End of training
Browse files- all_results.json +7 -0
- train_results.json +7 -0
- trainer_state.json +685 -0
all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.09,
|
| 3 |
+
"train_loss": 0.0831078713297844,
|
| 4 |
+
"train_runtime": 247222.6725,
|
| 5 |
+
"train_samples_per_second": 7.766,
|
| 6 |
+
"train_steps_per_second": 0.081
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.09,
|
| 3 |
+
"train_loss": 0.0831078713297844,
|
| 4 |
+
"train_runtime": 247222.6725,
|
| 5 |
+
"train_samples_per_second": 7.766,
|
| 6 |
+
"train_steps_per_second": 0.081
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,685 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 7.411630558722919,
|
| 3 |
+
"best_model_checkpoint": "../whisper-NST2-unfreeze-constanti-low-lr/checkpoint-14000",
|
| 4 |
+
"epoch": 3.09385,
|
| 5 |
+
"global_step": 20000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.01,
|
| 12 |
+
"learning_rate": 2.47e-06,
|
| 13 |
+
"loss": 1.7243,
|
| 14 |
+
"step": 250
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.03,
|
| 18 |
+
"learning_rate": 4.970000000000001e-06,
|
| 19 |
+
"loss": 0.3597,
|
| 20 |
+
"step": 500
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.04,
|
| 24 |
+
"learning_rate": 7.4700000000000005e-06,
|
| 25 |
+
"loss": 0.2237,
|
| 26 |
+
"step": 750
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.05,
|
| 30 |
+
"learning_rate": 9.970000000000001e-06,
|
| 31 |
+
"loss": 0.1901,
|
| 32 |
+
"step": 1000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.05,
|
| 36 |
+
"eval_loss": 0.30687370896339417,
|
| 37 |
+
"eval_runtime": 48.6375,
|
| 38 |
+
"eval_samples_per_second": 2.056,
|
| 39 |
+
"eval_steps_per_second": 0.144,
|
| 40 |
+
"eval_wer": 14.823261117445838,
|
| 41 |
+
"step": 1000
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.06,
|
| 45 |
+
"learning_rate": 1e-05,
|
| 46 |
+
"loss": 0.1714,
|
| 47 |
+
"step": 1250
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 0.07,
|
| 51 |
+
"learning_rate": 1e-05,
|
| 52 |
+
"loss": 0.1581,
|
| 53 |
+
"step": 1500
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.09,
|
| 57 |
+
"learning_rate": 1e-05,
|
| 58 |
+
"loss": 0.1451,
|
| 59 |
+
"step": 1750
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1,
|
| 63 |
+
"learning_rate": 1e-05,
|
| 64 |
+
"loss": 0.1323,
|
| 65 |
+
"step": 2000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1,
|
| 69 |
+
"eval_loss": 0.2687492072582245,
|
| 70 |
+
"eval_runtime": 59.4025,
|
| 71 |
+
"eval_samples_per_second": 1.683,
|
| 72 |
+
"eval_steps_per_second": 0.118,
|
| 73 |
+
"eval_wer": 11.2884834663626,
|
| 74 |
+
"step": 2000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.11,
|
| 78 |
+
"learning_rate": 1e-05,
|
| 79 |
+
"loss": 0.1283,
|
| 80 |
+
"step": 2250
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.12,
|
| 84 |
+
"learning_rate": 1e-05,
|
| 85 |
+
"loss": 0.1227,
|
| 86 |
+
"step": 2500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.14,
|
| 90 |
+
"learning_rate": 1e-05,
|
| 91 |
+
"loss": 0.1172,
|
| 92 |
+
"step": 2750
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.15,
|
| 96 |
+
"learning_rate": 1e-05,
|
| 97 |
+
"loss": 0.1137,
|
| 98 |
+
"step": 3000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.15,
|
| 102 |
+
"eval_loss": 0.26204466819763184,
|
| 103 |
+
"eval_runtime": 53.0968,
|
| 104 |
+
"eval_samples_per_second": 1.883,
|
| 105 |
+
"eval_steps_per_second": 0.132,
|
| 106 |
+
"eval_wer": 10.832383124287343,
|
| 107 |
+
"step": 3000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.16,
|
| 111 |
+
"learning_rate": 1e-05,
|
| 112 |
+
"loss": 0.1104,
|
| 113 |
+
"step": 3250
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.17,
|
| 117 |
+
"learning_rate": 1e-05,
|
| 118 |
+
"loss": 0.1071,
|
| 119 |
+
"step": 3500
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.19,
|
| 123 |
+
"learning_rate": 1e-05,
|
| 124 |
+
"loss": 0.1022,
|
| 125 |
+
"step": 3750
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.2,
|
| 129 |
+
"learning_rate": 1e-05,
|
| 130 |
+
"loss": 0.1022,
|
| 131 |
+
"step": 4000
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.2,
|
| 135 |
+
"eval_loss": 0.297575443983078,
|
| 136 |
+
"eval_runtime": 45.3592,
|
| 137 |
+
"eval_samples_per_second": 2.205,
|
| 138 |
+
"eval_steps_per_second": 0.154,
|
| 139 |
+
"eval_wer": 9.007981755986318,
|
| 140 |
+
"step": 4000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.21,
|
| 144 |
+
"learning_rate": 1e-05,
|
| 145 |
+
"loss": 0.096,
|
| 146 |
+
"step": 4250
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.23,
|
| 150 |
+
"learning_rate": 1e-05,
|
| 151 |
+
"loss": 0.0939,
|
| 152 |
+
"step": 4500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.24,
|
| 156 |
+
"learning_rate": 1e-05,
|
| 157 |
+
"loss": 0.0918,
|
| 158 |
+
"step": 4750
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.25,
|
| 162 |
+
"learning_rate": 1e-05,
|
| 163 |
+
"loss": 0.0937,
|
| 164 |
+
"step": 5000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.25,
|
| 168 |
+
"eval_loss": 0.25842034816741943,
|
| 169 |
+
"eval_runtime": 48.6209,
|
| 170 |
+
"eval_samples_per_second": 2.057,
|
| 171 |
+
"eval_steps_per_second": 0.144,
|
| 172 |
+
"eval_wer": 9.578107183580387,
|
| 173 |
+
"step": 5000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.26,
|
| 177 |
+
"learning_rate": 1e-05,
|
| 178 |
+
"loss": 0.0895,
|
| 179 |
+
"step": 5250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.28,
|
| 183 |
+
"learning_rate": 1e-05,
|
| 184 |
+
"loss": 0.0874,
|
| 185 |
+
"step": 5500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.29,
|
| 189 |
+
"learning_rate": 1e-05,
|
| 190 |
+
"loss": 0.0875,
|
| 191 |
+
"step": 5750
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3,
|
| 195 |
+
"learning_rate": 1e-05,
|
| 196 |
+
"loss": 0.0875,
|
| 197 |
+
"step": 6000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.3,
|
| 201 |
+
"eval_loss": 0.27036869525909424,
|
| 202 |
+
"eval_runtime": 76.4925,
|
| 203 |
+
"eval_samples_per_second": 1.307,
|
| 204 |
+
"eval_steps_per_second": 0.092,
|
| 205 |
+
"eval_wer": 20.296465222348917,
|
| 206 |
+
"step": 6000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.01,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 0.0816,
|
| 212 |
+
"step": 6250
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.02,
|
| 216 |
+
"learning_rate": 1e-05,
|
| 217 |
+
"loss": 0.0748,
|
| 218 |
+
"step": 6500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.04,
|
| 222 |
+
"learning_rate": 1e-05,
|
| 223 |
+
"loss": 0.0656,
|
| 224 |
+
"step": 6750
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.05,
|
| 228 |
+
"learning_rate": 1e-05,
|
| 229 |
+
"loss": 0.0592,
|
| 230 |
+
"step": 7000
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.05,
|
| 234 |
+
"eval_loss": 0.2750946283340454,
|
| 235 |
+
"eval_runtime": 45.0139,
|
| 236 |
+
"eval_samples_per_second": 2.222,
|
| 237 |
+
"eval_steps_per_second": 0.156,
|
| 238 |
+
"eval_wer": 9.007981755986318,
|
| 239 |
+
"step": 7000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.06,
|
| 243 |
+
"learning_rate": 1e-05,
|
| 244 |
+
"loss": 0.0553,
|
| 245 |
+
"step": 7250
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.07,
|
| 249 |
+
"learning_rate": 1e-05,
|
| 250 |
+
"loss": 0.0542,
|
| 251 |
+
"step": 7500
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.09,
|
| 255 |
+
"learning_rate": 1e-05,
|
| 256 |
+
"loss": 0.0533,
|
| 257 |
+
"step": 7750
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.1,
|
| 261 |
+
"learning_rate": 1e-05,
|
| 262 |
+
"loss": 0.0488,
|
| 263 |
+
"step": 8000
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.1,
|
| 267 |
+
"eval_loss": 0.27778884768486023,
|
| 268 |
+
"eval_runtime": 49.8405,
|
| 269 |
+
"eval_samples_per_second": 2.006,
|
| 270 |
+
"eval_steps_per_second": 0.14,
|
| 271 |
+
"eval_wer": 8.665906499429875,
|
| 272 |
+
"step": 8000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.11,
|
| 276 |
+
"learning_rate": 1e-05,
|
| 277 |
+
"loss": 0.0481,
|
| 278 |
+
"step": 8250
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.12,
|
| 282 |
+
"learning_rate": 1e-05,
|
| 283 |
+
"loss": 0.0477,
|
| 284 |
+
"step": 8500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.14,
|
| 288 |
+
"learning_rate": 1e-05,
|
| 289 |
+
"loss": 0.047,
|
| 290 |
+
"step": 8750
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.15,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 0.0475,
|
| 296 |
+
"step": 9000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.15,
|
| 300 |
+
"eval_loss": 0.2791878581047058,
|
| 301 |
+
"eval_runtime": 48.8356,
|
| 302 |
+
"eval_samples_per_second": 2.048,
|
| 303 |
+
"eval_steps_per_second": 0.143,
|
| 304 |
+
"eval_wer": 9.464082098061574,
|
| 305 |
+
"step": 9000
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.16,
|
| 309 |
+
"learning_rate": 1e-05,
|
| 310 |
+
"loss": 0.045,
|
| 311 |
+
"step": 9250
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.17,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 0.0464,
|
| 317 |
+
"step": 9500
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.19,
|
| 321 |
+
"learning_rate": 1e-05,
|
| 322 |
+
"loss": 0.0419,
|
| 323 |
+
"step": 9750
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 1.2,
|
| 327 |
+
"learning_rate": 1e-05,
|
| 328 |
+
"loss": 0.0439,
|
| 329 |
+
"step": 10000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.2,
|
| 333 |
+
"eval_loss": 0.2879847586154938,
|
| 334 |
+
"eval_runtime": 49.0674,
|
| 335 |
+
"eval_samples_per_second": 2.038,
|
| 336 |
+
"eval_steps_per_second": 0.143,
|
| 337 |
+
"eval_wer": 8.323831242873432,
|
| 338 |
+
"step": 10000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.21,
|
| 342 |
+
"learning_rate": 1e-05,
|
| 343 |
+
"loss": 0.0413,
|
| 344 |
+
"step": 10250
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.22,
|
| 348 |
+
"learning_rate": 1e-05,
|
| 349 |
+
"loss": 0.041,
|
| 350 |
+
"step": 10500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.24,
|
| 354 |
+
"learning_rate": 1e-05,
|
| 355 |
+
"loss": 0.0413,
|
| 356 |
+
"step": 10750
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.25,
|
| 360 |
+
"learning_rate": 1e-05,
|
| 361 |
+
"loss": 0.0425,
|
| 362 |
+
"step": 11000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.25,
|
| 366 |
+
"eval_loss": 0.2954462766647339,
|
| 367 |
+
"eval_runtime": 46.9453,
|
| 368 |
+
"eval_samples_per_second": 2.13,
|
| 369 |
+
"eval_steps_per_second": 0.149,
|
| 370 |
+
"eval_wer": 8.551881413911062,
|
| 371 |
+
"step": 11000
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 1.26,
|
| 375 |
+
"learning_rate": 1e-05,
|
| 376 |
+
"loss": 0.0409,
|
| 377 |
+
"step": 11250
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.27,
|
| 381 |
+
"learning_rate": 1e-05,
|
| 382 |
+
"loss": 0.0389,
|
| 383 |
+
"step": 11500
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.29,
|
| 387 |
+
"learning_rate": 1e-05,
|
| 388 |
+
"loss": 0.0403,
|
| 389 |
+
"step": 11750
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.3,
|
| 393 |
+
"learning_rate": 1e-05,
|
| 394 |
+
"loss": 0.0416,
|
| 395 |
+
"step": 12000
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.3,
|
| 399 |
+
"eval_loss": 0.2895849347114563,
|
| 400 |
+
"eval_runtime": 72.8965,
|
| 401 |
+
"eval_samples_per_second": 1.372,
|
| 402 |
+
"eval_steps_per_second": 0.096,
|
| 403 |
+
"eval_wer": 20.296465222348917,
|
| 404 |
+
"step": 12000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.01,
|
| 408 |
+
"learning_rate": 1e-05,
|
| 409 |
+
"loss": 0.0384,
|
| 410 |
+
"step": 12250
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.02,
|
| 414 |
+
"learning_rate": 1e-05,
|
| 415 |
+
"loss": 0.0355,
|
| 416 |
+
"step": 12500
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.03,
|
| 420 |
+
"learning_rate": 1e-05,
|
| 421 |
+
"loss": 0.0323,
|
| 422 |
+
"step": 12750
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.05,
|
| 426 |
+
"learning_rate": 1e-05,
|
| 427 |
+
"loss": 0.0289,
|
| 428 |
+
"step": 13000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.05,
|
| 432 |
+
"eval_loss": 0.29897502064704895,
|
| 433 |
+
"eval_runtime": 54.9916,
|
| 434 |
+
"eval_samples_per_second": 1.818,
|
| 435 |
+
"eval_steps_per_second": 0.127,
|
| 436 |
+
"eval_wer": 7.98175598631699,
|
| 437 |
+
"step": 13000
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.06,
|
| 441 |
+
"learning_rate": 1e-05,
|
| 442 |
+
"loss": 0.0266,
|
| 443 |
+
"step": 13250
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 2.07,
|
| 447 |
+
"learning_rate": 1e-05,
|
| 448 |
+
"loss": 0.0264,
|
| 449 |
+
"step": 13500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 2.08,
|
| 453 |
+
"learning_rate": 1e-05,
|
| 454 |
+
"loss": 0.0267,
|
| 455 |
+
"step": 13750
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 2.1,
|
| 459 |
+
"learning_rate": 1e-05,
|
| 460 |
+
"loss": 0.0229,
|
| 461 |
+
"step": 14000
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 2.1,
|
| 465 |
+
"eval_loss": 0.3026929497718811,
|
| 466 |
+
"eval_runtime": 48.2743,
|
| 467 |
+
"eval_samples_per_second": 2.071,
|
| 468 |
+
"eval_steps_per_second": 0.145,
|
| 469 |
+
"eval_wer": 7.411630558722919,
|
| 470 |
+
"step": 14000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.11,
|
| 474 |
+
"learning_rate": 1e-05,
|
| 475 |
+
"loss": 0.0228,
|
| 476 |
+
"step": 14250
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.12,
|
| 480 |
+
"learning_rate": 1e-05,
|
| 481 |
+
"loss": 0.024,
|
| 482 |
+
"step": 14500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.13,
|
| 486 |
+
"learning_rate": 1e-05,
|
| 487 |
+
"loss": 0.0238,
|
| 488 |
+
"step": 14750
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.15,
|
| 492 |
+
"learning_rate": 1e-05,
|
| 493 |
+
"loss": 0.0248,
|
| 494 |
+
"step": 15000
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.15,
|
| 498 |
+
"eval_loss": 0.29682403802871704,
|
| 499 |
+
"eval_runtime": 47.3876,
|
| 500 |
+
"eval_samples_per_second": 2.11,
|
| 501 |
+
"eval_steps_per_second": 0.148,
|
| 502 |
+
"eval_wer": 8.665906499429875,
|
| 503 |
+
"step": 15000
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.16,
|
| 507 |
+
"learning_rate": 1e-05,
|
| 508 |
+
"loss": 0.022,
|
| 509 |
+
"step": 15250
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 2.17,
|
| 513 |
+
"learning_rate": 1e-05,
|
| 514 |
+
"loss": 0.0241,
|
| 515 |
+
"step": 15500
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 2.18,
|
| 519 |
+
"learning_rate": 1e-05,
|
| 520 |
+
"loss": 0.021,
|
| 521 |
+
"step": 15750
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 2.2,
|
| 525 |
+
"learning_rate": 1e-05,
|
| 526 |
+
"loss": 0.0225,
|
| 527 |
+
"step": 16000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 2.2,
|
| 531 |
+
"eval_loss": 0.30999651551246643,
|
| 532 |
+
"eval_runtime": 56.9907,
|
| 533 |
+
"eval_samples_per_second": 1.755,
|
| 534 |
+
"eval_steps_per_second": 0.123,
|
| 535 |
+
"eval_wer": 8.551881413911062,
|
| 536 |
+
"step": 16000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.21,
|
| 540 |
+
"learning_rate": 1e-05,
|
| 541 |
+
"loss": 0.0211,
|
| 542 |
+
"step": 16250
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.22,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 0.0214,
|
| 548 |
+
"step": 16500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.23,
|
| 552 |
+
"learning_rate": 1e-05,
|
| 553 |
+
"loss": 0.0215,
|
| 554 |
+
"step": 16750
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.25,
|
| 558 |
+
"learning_rate": 1e-05,
|
| 559 |
+
"loss": 0.0222,
|
| 560 |
+
"step": 17000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.25,
|
| 564 |
+
"eval_loss": 0.3132005035877228,
|
| 565 |
+
"eval_runtime": 54.3277,
|
| 566 |
+
"eval_samples_per_second": 1.841,
|
| 567 |
+
"eval_steps_per_second": 0.129,
|
| 568 |
+
"eval_wer": 9.35005701254276,
|
| 569 |
+
"step": 17000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 2.26,
|
| 573 |
+
"learning_rate": 1e-05,
|
| 574 |
+
"loss": 0.0211,
|
| 575 |
+
"step": 17250
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 2.27,
|
| 579 |
+
"learning_rate": 1e-05,
|
| 580 |
+
"loss": 0.0199,
|
| 581 |
+
"step": 17500
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 2.28,
|
| 585 |
+
"learning_rate": 1e-05,
|
| 586 |
+
"loss": 0.021,
|
| 587 |
+
"step": 17750
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.3,
|
| 591 |
+
"learning_rate": 1e-05,
|
| 592 |
+
"loss": 0.0219,
|
| 593 |
+
"step": 18000
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 2.3,
|
| 597 |
+
"eval_loss": 0.3229832053184509,
|
| 598 |
+
"eval_runtime": 56.2842,
|
| 599 |
+
"eval_samples_per_second": 1.777,
|
| 600 |
+
"eval_steps_per_second": 0.124,
|
| 601 |
+
"eval_wer": 7.639680729760548,
|
| 602 |
+
"step": 18000
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 3.01,
|
| 606 |
+
"learning_rate": 1e-05,
|
| 607 |
+
"loss": 0.0205,
|
| 608 |
+
"step": 18250
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 3.02,
|
| 612 |
+
"learning_rate": 1e-05,
|
| 613 |
+
"loss": 0.0194,
|
| 614 |
+
"step": 18500
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.03,
|
| 618 |
+
"learning_rate": 1e-05,
|
| 619 |
+
"loss": 0.0178,
|
| 620 |
+
"step": 18750
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 3.04,
|
| 624 |
+
"learning_rate": 1e-05,
|
| 625 |
+
"loss": 0.0162,
|
| 626 |
+
"step": 19000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 3.04,
|
| 630 |
+
"eval_loss": 0.33802878856658936,
|
| 631 |
+
"eval_runtime": 56.712,
|
| 632 |
+
"eval_samples_per_second": 1.763,
|
| 633 |
+
"eval_steps_per_second": 0.123,
|
| 634 |
+
"eval_wer": 9.806157354618016,
|
| 635 |
+
"step": 19000
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 3.06,
|
| 639 |
+
"learning_rate": 1e-05,
|
| 640 |
+
"loss": 0.0144,
|
| 641 |
+
"step": 19250
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 3.07,
|
| 645 |
+
"learning_rate": 1e-05,
|
| 646 |
+
"loss": 0.0151,
|
| 647 |
+
"step": 19500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 3.08,
|
| 651 |
+
"learning_rate": 1e-05,
|
| 652 |
+
"loss": 0.0151,
|
| 653 |
+
"step": 19750
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 3.09,
|
| 657 |
+
"learning_rate": 1e-05,
|
| 658 |
+
"loss": 0.0132,
|
| 659 |
+
"step": 20000
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 3.09,
|
| 663 |
+
"eval_loss": 0.35617145895957947,
|
| 664 |
+
"eval_runtime": 42.4011,
|
| 665 |
+
"eval_samples_per_second": 2.358,
|
| 666 |
+
"eval_steps_per_second": 0.165,
|
| 667 |
+
"eval_wer": 8.551881413911062,
|
| 668 |
+
"step": 20000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 3.09,
|
| 672 |
+
"step": 20000,
|
| 673 |
+
"total_flos": 2.7701254859636736e+20,
|
| 674 |
+
"train_loss": 0.0831078713297844,
|
| 675 |
+
"train_runtime": 247222.6725,
|
| 676 |
+
"train_samples_per_second": 7.766,
|
| 677 |
+
"train_steps_per_second": 0.081
|
| 678 |
+
}
|
| 679 |
+
],
|
| 680 |
+
"max_steps": 20000,
|
| 681 |
+
"num_train_epochs": 9223372036854775807,
|
| 682 |
+
"total_flos": 2.7701254859636736e+20,
|
| 683 |
+
"trial_name": null,
|
| 684 |
+
"trial_params": null
|
| 685 |
+
}
|