Commit
•
1e53f84
1
Parent(s):
cae6e34
End of training
Browse files- all_results.json +14 -0
- eval_results.json +9 -0
- train_results.json +8 -0
- trainer_state.json +673 -0
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"eval_loss": 0.5773508548736572,
|
| 4 |
+
"eval_runtime": 1.9298,
|
| 5 |
+
"eval_samples": 100,
|
| 6 |
+
"eval_samples_per_second": 51.819,
|
| 7 |
+
"eval_steps_per_second": 6.736,
|
| 8 |
+
"eval_wer": 0.5078809106830122,
|
| 9 |
+
"train_loss": 0.5518865008413055,
|
| 10 |
+
"train_runtime": 5863.5283,
|
| 11 |
+
"train_samples": 13811,
|
| 12 |
+
"train_samples_per_second": 35.331,
|
| 13 |
+
"train_steps_per_second": 1.105
|
| 14 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"eval_loss": 0.5773508548736572,
|
| 4 |
+
"eval_runtime": 1.9298,
|
| 5 |
+
"eval_samples": 100,
|
| 6 |
+
"eval_samples_per_second": 51.819,
|
| 7 |
+
"eval_steps_per_second": 6.736,
|
| 8 |
+
"eval_wer": 0.5078809106830122
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 15.0,
|
| 3 |
+
"train_loss": 0.5518865008413055,
|
| 4 |
+
"train_runtime": 5863.5283,
|
| 5 |
+
"train_samples": 13811,
|
| 6 |
+
"train_samples_per_second": 35.331,
|
| 7 |
+
"train_steps_per_second": 1.105
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,673 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 15.0,
|
| 5 |
+
"global_step": 6480,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.23,
|
| 12 |
+
"eval_loss": 3.359571933746338,
|
| 13 |
+
"eval_runtime": 1.8892,
|
| 14 |
+
"eval_samples_per_second": 52.934,
|
| 15 |
+
"eval_steps_per_second": 6.881,
|
| 16 |
+
"eval_wer": 1.0,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.46,
|
| 21 |
+
"eval_loss": 2.9279825687408447,
|
| 22 |
+
"eval_runtime": 1.8734,
|
| 23 |
+
"eval_samples_per_second": 53.378,
|
| 24 |
+
"eval_steps_per_second": 6.939,
|
| 25 |
+
"eval_wer": 1.0,
|
| 26 |
+
"step": 200
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.69,
|
| 30 |
+
"eval_loss": 1.5090594291687012,
|
| 31 |
+
"eval_runtime": 1.876,
|
| 32 |
+
"eval_samples_per_second": 53.306,
|
| 33 |
+
"eval_steps_per_second": 6.93,
|
| 34 |
+
"eval_wer": 0.9649737302977233,
|
| 35 |
+
"step": 300
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.93,
|
| 39 |
+
"eval_loss": 0.994255006313324,
|
| 40 |
+
"eval_runtime": 1.8636,
|
| 41 |
+
"eval_samples_per_second": 53.659,
|
| 42 |
+
"eval_steps_per_second": 6.976,
|
| 43 |
+
"eval_wer": 0.9176882661996497,
|
| 44 |
+
"step": 400
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.16,
|
| 48 |
+
"learning_rate": 0.0002982,
|
| 49 |
+
"loss": 3.1184,
|
| 50 |
+
"step": 500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 1.16,
|
| 54 |
+
"eval_loss": 0.7590276598930359,
|
| 55 |
+
"eval_runtime": 1.9271,
|
| 56 |
+
"eval_samples_per_second": 51.893,
|
| 57 |
+
"eval_steps_per_second": 6.746,
|
| 58 |
+
"eval_wer": 0.7793345008756567,
|
| 59 |
+
"step": 500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 1.39,
|
| 63 |
+
"eval_loss": 0.7336117625236511,
|
| 64 |
+
"eval_runtime": 1.8782,
|
| 65 |
+
"eval_samples_per_second": 53.242,
|
| 66 |
+
"eval_steps_per_second": 6.921,
|
| 67 |
+
"eval_wer": 0.7408056042031523,
|
| 68 |
+
"step": 600
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 1.62,
|
| 72 |
+
"eval_loss": 0.7039847373962402,
|
| 73 |
+
"eval_runtime": 1.8772,
|
| 74 |
+
"eval_samples_per_second": 53.27,
|
| 75 |
+
"eval_steps_per_second": 6.925,
|
| 76 |
+
"eval_wer": 0.7618213660245184,
|
| 77 |
+
"step": 700
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 1.85,
|
| 81 |
+
"eval_loss": 0.6815319061279297,
|
| 82 |
+
"eval_runtime": 1.8953,
|
| 83 |
+
"eval_samples_per_second": 52.762,
|
| 84 |
+
"eval_steps_per_second": 6.859,
|
| 85 |
+
"eval_wer": 0.723292469352014,
|
| 86 |
+
"step": 800
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.08,
|
| 90 |
+
"eval_loss": 0.6456977128982544,
|
| 91 |
+
"eval_runtime": 1.8847,
|
| 92 |
+
"eval_samples_per_second": 53.058,
|
| 93 |
+
"eval_steps_per_second": 6.898,
|
| 94 |
+
"eval_wer": 0.6865148861646234,
|
| 95 |
+
"step": 900
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.31,
|
| 99 |
+
"learning_rate": 0.00027511705685618725,
|
| 100 |
+
"loss": 0.7917,
|
| 101 |
+
"step": 1000
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 2.31,
|
| 105 |
+
"eval_loss": 0.5704939365386963,
|
| 106 |
+
"eval_runtime": 1.8802,
|
| 107 |
+
"eval_samples_per_second": 53.187,
|
| 108 |
+
"eval_steps_per_second": 6.914,
|
| 109 |
+
"eval_wer": 0.681260945709282,
|
| 110 |
+
"step": 1000
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 2.55,
|
| 114 |
+
"eval_loss": 0.5708367824554443,
|
| 115 |
+
"eval_runtime": 1.8769,
|
| 116 |
+
"eval_samples_per_second": 53.28,
|
| 117 |
+
"eval_steps_per_second": 6.926,
|
| 118 |
+
"eval_wer": 0.6619964973730298,
|
| 119 |
+
"step": 1100
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 2.78,
|
| 123 |
+
"eval_loss": 0.5888168811798096,
|
| 124 |
+
"eval_runtime": 1.8801,
|
| 125 |
+
"eval_samples_per_second": 53.188,
|
| 126 |
+
"eval_steps_per_second": 6.914,
|
| 127 |
+
"eval_wer": 0.6462346760070052,
|
| 128 |
+
"step": 1200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.01,
|
| 132 |
+
"eval_loss": 0.6509290337562561,
|
| 133 |
+
"eval_runtime": 1.8971,
|
| 134 |
+
"eval_samples_per_second": 52.711,
|
| 135 |
+
"eval_steps_per_second": 6.852,
|
| 136 |
+
"eval_wer": 0.6970227670753065,
|
| 137 |
+
"step": 1300
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 3.24,
|
| 141 |
+
"eval_loss": 0.5871497392654419,
|
| 142 |
+
"eval_runtime": 1.8883,
|
| 143 |
+
"eval_samples_per_second": 52.957,
|
| 144 |
+
"eval_steps_per_second": 6.884,
|
| 145 |
+
"eval_wer": 0.6462346760070052,
|
| 146 |
+
"step": 1400
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.47,
|
| 150 |
+
"learning_rate": 0.00025008361204013375,
|
| 151 |
+
"loss": 0.5909,
|
| 152 |
+
"step": 1500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 3.47,
|
| 156 |
+
"eval_loss": 0.619912326335907,
|
| 157 |
+
"eval_runtime": 1.8867,
|
| 158 |
+
"eval_samples_per_second": 53.002,
|
| 159 |
+
"eval_steps_per_second": 6.89,
|
| 160 |
+
"eval_wer": 0.681260945709282,
|
| 161 |
+
"step": 1500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 3.7,
|
| 165 |
+
"eval_loss": 0.6230051517486572,
|
| 166 |
+
"eval_runtime": 1.8866,
|
| 167 |
+
"eval_samples_per_second": 53.006,
|
| 168 |
+
"eval_steps_per_second": 6.891,
|
| 169 |
+
"eval_wer": 0.5919439579684763,
|
| 170 |
+
"step": 1600
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.94,
|
| 174 |
+
"eval_loss": 0.572126030921936,
|
| 175 |
+
"eval_runtime": 1.896,
|
| 176 |
+
"eval_samples_per_second": 52.741,
|
| 177 |
+
"eval_steps_per_second": 6.856,
|
| 178 |
+
"eval_wer": 0.6427320490367776,
|
| 179 |
+
"step": 1700
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 4.17,
|
| 183 |
+
"eval_loss": 0.5330815315246582,
|
| 184 |
+
"eval_runtime": 1.8961,
|
| 185 |
+
"eval_samples_per_second": 52.741,
|
| 186 |
+
"eval_steps_per_second": 6.856,
|
| 187 |
+
"eval_wer": 0.5866900175131349,
|
| 188 |
+
"step": 1800
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 4.4,
|
| 192 |
+
"eval_loss": 0.5560976266860962,
|
| 193 |
+
"eval_runtime": 1.9173,
|
| 194 |
+
"eval_samples_per_second": 52.157,
|
| 195 |
+
"eval_steps_per_second": 6.78,
|
| 196 |
+
"eval_wer": 0.6007005253940455,
|
| 197 |
+
"step": 1900
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 4.63,
|
| 201 |
+
"learning_rate": 0.00022510033444816051,
|
| 202 |
+
"loss": 0.4607,
|
| 203 |
+
"step": 2000
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 4.63,
|
| 207 |
+
"eval_loss": 0.5414324402809143,
|
| 208 |
+
"eval_runtime": 1.8832,
|
| 209 |
+
"eval_samples_per_second": 53.102,
|
| 210 |
+
"eval_steps_per_second": 6.903,
|
| 211 |
+
"eval_wer": 0.5849387040280211,
|
| 212 |
+
"step": 2000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 4.86,
|
| 216 |
+
"eval_loss": 0.5389540791511536,
|
| 217 |
+
"eval_runtime": 1.8801,
|
| 218 |
+
"eval_samples_per_second": 53.19,
|
| 219 |
+
"eval_steps_per_second": 6.915,
|
| 220 |
+
"eval_wer": 0.5586690017513135,
|
| 221 |
+
"step": 2100
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 5.09,
|
| 225 |
+
"eval_loss": 0.5313354730606079,
|
| 226 |
+
"eval_runtime": 1.8925,
|
| 227 |
+
"eval_samples_per_second": 52.842,
|
| 228 |
+
"eval_steps_per_second": 6.869,
|
| 229 |
+
"eval_wer": 0.5569176882661997,
|
| 230 |
+
"step": 2200
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 5.32,
|
| 234 |
+
"eval_loss": 0.5893406271934509,
|
| 235 |
+
"eval_runtime": 1.8839,
|
| 236 |
+
"eval_samples_per_second": 53.083,
|
| 237 |
+
"eval_steps_per_second": 6.901,
|
| 238 |
+
"eval_wer": 0.5796847635726795,
|
| 239 |
+
"step": 2300
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 5.56,
|
| 243 |
+
"eval_loss": 0.550654411315918,
|
| 244 |
+
"eval_runtime": 1.8741,
|
| 245 |
+
"eval_samples_per_second": 53.359,
|
| 246 |
+
"eval_steps_per_second": 6.937,
|
| 247 |
+
"eval_wer": 0.5954465849387041,
|
| 248 |
+
"step": 2400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.79,
|
| 252 |
+
"learning_rate": 0.00020006688963210701,
|
| 253 |
+
"loss": 0.3933,
|
| 254 |
+
"step": 2500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 5.79,
|
| 258 |
+
"eval_loss": 0.552050769329071,
|
| 259 |
+
"eval_runtime": 1.8818,
|
| 260 |
+
"eval_samples_per_second": 53.14,
|
| 261 |
+
"eval_steps_per_second": 6.908,
|
| 262 |
+
"eval_wer": 0.6024518388791593,
|
| 263 |
+
"step": 2500
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 6.02,
|
| 267 |
+
"eval_loss": 0.5662926435470581,
|
| 268 |
+
"eval_runtime": 1.9296,
|
| 269 |
+
"eval_samples_per_second": 51.824,
|
| 270 |
+
"eval_steps_per_second": 6.737,
|
| 271 |
+
"eval_wer": 0.5989492119089317,
|
| 272 |
+
"step": 2600
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 6.25,
|
| 276 |
+
"eval_loss": 0.5636402368545532,
|
| 277 |
+
"eval_runtime": 1.8763,
|
| 278 |
+
"eval_samples_per_second": 53.297,
|
| 279 |
+
"eval_steps_per_second": 6.929,
|
| 280 |
+
"eval_wer": 0.5831873905429071,
|
| 281 |
+
"step": 2700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 6.48,
|
| 285 |
+
"eval_loss": 0.5463794469833374,
|
| 286 |
+
"eval_runtime": 1.8753,
|
| 287 |
+
"eval_samples_per_second": 53.325,
|
| 288 |
+
"eval_steps_per_second": 6.932,
|
| 289 |
+
"eval_wer": 0.5919439579684763,
|
| 290 |
+
"step": 2800
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 6.71,
|
| 294 |
+
"eval_loss": 0.5623293519020081,
|
| 295 |
+
"eval_runtime": 1.8826,
|
| 296 |
+
"eval_samples_per_second": 53.118,
|
| 297 |
+
"eval_steps_per_second": 6.905,
|
| 298 |
+
"eval_wer": 0.5831873905429071,
|
| 299 |
+
"step": 2900
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 6.94,
|
| 303 |
+
"learning_rate": 0.00017503344481605351,
|
| 304 |
+
"loss": 0.3367,
|
| 305 |
+
"step": 3000
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 6.94,
|
| 309 |
+
"eval_loss": 0.5324317812919617,
|
| 310 |
+
"eval_runtime": 1.8849,
|
| 311 |
+
"eval_samples_per_second": 53.053,
|
| 312 |
+
"eval_steps_per_second": 6.897,
|
| 313 |
+
"eval_wer": 0.5691768826619965,
|
| 314 |
+
"step": 3000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 7.18,
|
| 318 |
+
"eval_loss": 0.5907294750213623,
|
| 319 |
+
"eval_runtime": 1.8884,
|
| 320 |
+
"eval_samples_per_second": 52.954,
|
| 321 |
+
"eval_steps_per_second": 6.884,
|
| 322 |
+
"eval_wer": 0.5394045534150613,
|
| 323 |
+
"step": 3100
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 7.41,
|
| 327 |
+
"eval_loss": 0.5653238892555237,
|
| 328 |
+
"eval_runtime": 1.8757,
|
| 329 |
+
"eval_samples_per_second": 53.314,
|
| 330 |
+
"eval_steps_per_second": 6.931,
|
| 331 |
+
"eval_wer": 0.5814360770577933,
|
| 332 |
+
"step": 3200
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 7.64,
|
| 336 |
+
"eval_loss": 0.5707294940948486,
|
| 337 |
+
"eval_runtime": 1.9074,
|
| 338 |
+
"eval_samples_per_second": 52.427,
|
| 339 |
+
"eval_steps_per_second": 6.816,
|
| 340 |
+
"eval_wer": 0.5814360770577933,
|
| 341 |
+
"step": 3300
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 7.87,
|
| 345 |
+
"eval_loss": 0.5753923654556274,
|
| 346 |
+
"eval_runtime": 1.8899,
|
| 347 |
+
"eval_samples_per_second": 52.912,
|
| 348 |
+
"eval_steps_per_second": 6.879,
|
| 349 |
+
"eval_wer": 0.542907180385289,
|
| 350 |
+
"step": 3400
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 8.1,
|
| 354 |
+
"learning_rate": 0.00015,
|
| 355 |
+
"loss": 0.2856,
|
| 356 |
+
"step": 3500
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 8.1,
|
| 360 |
+
"eval_loss": 0.5953063368797302,
|
| 361 |
+
"eval_runtime": 1.8933,
|
| 362 |
+
"eval_samples_per_second": 52.818,
|
| 363 |
+
"eval_steps_per_second": 6.866,
|
| 364 |
+
"eval_wer": 0.5569176882661997,
|
| 365 |
+
"step": 3500
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 8.33,
|
| 369 |
+
"eval_loss": 0.6274660229682922,
|
| 370 |
+
"eval_runtime": 1.8873,
|
| 371 |
+
"eval_samples_per_second": 52.985,
|
| 372 |
+
"eval_steps_per_second": 6.888,
|
| 373 |
+
"eval_wer": 0.5394045534150613,
|
| 374 |
+
"step": 3600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 8.56,
|
| 378 |
+
"eval_loss": 0.6253136992454529,
|
| 379 |
+
"eval_runtime": 1.8922,
|
| 380 |
+
"eval_samples_per_second": 52.847,
|
| 381 |
+
"eval_steps_per_second": 6.87,
|
| 382 |
+
"eval_wer": 0.5569176882661997,
|
| 383 |
+
"step": 3700
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 8.8,
|
| 387 |
+
"eval_loss": 0.5930343866348267,
|
| 388 |
+
"eval_runtime": 1.9165,
|
| 389 |
+
"eval_samples_per_second": 52.177,
|
| 390 |
+
"eval_steps_per_second": 6.783,
|
| 391 |
+
"eval_wer": 0.542907180385289,
|
| 392 |
+
"step": 3800
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 9.03,
|
| 396 |
+
"eval_loss": 0.6082107424736023,
|
| 397 |
+
"eval_runtime": 1.9067,
|
| 398 |
+
"eval_samples_per_second": 52.448,
|
| 399 |
+
"eval_steps_per_second": 6.818,
|
| 400 |
+
"eval_wer": 0.521891418563923,
|
| 401 |
+
"step": 3900
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 9.26,
|
| 405 |
+
"learning_rate": 0.00012496655518394646,
|
| 406 |
+
"loss": 0.2522,
|
| 407 |
+
"step": 4000
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 9.26,
|
| 411 |
+
"eval_loss": 0.6026180982589722,
|
| 412 |
+
"eval_runtime": 1.9243,
|
| 413 |
+
"eval_samples_per_second": 51.967,
|
| 414 |
+
"eval_steps_per_second": 6.756,
|
| 415 |
+
"eval_wer": 0.5446584938704028,
|
| 416 |
+
"step": 4000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 9.49,
|
| 420 |
+
"eval_loss": 0.6052154302597046,
|
| 421 |
+
"eval_runtime": 1.9113,
|
| 422 |
+
"eval_samples_per_second": 52.32,
|
| 423 |
+
"eval_steps_per_second": 6.802,
|
| 424 |
+
"eval_wer": 0.5271453590192644,
|
| 425 |
+
"step": 4100
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 9.72,
|
| 429 |
+
"eval_loss": 0.5870827436447144,
|
| 430 |
+
"eval_runtime": 1.8877,
|
| 431 |
+
"eval_samples_per_second": 52.973,
|
| 432 |
+
"eval_steps_per_second": 6.887,
|
| 433 |
+
"eval_wer": 0.521891418563923,
|
| 434 |
+
"step": 4200
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 9.95,
|
| 438 |
+
"eval_loss": 0.5870257019996643,
|
| 439 |
+
"eval_runtime": 1.8862,
|
| 440 |
+
"eval_samples_per_second": 53.017,
|
| 441 |
+
"eval_steps_per_second": 6.892,
|
| 442 |
+
"eval_wer": 0.5236427320490368,
|
| 443 |
+
"step": 4300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 10.19,
|
| 447 |
+
"eval_loss": 0.5880929231643677,
|
| 448 |
+
"eval_runtime": 1.9034,
|
| 449 |
+
"eval_samples_per_second": 52.538,
|
| 450 |
+
"eval_steps_per_second": 6.83,
|
| 451 |
+
"eval_wer": 0.5131348511383538,
|
| 452 |
+
"step": 4400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 10.42,
|
| 456 |
+
"learning_rate": 9.993311036789297e-05,
|
| 457 |
+
"loss": 0.2167,
|
| 458 |
+
"step": 4500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 10.42,
|
| 462 |
+
"eval_loss": 0.6122171878814697,
|
| 463 |
+
"eval_runtime": 1.8922,
|
| 464 |
+
"eval_samples_per_second": 52.847,
|
| 465 |
+
"eval_steps_per_second": 6.87,
|
| 466 |
+
"eval_wer": 0.5288966725043783,
|
| 467 |
+
"step": 4500
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 10.65,
|
| 471 |
+
"eval_loss": 0.612826406955719,
|
| 472 |
+
"eval_runtime": 1.891,
|
| 473 |
+
"eval_samples_per_second": 52.882,
|
| 474 |
+
"eval_steps_per_second": 6.875,
|
| 475 |
+
"eval_wer": 0.5166374781085814,
|
| 476 |
+
"step": 4600
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 10.88,
|
| 480 |
+
"eval_loss": 0.6134529113769531,
|
| 481 |
+
"eval_runtime": 1.891,
|
| 482 |
+
"eval_samples_per_second": 52.883,
|
| 483 |
+
"eval_steps_per_second": 6.875,
|
| 484 |
+
"eval_wer": 0.5376532399299475,
|
| 485 |
+
"step": 4700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 11.11,
|
| 489 |
+
"eval_loss": 0.6054602265357971,
|
| 490 |
+
"eval_runtime": 1.8917,
|
| 491 |
+
"eval_samples_per_second": 52.863,
|
| 492 |
+
"eval_steps_per_second": 6.872,
|
| 493 |
+
"eval_wer": 0.5183887915936952,
|
| 494 |
+
"step": 4800
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 11.34,
|
| 498 |
+
"eval_loss": 0.6724901795387268,
|
| 499 |
+
"eval_runtime": 1.8847,
|
| 500 |
+
"eval_samples_per_second": 53.058,
|
| 501 |
+
"eval_steps_per_second": 6.898,
|
| 502 |
+
"eval_wer": 0.5569176882661997,
|
| 503 |
+
"step": 4900
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 11.57,
|
| 507 |
+
"learning_rate": 7.489966555183946e-05,
|
| 508 |
+
"loss": 0.1965,
|
| 509 |
+
"step": 5000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 11.57,
|
| 513 |
+
"eval_loss": 0.6481964588165283,
|
| 514 |
+
"eval_runtime": 1.9092,
|
| 515 |
+
"eval_samples_per_second": 52.378,
|
| 516 |
+
"eval_steps_per_second": 6.809,
|
| 517 |
+
"eval_wer": 0.542907180385289,
|
| 518 |
+
"step": 5000
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 11.81,
|
| 522 |
+
"eval_loss": 0.6037153601646423,
|
| 523 |
+
"eval_runtime": 1.8885,
|
| 524 |
+
"eval_samples_per_second": 52.953,
|
| 525 |
+
"eval_steps_per_second": 6.884,
|
| 526 |
+
"eval_wer": 0.5096322241681261,
|
| 527 |
+
"step": 5100
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 12.04,
|
| 531 |
+
"eval_loss": 0.5931165218353271,
|
| 532 |
+
"eval_runtime": 1.9024,
|
| 533 |
+
"eval_samples_per_second": 52.566,
|
| 534 |
+
"eval_steps_per_second": 6.834,
|
| 535 |
+
"eval_wer": 0.5131348511383538,
|
| 536 |
+
"step": 5200
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 12.27,
|
| 540 |
+
"eval_loss": 0.5853209495544434,
|
| 541 |
+
"eval_runtime": 1.9149,
|
| 542 |
+
"eval_samples_per_second": 52.223,
|
| 543 |
+
"eval_steps_per_second": 6.789,
|
| 544 |
+
"eval_wer": 0.51138353765324,
|
| 545 |
+
"step": 5300
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 12.5,
|
| 549 |
+
"eval_loss": 0.5798044800758362,
|
| 550 |
+
"eval_runtime": 1.9229,
|
| 551 |
+
"eval_samples_per_second": 52.006,
|
| 552 |
+
"eval_steps_per_second": 6.761,
|
| 553 |
+
"eval_wer": 0.521891418563923,
|
| 554 |
+
"step": 5400
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 12.73,
|
| 558 |
+
"learning_rate": 4.991638795986622e-05,
|
| 559 |
+
"loss": 0.172,
|
| 560 |
+
"step": 5500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 12.73,
|
| 564 |
+
"eval_loss": 0.5774720907211304,
|
| 565 |
+
"eval_runtime": 1.8968,
|
| 566 |
+
"eval_samples_per_second": 52.72,
|
| 567 |
+
"eval_steps_per_second": 6.854,
|
| 568 |
+
"eval_wer": 0.500875656742557,
|
| 569 |
+
"step": 5500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 12.96,
|
| 573 |
+
"eval_loss": 0.5782347321510315,
|
| 574 |
+
"eval_runtime": 1.9216,
|
| 575 |
+
"eval_samples_per_second": 52.041,
|
| 576 |
+
"eval_steps_per_second": 6.765,
|
| 577 |
+
"eval_wer": 0.5043782837127846,
|
| 578 |
+
"step": 5600
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 13.19,
|
| 582 |
+
"eval_loss": 0.5803666114807129,
|
| 583 |
+
"eval_runtime": 1.9303,
|
| 584 |
+
"eval_samples_per_second": 51.806,
|
| 585 |
+
"eval_steps_per_second": 6.735,
|
| 586 |
+
"eval_wer": 0.5183887915936952,
|
| 587 |
+
"step": 5700
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 13.43,
|
| 591 |
+
"eval_loss": 0.5976961255073547,
|
| 592 |
+
"eval_runtime": 1.8886,
|
| 593 |
+
"eval_samples_per_second": 52.948,
|
| 594 |
+
"eval_steps_per_second": 6.883,
|
| 595 |
+
"eval_wer": 0.521891418563923,
|
| 596 |
+
"step": 5800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 13.66,
|
| 600 |
+
"eval_loss": 0.6069247722625732,
|
| 601 |
+
"eval_runtime": 1.9188,
|
| 602 |
+
"eval_samples_per_second": 52.117,
|
| 603 |
+
"eval_steps_per_second": 6.775,
|
| 604 |
+
"eval_wer": 0.5236427320490368,
|
| 605 |
+
"step": 5900
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 13.89,
|
| 609 |
+
"learning_rate": 2.488294314381271e-05,
|
| 610 |
+
"loss": 0.1622,
|
| 611 |
+
"step": 6000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 13.89,
|
| 615 |
+
"eval_loss": 0.5849621295928955,
|
| 616 |
+
"eval_runtime": 1.8831,
|
| 617 |
+
"eval_samples_per_second": 53.103,
|
| 618 |
+
"eval_steps_per_second": 6.903,
|
| 619 |
+
"eval_wer": 0.5131348511383538,
|
| 620 |
+
"step": 6000
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 14.12,
|
| 624 |
+
"eval_loss": 0.5757902264595032,
|
| 625 |
+
"eval_runtime": 1.9024,
|
| 626 |
+
"eval_samples_per_second": 52.566,
|
| 627 |
+
"eval_steps_per_second": 6.834,
|
| 628 |
+
"eval_wer": 0.5096322241681261,
|
| 629 |
+
"step": 6100
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 14.35,
|
| 633 |
+
"eval_loss": 0.5752313137054443,
|
| 634 |
+
"eval_runtime": 1.8854,
|
| 635 |
+
"eval_samples_per_second": 53.038,
|
| 636 |
+
"eval_steps_per_second": 6.895,
|
| 637 |
+
"eval_wer": 0.500875656742557,
|
| 638 |
+
"step": 6200
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 14.58,
|
| 642 |
+
"eval_loss": 0.5727171301841736,
|
| 643 |
+
"eval_runtime": 1.9071,
|
| 644 |
+
"eval_samples_per_second": 52.437,
|
| 645 |
+
"eval_steps_per_second": 6.817,
|
| 646 |
+
"eval_wer": 0.5183887915936952,
|
| 647 |
+
"step": 6300
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 14.81,
|
| 651 |
+
"eval_loss": 0.5795398354530334,
|
| 652 |
+
"eval_runtime": 1.8851,
|
| 653 |
+
"eval_samples_per_second": 53.048,
|
| 654 |
+
"eval_steps_per_second": 6.896,
|
| 655 |
+
"eval_wer": 0.5043782837127846,
|
| 656 |
+
"step": 6400
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 15.0,
|
| 660 |
+
"step": 6480,
|
| 661 |
+
"total_flos": 1.5048474592820595e+19,
|
| 662 |
+
"train_loss": 0.5518865008413055,
|
| 663 |
+
"train_runtime": 5863.5283,
|
| 664 |
+
"train_samples_per_second": 35.331,
|
| 665 |
+
"train_steps_per_second": 1.105
|
| 666 |
+
}
|
| 667 |
+
],
|
| 668 |
+
"max_steps": 6480,
|
| 669 |
+
"num_train_epochs": 15,
|
| 670 |
+
"total_flos": 1.5048474592820595e+19,
|
| 671 |
+
"trial_name": null,
|
| 672 |
+
"trial_params": null
|
| 673 |
+
}
|