Commit
•
3db4946
1
Parent(s):
3606b39
End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +664 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_loss": 0.1659538894891739,
|
| 4 |
+
"eval_runtime": 68.9631,
|
| 5 |
+
"eval_samples": 2148,
|
| 6 |
+
"eval_samples_per_second": 31.147,
|
| 7 |
+
"eval_steps_per_second": 1.958,
|
| 8 |
+
"eval_wer": 0.1704645476772616,
|
| 9 |
+
"train_loss": 1.4054935827010717,
|
| 10 |
+
"train_runtime": 11681.5311,
|
| 11 |
+
"train_samples": 4963,
|
| 12 |
+
"train_samples_per_second": 21.243,
|
| 13 |
+
"train_steps_per_second": 0.668
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_loss": 0.1659538894891739,
|
| 4 |
+
"eval_runtime": 68.9631,
|
| 5 |
+
"eval_samples": 2148,
|
| 6 |
+
"eval_samples_per_second": 31.147,
|
| 7 |
+
"eval_steps_per_second": 1.958,
|
| 8 |
+
"eval_wer": 0.1704645476772616
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"train_loss": 1.4054935827010717,
|
| 4 |
+
"train_runtime": 11681.5311,
|
| 5 |
+
"train_samples": 4963,
|
| 6 |
+
"train_samples_per_second": 21.243,
|
| 7 |
+
"train_steps_per_second": 0.668
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,664 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"global_step": 7800,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.64,
|
| 12 |
+
"learning_rate": 7.275e-06,
|
| 13 |
+
"loss": 16.819,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.28,
|
| 18 |
+
"learning_rate": 1.4775e-05,
|
| 19 |
+
"loss": 6.1908,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.92,
|
| 24 |
+
"learning_rate": 2.2274999999999996e-05,
|
| 25 |
+
"loss": 4.1987,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.56,
|
| 30 |
+
"learning_rate": 2.9775e-05,
|
| 31 |
+
"loss": 3.489,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.56,
|
| 36 |
+
"eval_loss": 3.3590216636657715,
|
| 37 |
+
"eval_runtime": 66.6533,
|
| 38 |
+
"eval_samples_per_second": 32.226,
|
| 39 |
+
"eval_steps_per_second": 2.025,
|
| 40 |
+
"eval_wer": 1.0,
|
| 41 |
+
"step": 400
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 3.21,
|
| 45 |
+
"learning_rate": 3.7275e-05,
|
| 46 |
+
"loss": 3.2231,
|
| 47 |
+
"step": 500
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.85,
|
| 51 |
+
"learning_rate": 4.4775e-05,
|
| 52 |
+
"loss": 3.1038,
|
| 53 |
+
"step": 600
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 4.49,
|
| 57 |
+
"learning_rate": 5.227499999999999e-05,
|
| 58 |
+
"loss": 3.0217,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 5.13,
|
| 63 |
+
"learning_rate": 5.9774999999999996e-05,
|
| 64 |
+
"loss": 2.9903,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 5.13,
|
| 69 |
+
"eval_loss": 2.970390796661377,
|
| 70 |
+
"eval_runtime": 66.5925,
|
| 71 |
+
"eval_samples_per_second": 32.256,
|
| 72 |
+
"eval_steps_per_second": 2.027,
|
| 73 |
+
"eval_wer": 1.0000977995110025,
|
| 74 |
+
"step": 800
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 5.77,
|
| 78 |
+
"learning_rate": 6.7275e-05,
|
| 79 |
+
"loss": 2.906,
|
| 80 |
+
"step": 900
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 6.41,
|
| 84 |
+
"learning_rate": 7.477499999999999e-05,
|
| 85 |
+
"loss": 2.5346,
|
| 86 |
+
"step": 1000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 7.05,
|
| 90 |
+
"learning_rate": 7.393014705882353e-05,
|
| 91 |
+
"loss": 1.8923,
|
| 92 |
+
"step": 1100
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 7.69,
|
| 96 |
+
"learning_rate": 7.282720588235294e-05,
|
| 97 |
+
"loss": 1.6712,
|
| 98 |
+
"step": 1200
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 7.69,
|
| 102 |
+
"eval_loss": 0.6178616285324097,
|
| 103 |
+
"eval_runtime": 66.5699,
|
| 104 |
+
"eval_samples_per_second": 32.267,
|
| 105 |
+
"eval_steps_per_second": 2.028,
|
| 106 |
+
"eval_wer": 0.6566259168704156,
|
| 107 |
+
"step": 1200
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 8.33,
|
| 111 |
+
"learning_rate": 7.172426470588234e-05,
|
| 112 |
+
"loss": 1.5317,
|
| 113 |
+
"step": 1300
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 8.97,
|
| 117 |
+
"learning_rate": 7.062132352941176e-05,
|
| 118 |
+
"loss": 1.4151,
|
| 119 |
+
"step": 1400
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 9.62,
|
| 123 |
+
"learning_rate": 6.951838235294117e-05,
|
| 124 |
+
"loss": 1.3153,
|
| 125 |
+
"step": 1500
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 10.26,
|
| 129 |
+
"learning_rate": 6.841544117647059e-05,
|
| 130 |
+
"loss": 1.2635,
|
| 131 |
+
"step": 1600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 10.26,
|
| 135 |
+
"eval_loss": 0.31762251257896423,
|
| 136 |
+
"eval_runtime": 67.311,
|
| 137 |
+
"eval_samples_per_second": 31.912,
|
| 138 |
+
"eval_steps_per_second": 2.006,
|
| 139 |
+
"eval_wer": 0.45310513447432765,
|
| 140 |
+
"step": 1600
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 10.9,
|
| 144 |
+
"learning_rate": 6.731249999999999e-05,
|
| 145 |
+
"loss": 1.1929,
|
| 146 |
+
"step": 1700
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 11.54,
|
| 150 |
+
"learning_rate": 6.62095588235294e-05,
|
| 151 |
+
"loss": 1.166,
|
| 152 |
+
"step": 1800
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 12.18,
|
| 156 |
+
"learning_rate": 6.510661764705882e-05,
|
| 157 |
+
"loss": 1.0971,
|
| 158 |
+
"step": 1900
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 12.82,
|
| 162 |
+
"learning_rate": 6.400367647058824e-05,
|
| 163 |
+
"loss": 1.0819,
|
| 164 |
+
"step": 2000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 12.82,
|
| 168 |
+
"eval_loss": 0.2516830265522003,
|
| 169 |
+
"eval_runtime": 66.7082,
|
| 170 |
+
"eval_samples_per_second": 32.2,
|
| 171 |
+
"eval_steps_per_second": 2.024,
|
| 172 |
+
"eval_wer": 0.35080684596577016,
|
| 173 |
+
"step": 2000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 13.46,
|
| 177 |
+
"learning_rate": 6.290073529411764e-05,
|
| 178 |
+
"loss": 1.0892,
|
| 179 |
+
"step": 2100
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 14.1,
|
| 183 |
+
"learning_rate": 6.179779411764705e-05,
|
| 184 |
+
"loss": 1.0534,
|
| 185 |
+
"step": 2200
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 14.74,
|
| 189 |
+
"learning_rate": 6.069485294117646e-05,
|
| 190 |
+
"loss": 1.012,
|
| 191 |
+
"step": 2300
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 15.38,
|
| 195 |
+
"learning_rate": 5.9591911764705876e-05,
|
| 196 |
+
"loss": 1.0136,
|
| 197 |
+
"step": 2400
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 15.38,
|
| 201 |
+
"eval_loss": 0.22566433250904083,
|
| 202 |
+
"eval_runtime": 66.3228,
|
| 203 |
+
"eval_samples_per_second": 32.387,
|
| 204 |
+
"eval_steps_per_second": 2.036,
|
| 205 |
+
"eval_wer": 0.3123716381418093,
|
| 206 |
+
"step": 2400
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 16.03,
|
| 210 |
+
"learning_rate": 5.848897058823529e-05,
|
| 211 |
+
"loss": 0.9991,
|
| 212 |
+
"step": 2500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 16.67,
|
| 216 |
+
"learning_rate": 5.73860294117647e-05,
|
| 217 |
+
"loss": 0.9722,
|
| 218 |
+
"step": 2600
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 17.31,
|
| 222 |
+
"learning_rate": 5.6283088235294115e-05,
|
| 223 |
+
"loss": 0.9794,
|
| 224 |
+
"step": 2700
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 17.95,
|
| 228 |
+
"learning_rate": 5.5180147058823523e-05,
|
| 229 |
+
"loss": 0.9625,
|
| 230 |
+
"step": 2800
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 17.95,
|
| 234 |
+
"eval_loss": 0.19747723639011383,
|
| 235 |
+
"eval_runtime": 67.1687,
|
| 236 |
+
"eval_samples_per_second": 31.979,
|
| 237 |
+
"eval_steps_per_second": 2.01,
|
| 238 |
+
"eval_wer": 0.2311002444987775,
|
| 239 |
+
"step": 2800
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 18.59,
|
| 243 |
+
"learning_rate": 5.407720588235294e-05,
|
| 244 |
+
"loss": 0.9505,
|
| 245 |
+
"step": 2900
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 19.23,
|
| 249 |
+
"learning_rate": 5.297426470588235e-05,
|
| 250 |
+
"loss": 0.9337,
|
| 251 |
+
"step": 3000
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 19.87,
|
| 255 |
+
"learning_rate": 5.187132352941176e-05,
|
| 256 |
+
"loss": 0.936,
|
| 257 |
+
"step": 3100
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 20.51,
|
| 261 |
+
"learning_rate": 5.076838235294117e-05,
|
| 262 |
+
"loss": 0.901,
|
| 263 |
+
"step": 3200
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 20.51,
|
| 267 |
+
"eval_loss": 0.19860759377479553,
|
| 268 |
+
"eval_runtime": 66.128,
|
| 269 |
+
"eval_samples_per_second": 32.482,
|
| 270 |
+
"eval_steps_per_second": 2.041,
|
| 271 |
+
"eval_wer": 0.20968215158924206,
|
| 272 |
+
"step": 3200
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 21.15,
|
| 276 |
+
"learning_rate": 4.9665441176470586e-05,
|
| 277 |
+
"loss": 0.9179,
|
| 278 |
+
"step": 3300
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 21.79,
|
| 282 |
+
"learning_rate": 4.8562499999999995e-05,
|
| 283 |
+
"loss": 0.8963,
|
| 284 |
+
"step": 3400
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 22.44,
|
| 288 |
+
"learning_rate": 4.745955882352941e-05,
|
| 289 |
+
"loss": 0.8826,
|
| 290 |
+
"step": 3500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 23.08,
|
| 294 |
+
"learning_rate": 4.635661764705882e-05,
|
| 295 |
+
"loss": 0.8842,
|
| 296 |
+
"step": 3600
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 23.08,
|
| 300 |
+
"eval_loss": 0.19037693738937378,
|
| 301 |
+
"eval_runtime": 67.9694,
|
| 302 |
+
"eval_samples_per_second": 31.602,
|
| 303 |
+
"eval_steps_per_second": 1.986,
|
| 304 |
+
"eval_wer": 0.2039119804400978,
|
| 305 |
+
"step": 3600
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 23.72,
|
| 309 |
+
"learning_rate": 4.5253676470588234e-05,
|
| 310 |
+
"loss": 0.8683,
|
| 311 |
+
"step": 3700
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 24.36,
|
| 315 |
+
"learning_rate": 4.416176470588235e-05,
|
| 316 |
+
"loss": 0.8599,
|
| 317 |
+
"step": 3800
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 25.0,
|
| 321 |
+
"learning_rate": 4.305882352941176e-05,
|
| 322 |
+
"loss": 0.8658,
|
| 323 |
+
"step": 3900
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 25.64,
|
| 327 |
+
"learning_rate": 4.1955882352941173e-05,
|
| 328 |
+
"loss": 0.8542,
|
| 329 |
+
"step": 4000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 25.64,
|
| 333 |
+
"eval_loss": 0.1846681386232376,
|
| 334 |
+
"eval_runtime": 67.5681,
|
| 335 |
+
"eval_samples_per_second": 31.79,
|
| 336 |
+
"eval_steps_per_second": 1.998,
|
| 337 |
+
"eval_wer": 0.19814180929095354,
|
| 338 |
+
"step": 4000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 26.28,
|
| 342 |
+
"learning_rate": 4.085294117647058e-05,
|
| 343 |
+
"loss": 0.8468,
|
| 344 |
+
"step": 4100
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 26.92,
|
| 348 |
+
"learning_rate": 3.975e-05,
|
| 349 |
+
"loss": 0.8462,
|
| 350 |
+
"step": 4200
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 27.56,
|
| 354 |
+
"learning_rate": 3.8647058823529406e-05,
|
| 355 |
+
"loss": 0.8284,
|
| 356 |
+
"step": 4300
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 28.21,
|
| 360 |
+
"learning_rate": 3.754411764705882e-05,
|
| 361 |
+
"loss": 0.8244,
|
| 362 |
+
"step": 4400
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 28.21,
|
| 366 |
+
"eval_loss": 0.1804967224597931,
|
| 367 |
+
"eval_runtime": 66.5917,
|
| 368 |
+
"eval_samples_per_second": 32.256,
|
| 369 |
+
"eval_steps_per_second": 2.027,
|
| 370 |
+
"eval_wer": 0.18474327628361858,
|
| 371 |
+
"step": 4400
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 28.85,
|
| 375 |
+
"learning_rate": 3.6441176470588236e-05,
|
| 376 |
+
"loss": 0.8288,
|
| 377 |
+
"step": 4500
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 29.49,
|
| 381 |
+
"learning_rate": 3.5338235294117645e-05,
|
| 382 |
+
"loss": 0.7904,
|
| 383 |
+
"step": 4600
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 30.13,
|
| 387 |
+
"learning_rate": 3.423529411764706e-05,
|
| 388 |
+
"loss": 0.8157,
|
| 389 |
+
"step": 4700
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 30.77,
|
| 393 |
+
"learning_rate": 3.313235294117647e-05,
|
| 394 |
+
"loss": 0.7689,
|
| 395 |
+
"step": 4800
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 30.77,
|
| 399 |
+
"eval_loss": 0.17355979979038239,
|
| 400 |
+
"eval_runtime": 69.3305,
|
| 401 |
+
"eval_samples_per_second": 30.982,
|
| 402 |
+
"eval_steps_per_second": 1.947,
|
| 403 |
+
"eval_wer": 0.18317848410757948,
|
| 404 |
+
"step": 4800
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 31.41,
|
| 408 |
+
"learning_rate": 3.202941176470588e-05,
|
| 409 |
+
"loss": 0.7948,
|
| 410 |
+
"step": 4900
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 32.05,
|
| 414 |
+
"learning_rate": 3.092647058823529e-05,
|
| 415 |
+
"loss": 0.8035,
|
| 416 |
+
"step": 5000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 32.69,
|
| 420 |
+
"learning_rate": 2.9823529411764704e-05,
|
| 421 |
+
"loss": 0.7754,
|
| 422 |
+
"step": 5100
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 33.33,
|
| 426 |
+
"learning_rate": 2.8720588235294116e-05,
|
| 427 |
+
"loss": 0.7825,
|
| 428 |
+
"step": 5200
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 33.33,
|
| 432 |
+
"eval_loss": 0.1698261797428131,
|
| 433 |
+
"eval_runtime": 67.6625,
|
| 434 |
+
"eval_samples_per_second": 31.746,
|
| 435 |
+
"eval_steps_per_second": 1.995,
|
| 436 |
+
"eval_wer": 0.18210268948655256,
|
| 437 |
+
"step": 5200
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 33.97,
|
| 441 |
+
"learning_rate": 2.7617647058823528e-05,
|
| 442 |
+
"loss": 0.7748,
|
| 443 |
+
"step": 5300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 34.62,
|
| 447 |
+
"learning_rate": 2.651470588235294e-05,
|
| 448 |
+
"loss": 0.7704,
|
| 449 |
+
"step": 5400
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 35.26,
|
| 453 |
+
"learning_rate": 2.5411764705882348e-05,
|
| 454 |
+
"loss": 0.7597,
|
| 455 |
+
"step": 5500
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 35.9,
|
| 459 |
+
"learning_rate": 2.430882352941176e-05,
|
| 460 |
+
"loss": 0.7817,
|
| 461 |
+
"step": 5600
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 35.9,
|
| 465 |
+
"eval_loss": 0.17581327259540558,
|
| 466 |
+
"eval_runtime": 67.4489,
|
| 467 |
+
"eval_samples_per_second": 31.846,
|
| 468 |
+
"eval_steps_per_second": 2.002,
|
| 469 |
+
"eval_wer": 0.18034229828850856,
|
| 470 |
+
"step": 5600
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 36.54,
|
| 474 |
+
"learning_rate": 2.3205882352941172e-05,
|
| 475 |
+
"loss": 0.744,
|
| 476 |
+
"step": 5700
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 37.18,
|
| 480 |
+
"learning_rate": 2.2102941176470584e-05,
|
| 481 |
+
"loss": 0.7508,
|
| 482 |
+
"step": 5800
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 37.82,
|
| 486 |
+
"learning_rate": 2.1e-05,
|
| 487 |
+
"loss": 0.746,
|
| 488 |
+
"step": 5900
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 38.46,
|
| 492 |
+
"learning_rate": 1.989705882352941e-05,
|
| 493 |
+
"loss": 0.7488,
|
| 494 |
+
"step": 6000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 38.46,
|
| 498 |
+
"eval_loss": 0.16634251177310944,
|
| 499 |
+
"eval_runtime": 67.3054,
|
| 500 |
+
"eval_samples_per_second": 31.914,
|
| 501 |
+
"eval_steps_per_second": 2.006,
|
| 502 |
+
"eval_wer": 0.17603911980440098,
|
| 503 |
+
"step": 6000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 39.1,
|
| 507 |
+
"learning_rate": 1.8794117647058823e-05,
|
| 508 |
+
"loss": 0.7486,
|
| 509 |
+
"step": 6100
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 39.74,
|
| 513 |
+
"learning_rate": 1.7691176470588234e-05,
|
| 514 |
+
"loss": 0.7281,
|
| 515 |
+
"step": 6200
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 40.38,
|
| 519 |
+
"learning_rate": 1.6588235294117646e-05,
|
| 520 |
+
"loss": 0.7317,
|
| 521 |
+
"step": 6300
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 41.03,
|
| 525 |
+
"learning_rate": 1.5485294117647058e-05,
|
| 526 |
+
"loss": 0.7171,
|
| 527 |
+
"step": 6400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 41.03,
|
| 531 |
+
"eval_loss": 0.16363976895809174,
|
| 532 |
+
"eval_runtime": 68.4211,
|
| 533 |
+
"eval_samples_per_second": 31.394,
|
| 534 |
+
"eval_steps_per_second": 1.973,
|
| 535 |
+
"eval_wer": 0.17212713936430318,
|
| 536 |
+
"step": 6400
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 41.67,
|
| 540 |
+
"learning_rate": 1.438235294117647e-05,
|
| 541 |
+
"loss": 0.7196,
|
| 542 |
+
"step": 6500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 42.31,
|
| 546 |
+
"learning_rate": 1.3279411764705882e-05,
|
| 547 |
+
"loss": 0.7292,
|
| 548 |
+
"step": 6600
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 42.95,
|
| 552 |
+
"learning_rate": 1.2176470588235294e-05,
|
| 553 |
+
"loss": 0.7319,
|
| 554 |
+
"step": 6700
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 43.59,
|
| 558 |
+
"learning_rate": 1.1073529411764704e-05,
|
| 559 |
+
"loss": 0.7222,
|
| 560 |
+
"step": 6800
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 43.59,
|
| 564 |
+
"eval_loss": 0.16627496480941772,
|
| 565 |
+
"eval_runtime": 67.2415,
|
| 566 |
+
"eval_samples_per_second": 31.945,
|
| 567 |
+
"eval_steps_per_second": 2.008,
|
| 568 |
+
"eval_wer": 0.17290953545232274,
|
| 569 |
+
"step": 6800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 44.23,
|
| 573 |
+
"learning_rate": 9.970588235294116e-06,
|
| 574 |
+
"loss": 0.709,
|
| 575 |
+
"step": 6900
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 44.87,
|
| 579 |
+
"learning_rate": 8.86764705882353e-06,
|
| 580 |
+
"loss": 0.7162,
|
| 581 |
+
"step": 7000
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 45.51,
|
| 585 |
+
"learning_rate": 7.764705882352941e-06,
|
| 586 |
+
"loss": 0.7282,
|
| 587 |
+
"step": 7100
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 46.15,
|
| 591 |
+
"learning_rate": 6.6838235294117634e-06,
|
| 592 |
+
"loss": 0.7156,
|
| 593 |
+
"step": 7200
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 46.15,
|
| 597 |
+
"eval_loss": 0.16330638527870178,
|
| 598 |
+
"eval_runtime": 67.815,
|
| 599 |
+
"eval_samples_per_second": 31.674,
|
| 600 |
+
"eval_steps_per_second": 1.991,
|
| 601 |
+
"eval_wer": 0.17154034229828852,
|
| 602 |
+
"step": 7200
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 46.79,
|
| 606 |
+
"learning_rate": 5.580882352941176e-06,
|
| 607 |
+
"loss": 0.7001,
|
| 608 |
+
"step": 7300
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 47.44,
|
| 612 |
+
"learning_rate": 4.477941176470588e-06,
|
| 613 |
+
"loss": 0.7034,
|
| 614 |
+
"step": 7400
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 48.08,
|
| 618 |
+
"learning_rate": 3.3749999999999995e-06,
|
| 619 |
+
"loss": 0.6991,
|
| 620 |
+
"step": 7500
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 48.72,
|
| 624 |
+
"learning_rate": 2.2720588235294113e-06,
|
| 625 |
+
"loss": 0.7121,
|
| 626 |
+
"step": 7600
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 48.72,
|
| 630 |
+
"eval_loss": 0.16656027734279633,
|
| 631 |
+
"eval_runtime": 66.7215,
|
| 632 |
+
"eval_samples_per_second": 32.194,
|
| 633 |
+
"eval_steps_per_second": 2.023,
|
| 634 |
+
"eval_wer": 0.17183374083129585,
|
| 635 |
+
"step": 7600
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 49.36,
|
| 639 |
+
"learning_rate": 1.1691176470588234e-06,
|
| 640 |
+
"loss": 0.7012,
|
| 641 |
+
"step": 7700
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 50.0,
|
| 645 |
+
"learning_rate": 6.617647058823529e-08,
|
| 646 |
+
"loss": 0.6972,
|
| 647 |
+
"step": 7800
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 50.0,
|
| 651 |
+
"step": 7800,
|
| 652 |
+
"total_flos": 2.5302348778981876e+19,
|
| 653 |
+
"train_loss": 1.4054935827010717,
|
| 654 |
+
"train_runtime": 11681.5311,
|
| 655 |
+
"train_samples_per_second": 21.243,
|
| 656 |
+
"train_steps_per_second": 0.668
|
| 657 |
+
}
|
| 658 |
+
],
|
| 659 |
+
"max_steps": 7800,
|
| 660 |
+
"num_train_epochs": 50,
|
| 661 |
+
"total_flos": 2.5302348778981876e+19,
|
| 662 |
+
"trial_name": null,
|
| 663 |
+
"trial_params": null
|
| 664 |
+
}
|