 124 |
+
"step": 3200
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 6.19,
|
| 128 |
+
"learning_rate": 0.00024534346504559266,
|
| 129 |
+
"loss": 0.2143,
|
| 130 |
+
"step": 3500
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 6.37,
|
| 134 |
+
"eval_cer": 0.09717821339707125,
|
| 135 |
+
"eval_loss": 0.3369796872138977,
|
| 136 |
+
"eval_runtime": 355.7454,
|
| 137 |
+
"eval_samples_per_second": 28.512,
|
| 138 |
+
"eval_steps_per_second": 3.564,
|
| 139 |
+
"eval_wer": 0.395601622702304,
|
| 140 |
+
"step": 3600
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 7.08,
|
| 144 |
+
"learning_rate": 0.00023622492401215801,
|
| 145 |
+
"loss": 0.1955,
|
| 146 |
+
"step": 4000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 7.08,
|
| 150 |
+
"eval_cer": 0.09977020906074034,
|
| 151 |
+
"eval_loss": 0.3401270806789398,
|
| 152 |
+
"eval_runtime": 359.7076,
|
| 153 |
+
"eval_samples_per_second": 28.198,
|
| 154 |
+
"eval_steps_per_second": 3.525,
|
| 155 |
+
"eval_wer": 0.40330751761486056,
|
| 156 |
+
"step": 4000
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 7.79,
|
| 160 |
+
"eval_cer": 0.09622099141671736,
|
| 161 |
+
"eval_loss": 0.33754295110702515,
|
| 162 |
+
"eval_runtime": 352.1651,
|
| 163 |
+
"eval_samples_per_second": 28.802,
|
| 164 |
+
"eval_steps_per_second": 3.601,
|
| 165 |
+
"eval_wer": 0.3889244744657311,
|
| 166 |
+
"step": 4400
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 7.96,
|
| 170 |
+
"learning_rate": 0.0002271063829787234,
|
| 171 |
+
"loss": 0.1845,
|
| 172 |
+
"step": 4500
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 8.5,
|
| 176 |
+
"eval_cer": 0.09233155632184571,
|
| 177 |
+
"eval_loss": 0.34551626443862915,
|
| 178 |
+
"eval_runtime": 357.5651,
|
| 179 |
+
"eval_samples_per_second": 28.367,
|
| 180 |
+
"eval_steps_per_second": 3.546,
|
| 181 |
+
"eval_wer": 0.37524020264368485,
|
| 182 |
+
"step": 4800
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 8.85,
|
| 186 |
+
"learning_rate": 0.00021798784194528871,
|
| 187 |
+
"loss": 0.1752,
|
| 188 |
+
"step": 5000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 9.2,
|
| 192 |
+
"eval_cer": 0.09245553386749396,
|
| 193 |
+
"eval_loss": 0.3335849642753601,
|
| 194 |
+
"eval_runtime": 351.3613,
|
| 195 |
+
"eval_samples_per_second": 28.868,
|
| 196 |
+
"eval_steps_per_second": 3.609,
|
| 197 |
+
"eval_wer": 0.37176575632291,
|
| 198 |
+
"step": 5200
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 9.73,
|
| 202 |
+
"learning_rate": 0.0002088693009118541,
|
| 203 |
+
"loss": 0.1705,
|
| 204 |
+
"step": 5500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 9.91,
|
| 208 |
+
"eval_cer": 0.08918598650086351,
|
| 209 |
+
"eval_loss": 0.3145359754562378,
|
| 210 |
+
"eval_runtime": 355.7051,
|
| 211 |
+
"eval_samples_per_second": 28.515,
|
| 212 |
+
"eval_steps_per_second": 3.565,
|
| 213 |
+
"eval_wer": 0.3653021215473903,
|
| 214 |
+
"step": 5600
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 10.62,
|
| 218 |
+
"learning_rate": 0.00019975075987841941,
|
| 219 |
+
"loss": 0.1585,
|
| 220 |
+
"step": 6000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 10.62,
|
| 224 |
+
"eval_cer": 0.09218451318630942,
|
| 225 |
+
"eval_loss": 0.34097233414649963,
|
| 226 |
+
"eval_runtime": 352.3797,
|
| 227 |
+
"eval_samples_per_second": 28.784,
|
| 228 |
+
"eval_steps_per_second": 3.598,
|
| 229 |
+
"eval_wer": 0.37370678778703004,
|
| 230 |
+
"step": 6000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 11.33,
|
| 234 |
+
"eval_cer": 0.08989237019118491,
|
| 235 |
+
"eval_loss": 0.3296053409576416,
|
| 236 |
+
"eval_runtime": 350.4695,
|
| 237 |
+
"eval_samples_per_second": 28.941,
|
| 238 |
+
"eval_steps_per_second": 3.618,
|
| 239 |
+
"eval_wer": 0.3664279197965799,
|
| 240 |
+
"step": 6400
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 11.5,
|
| 244 |
+
"learning_rate": 0.0001906322188449848,
|
| 245 |
+
"loss": 0.1474,
|
| 246 |
+
"step": 6500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 12.04,
|
| 250 |
+
"eval_cer": 0.08988083739624089,
|
| 251 |
+
"eval_loss": 0.34918734431266785,
|
| 252 |
+
"eval_runtime": 352.3033,
|
| 253 |
+
"eval_samples_per_second": 28.791,
|
| 254 |
+
"eval_steps_per_second": 3.599,
|
| 255 |
+
"eval_wer": 0.3589937692890002,
|
| 256 |
+
"step": 6800
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 12.39,
|
| 260 |
+
"learning_rate": 0.00018153191489361702,
|
| 261 |
+
"loss": 0.1485,
|
| 262 |
+
"step": 7000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 12.74,
|
| 266 |
+
"eval_cer": 0.08672950117778669,
|
| 267 |
+
"eval_loss": 0.31763964891433716,
|
| 268 |
+
"eval_runtime": 352.8906,
|
| 269 |
+
"eval_samples_per_second": 28.743,
|
| 270 |
+
"eval_steps_per_second": 3.593,
|
| 271 |
+
"eval_wer": 0.3506085133640016,
|
| 272 |
+
"step": 7200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 13.27,
|
| 276 |
+
"learning_rate": 0.00017241337386018235,
|
| 277 |
+
"loss": 0.137,
|
| 278 |
+
"step": 7500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 13.45,
|
| 282 |
+
"eval_cer": 0.08901587777543918,
|
| 283 |
+
"eval_loss": 0.3532153367996216,
|
| 284 |
+
"eval_runtime": 357.5465,
|
| 285 |
+
"eval_samples_per_second": 28.368,
|
| 286 |
+
"eval_steps_per_second": 3.546,
|
| 287 |
+
"eval_wer": 0.360041926279625,
|
| 288 |
+
"step": 7600
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 14.16,
|
| 292 |
+
"learning_rate": 0.00016329483282674772,
|
| 293 |
+
"loss": 0.1291,
|
| 294 |
+
"step": 8000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 14.16,
|
| 298 |
+
"eval_cer": 0.08733497291234787,
|
| 299 |
+
"eval_loss": 0.33181944489479065,
|
| 300 |
+
"eval_runtime": 352.5023,
|
| 301 |
+
"eval_samples_per_second": 28.774,
|
| 302 |
+
"eval_steps_per_second": 3.597,
|
| 303 |
+
"eval_wer": 0.3570527378248801,
|
| 304 |
+
"step": 8000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 14.87,
|
| 308 |
+
"eval_cer": 0.08829796129017377,
|
| 309 |
+
"eval_loss": 0.33532437682151794,
|
| 310 |
+
"eval_runtime": 357.1338,
|
| 311 |
+
"eval_samples_per_second": 28.401,
|
| 312 |
+
"eval_steps_per_second": 3.55,
|
| 313 |
+
"eval_wer": 0.3547623206972185,
|
| 314 |
+
"step": 8400
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 15.04,
|
| 318 |
+
"learning_rate": 0.00015417629179331305,
|
| 319 |
+
"loss": 0.1274,
|
| 320 |
+
"step": 8500
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 15.58,
|
| 324 |
+
"eval_cer": 0.08226342633571389,
|
| 325 |
+
"eval_loss": 0.32346823811531067,
|
| 326 |
+
"eval_runtime": 349.5389,
|
| 327 |
+
"eval_samples_per_second": 29.018,
|
| 328 |
+
"eval_steps_per_second": 3.628,
|
| 329 |
+
"eval_wer": 0.339602864962441,
|
| 330 |
+
"step": 8800
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 15.93,
|
| 334 |
+
"learning_rate": 0.00014505775075987842,
|
| 335 |
+
"loss": 0.1198,
|
| 336 |
+
"step": 9000
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 16.28,
|
| 340 |
+
"eval_cer": 0.08322353151480379,
|
| 341 |
+
"eval_loss": 0.32590439915657043,
|
| 342 |
+
"eval_runtime": 352.9664,
|
| 343 |
+
"eval_samples_per_second": 28.736,
|
| 344 |
+
"eval_steps_per_second": 3.592,
|
| 345 |
+
"eval_wer": 0.33894291426464024,
|
| 346 |
+
"step": 9200
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 16.81,
|
| 350 |
+
"learning_rate": 0.00013595744680851063,
|
| 351 |
+
"loss": 0.1164,
|
| 352 |
+
"step": 9500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 16.99,
|
| 356 |
+
"eval_cer": 0.084353745419318,
|
| 357 |
+
"eval_loss": 0.32632604241371155,
|
| 358 |
+
"eval_runtime": 355.0498,
|
| 359 |
+
"eval_samples_per_second": 28.568,
|
| 360 |
+
"eval_steps_per_second": 3.571,
|
| 361 |
+
"eval_wer": 0.3411362798190959,
|
| 362 |
+
"step": 9600
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 17.7,
|
| 366 |
+
"learning_rate": 0.00012683890577507598,
|
| 367 |
+
"loss": 0.1119,
|
| 368 |
+
"step": 10000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 17.7,
|
| 372 |
+
"eval_cer": 0.08243353506113824,
|
| 373 |
+
"eval_loss": 0.32535773515701294,
|
| 374 |
+
"eval_runtime": 352.3077,
|
| 375 |
+
"eval_samples_per_second": 28.79,
|
| 376 |
+
"eval_steps_per_second": 3.599,
|
| 377 |
+
"eval_wer": 0.3377006541276034,
|
| 378 |
+
"step": 10000
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 18.41,
|
| 382 |
+
"eval_cer": 0.0811533948223517,
|
| 383 |
+
"eval_loss": 0.3243008255958557,
|
| 384 |
+
"eval_runtime": 356.0414,
|
| 385 |
+
"eval_samples_per_second": 28.488,
|
| 386 |
+
"eval_steps_per_second": 3.561,
|
| 387 |
+
"eval_wer": 0.3330615889283565,
|
| 388 |
+
"step": 10400
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 18.58,
|
| 392 |
+
"learning_rate": 0.00011772036474164133,
|
| 393 |
+
"loss": 0.1054,
|
| 394 |
+
"step": 10500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 19.12,
|
| 398 |
+
"eval_cer": 0.07895063098804338,
|
| 399 |
+
"eval_loss": 0.32234683632850647,
|
| 400 |
+
"eval_runtime": 353.14,
|
| 401 |
+
"eval_samples_per_second": 28.722,
|
| 402 |
+
"eval_steps_per_second": 3.591,
|
| 403 |
+
"eval_wer": 0.3239387410469924,
|
| 404 |
+
"step": 10800
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 19.47,
|
| 408 |
+
"learning_rate": 0.00010860182370820666,
|
| 409 |
+
"loss": 0.1017,
|
| 410 |
+
"step": 11000
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 19.82,
|
| 414 |
+
"eval_cer": 0.07741965245922436,
|
| 415 |
+
"eval_loss": 0.305361270904541,
|
| 416 |
+
"eval_runtime": 348.3343,
|
| 417 |
+
"eval_samples_per_second": 29.119,
|
| 418 |
+
"eval_steps_per_second": 3.64,
|
| 419 |
+
"eval_wer": 0.3189502901842039,
|
| 420 |
+
"step": 11200
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 20.35,
|
| 424 |
+
"learning_rate": 9.948328267477204e-05,
|
| 425 |
+
"loss": 0.0964,
|
| 426 |
+
"step": 11500
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 20.53,
|
| 430 |
+
"eval_cer": 0.07850373518396249,
|
| 431 |
+
"eval_loss": 0.32777705788612366,
|
| 432 |
+
"eval_runtime": 353.3356,
|
| 433 |
+
"eval_samples_per_second": 28.706,
|
| 434 |
+
"eval_steps_per_second": 3.589,
|
| 435 |
+
"eval_wer": 0.3236669966420156,
|
| 436 |
+
"step": 11600
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 21.24,
|
| 440 |
+
"learning_rate": 9.036474164133739e-05,
|
| 441 |
+
"loss": 0.0903,
|
| 442 |
+
"step": 12000
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 21.24,
|
| 446 |
+
"eval_cer": 0.07744560124784841,
|
| 447 |
+
"eval_loss": 0.3166551887989044,
|
| 448 |
+
"eval_runtime": 353.0166,
|
| 449 |
+
"eval_samples_per_second": 28.732,
|
| 450 |
+
"eval_steps_per_second": 3.592,
|
| 451 |
+
"eval_wer": 0.3177274403618083,
|
| 452 |
+
"step": 12000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 21.95,
|
| 456 |
+
"eval_cer": 0.07655469283842266,
|
| 457 |
+
"eval_loss": 0.33310163021087646,
|
| 458 |
+
"eval_runtime": 354.4672,
|
| 459 |
+
"eval_samples_per_second": 28.615,
|
| 460 |
+
"eval_steps_per_second": 3.577,
|
| 461 |
+
"eval_wer": 0.3124478347794018,
|
| 462 |
+
"step": 12400
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 22.12,
|
| 466 |
+
"learning_rate": 8.124620060790274e-05,
|
| 467 |
+
"loss": 0.0886,
|
| 468 |
+
"step": 12500
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 22.65,
|
| 472 |
+
"eval_cer": 0.07452492092827466,
|
| 473 |
+
"eval_loss": 0.3098578155040741,
|
| 474 |
+
"eval_runtime": 354.4398,
|
| 475 |
+
"eval_samples_per_second": 28.617,
|
| 476 |
+
"eval_steps_per_second": 3.577,
|
| 477 |
+
"eval_wer": 0.30889574720006213,
|
| 478 |
+
"step": 12800
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 23.01,
|
| 482 |
+
"learning_rate": 7.214589665653494e-05,
|
| 483 |
+
"loss": 0.0836,
|
| 484 |
+
"step": 13000
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 23.36,
|
| 488 |
+
"eval_cer": 0.07314963513119996,
|
| 489 |
+
"eval_loss": 0.3170570433139801,
|
| 490 |
+
"eval_runtime": 351.2874,
|
| 491 |
+
"eval_samples_per_second": 28.874,
|
| 492 |
+
"eval_steps_per_second": 3.61,
|
| 493 |
+
"eval_wer": 0.3047613501814864,
|
| 494 |
+
"step": 13200
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 23.89,
|
| 498 |
+
"learning_rate": 6.30273556231003e-05,
|
| 499 |
+
"loss": 0.0796,
|
| 500 |
+
"step": 13500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 24.07,
|
| 504 |
+
"eval_cer": 0.07325919668316817,
|
| 505 |
+
"eval_loss": 0.315768837928772,
|
| 506 |
+
"eval_runtime": 354.0965,
|
| 507 |
+
"eval_samples_per_second": 28.645,
|
| 508 |
+
"eval_steps_per_second": 3.581,
|
| 509 |
+
"eval_wer": 0.30410139948368564,
|
| 510 |
+
"step": 13600
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 24.78,
|
| 514 |
+
"learning_rate": 5.390881458966565e-05,
|
| 515 |
+
"loss": 0.0739,
|
| 516 |
+
"step": 14000
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 24.78,
|
| 520 |
+
"eval_cer": 0.07206266920772582,
|
| 521 |
+
"eval_loss": 0.3202644884586334,
|
| 522 |
+
"eval_runtime": 351.0812,
|
| 523 |
+
"eval_samples_per_second": 28.891,
|
| 524 |
+
"eval_steps_per_second": 3.612,
|
| 525 |
+
"eval_wer": 0.3002775674993692,
|
| 526 |
+
"step": 14000
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 25.49,
|
| 530 |
+
"eval_cer": 0.07125249036290822,
|
| 531 |
+
"eval_loss": 0.3138331472873688,
|
| 532 |
+
"eval_runtime": 353.8044,
|
| 533 |
+
"eval_samples_per_second": 28.668,
|
| 534 |
+
"eval_steps_per_second": 3.584,
|
| 535 |
+
"eval_wer": 0.2973660203031891,
|
| 536 |
+
"step": 14400
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 25.66,
|
| 540 |
+
"learning_rate": 4.4790273556231e-05,
|
| 541 |
+
"loss": 0.0742,
|
| 542 |
+
"step": 14500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 26.19,
|
| 546 |
+
"eval_cer": 0.07109103123369191,
|
| 547 |
+
"eval_loss": 0.3196839392185211,
|
| 548 |
+
"eval_runtime": 353.996,
|
| 549 |
+
"eval_samples_per_second": 28.653,
|
| 550 |
+
"eval_steps_per_second": 3.582,
|
| 551 |
+
"eval_wer": 0.2958520157611755,
|
| 552 |
+
"step": 14800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 26.55,
|
| 556 |
+
"learning_rate": 3.568996960486322e-05,
|
| 557 |
+
"loss": 0.07,
|
| 558 |
+
"step": 15000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 26.9,
|
| 562 |
+
"eval_cer": 0.07031256757497037,
|
| 563 |
+
"eval_loss": 0.3232352137565613,
|
| 564 |
+
"eval_runtime": 355.2159,
|
| 565 |
+
"eval_samples_per_second": 28.554,
|
| 566 |
+
"eval_steps_per_second": 3.57,
|
| 567 |
+
"eval_wer": 0.2951726547487335,
|
| 568 |
+
"step": 15200
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 27.43,
|
| 572 |
+
"learning_rate": 2.6571428571428566e-05,
|
| 573 |
+
"loss": 0.0654,
|
| 574 |
+
"step": 15500
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 27.61,
|
| 578 |
+
"eval_cer": 0.07008479487482593,
|
| 579 |
+
"eval_loss": 0.3242589831352234,
|
| 580 |
+
"eval_runtime": 352.5643,
|
| 581 |
+
"eval_samples_per_second": 28.769,
|
| 582 |
+
"eval_steps_per_second": 3.597,
|
| 583 |
+
"eval_wer": 0.2938527533531319,
|
| 584 |
+
"step": 15600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 28.32,
|
| 588 |
+
"learning_rate": 1.745288753799392e-05,
|
| 589 |
+
"loss": 0.0631,
|
| 590 |
+
"step": 16000
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 28.32,
|
| 594 |
+
"eval_cer": 0.06881907062971944,
|
| 595 |
+
"eval_loss": 0.3212898373603821,
|
| 596 |
+
"eval_runtime": 352.1058,
|
| 597 |
+
"eval_samples_per_second": 28.807,
|
| 598 |
+
"eval_steps_per_second": 3.601,
|
| 599 |
+
"eval_wer": 0.2875638114093829,
|
| 600 |
+
"step": 16000
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 29.03,
|
| 604 |
+
"eval_cer": 0.06853940035232689,
|
| 605 |
+
"eval_loss": 0.3151107728481293,
|
| 606 |
+
"eval_runtime": 353.743,
|
| 607 |
+
"eval_samples_per_second": 28.673,
|
| 608 |
+
"eval_steps_per_second": 3.585,
|
| 609 |
+
"eval_wer": 0.28799083833148936,
|
| 610 |
+
"step": 16400
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 29.2,
|
| 614 |
+
"learning_rate": 8.33434650455927e-06,
|
| 615 |
+
"loss": 0.0607,
|
| 616 |
+
"step": 16500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 29.73,
|
| 620 |
+
"eval_cer": 0.06810115414445403,
|
| 621 |
+
"eval_loss": 0.31835824251174927,
|
| 622 |
+
"eval_runtime": 352.8555,
|
| 623 |
+
"eval_samples_per_second": 28.745,
|
| 624 |
+
"eval_steps_per_second": 3.594,
|
| 625 |
+
"eval_wer": 0.28665152662124654,
|
| 626 |
+
"step": 16800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 30.0,
|
| 630 |
+
"step": 16950,
|
| 631 |
+
"total_flos": 1.1955997003691401e+20,
|
| 632 |
+
"train_loss": 0.2775966154790558,
|
| 633 |
+
"train_runtime": 58895.0057,
|
| 634 |
+
"train_samples_per_second": 18.401,
|
| 635 |
+
"train_steps_per_second": 0.288
|
| 636 |
+
}
|
| 637 |
+
],
|
| 638 |
+
"max_steps": 16950,
|
| 639 |
+
"num_train_epochs": 30,
|
| 640 |
+
"total_flos": 1.1955997003691401e+20,
|
| 641 |
+
"trial_name": null,
|
| 642 |
+
"trial_params": null
|
| 643 |
+
}
|