Upload trainer_state.json
Browse files- trainer_state.json +944 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,944 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 198.62941847206386,
|
| 5 |
+
"global_step": 87000,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 3.42,
|
| 12 |
+
"learning_rate": 5.988000000000001e-05,
|
| 13 |
+
"loss": 1.7702,
|
| 14 |
+
"step": 1500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 3.42,
|
| 18 |
+
"eval_cer": 0.8902483010528606,
|
| 19 |
+
"eval_loss": 1.379656434059143,
|
| 20 |
+
"eval_runtime": 157.8178,
|
| 21 |
+
"eval_samples_per_second": 23.134,
|
| 22 |
+
"eval_steps_per_second": 1.451,
|
| 23 |
+
"eval_wer": 0.9197613252728272,
|
| 24 |
+
"step": 1500
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 6.85,
|
| 28 |
+
"learning_rate": 7.906822429906543e-05,
|
| 29 |
+
"loss": 0.6423,
|
| 30 |
+
"step": 3000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 6.85,
|
| 34 |
+
"eval_cer": 0.8909467730023416,
|
| 35 |
+
"eval_loss": 1.096530795097351,
|
| 36 |
+
"eval_runtime": 155.6392,
|
| 37 |
+
"eval_samples_per_second": 23.458,
|
| 38 |
+
"eval_steps_per_second": 1.471,
|
| 39 |
+
"eval_wer": 0.92135398557633,
|
| 40 |
+
"step": 3000
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 10.27,
|
| 44 |
+
"learning_rate": 7.766728971962618e-05,
|
| 45 |
+
"loss": 0.5752,
|
| 46 |
+
"step": 4500
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 10.27,
|
| 50 |
+
"eval_cer": 0.8896414319820001,
|
| 51 |
+
"eval_loss": 0.9504674077033997,
|
| 52 |
+
"eval_runtime": 156.6885,
|
| 53 |
+
"eval_samples_per_second": 23.301,
|
| 54 |
+
"eval_steps_per_second": 1.461,
|
| 55 |
+
"eval_wer": 0.9201763142251483,
|
| 56 |
+
"step": 4500
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 13.7,
|
| 60 |
+
"learning_rate": 7.626728971962617e-05,
|
| 61 |
+
"loss": 0.5136,
|
| 62 |
+
"step": 6000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 13.7,
|
| 66 |
+
"eval_cer": 0.8656071267039567,
|
| 67 |
+
"eval_loss": 0.49109727144241333,
|
| 68 |
+
"eval_runtime": 157.3097,
|
| 69 |
+
"eval_samples_per_second": 23.209,
|
| 70 |
+
"eval_steps_per_second": 1.456,
|
| 71 |
+
"eval_wer": 0.8955349431913772,
|
| 72 |
+
"step": 6000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 17.12,
|
| 76 |
+
"learning_rate": 7.486542056074767e-05,
|
| 77 |
+
"loss": 0.5015,
|
| 78 |
+
"step": 7500
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 17.12,
|
| 82 |
+
"eval_cer": 0.818803780908815,
|
| 83 |
+
"eval_loss": 0.4844619333744049,
|
| 84 |
+
"eval_runtime": 157.3062,
|
| 85 |
+
"eval_samples_per_second": 23.21,
|
| 86 |
+
"eval_steps_per_second": 1.456,
|
| 87 |
+
"eval_wer": 0.849381442142689,
|
| 88 |
+
"step": 7500
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 20.55,
|
| 92 |
+
"learning_rate": 7.346355140186916e-05,
|
| 93 |
+
"loss": 0.462,
|
| 94 |
+
"step": 9000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 20.55,
|
| 98 |
+
"eval_cer": 0.8027961778698895,
|
| 99 |
+
"eval_loss": 0.4619074761867523,
|
| 100 |
+
"eval_runtime": 157.0856,
|
| 101 |
+
"eval_samples_per_second": 23.242,
|
| 102 |
+
"eval_steps_per_second": 1.458,
|
| 103 |
+
"eval_wer": 0.8343072488475645,
|
| 104 |
+
"step": 9000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 23.97,
|
| 108 |
+
"learning_rate": 7.206168224299066e-05,
|
| 109 |
+
"loss": 0.444,
|
| 110 |
+
"step": 10500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 23.97,
|
| 114 |
+
"eval_cer": 0.7654622796521381,
|
| 115 |
+
"eval_loss": 0.4942396581172943,
|
| 116 |
+
"eval_runtime": 156.8664,
|
| 117 |
+
"eval_samples_per_second": 23.275,
|
| 118 |
+
"eval_steps_per_second": 1.46,
|
| 119 |
+
"eval_wer": 0.7982929373366682,
|
| 120 |
+
"step": 10500
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 27.4,
|
| 124 |
+
"learning_rate": 7.066074766355141e-05,
|
| 125 |
+
"loss": 0.4318,
|
| 126 |
+
"step": 12000
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 27.4,
|
| 130 |
+
"eval_cer": 0.7344947814985086,
|
| 131 |
+
"eval_loss": 0.4564968943595886,
|
| 132 |
+
"eval_runtime": 156.676,
|
| 133 |
+
"eval_samples_per_second": 23.303,
|
| 134 |
+
"eval_steps_per_second": 1.462,
|
| 135 |
+
"eval_wer": 0.7668771520541953,
|
| 136 |
+
"step": 12000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 30.82,
|
| 140 |
+
"learning_rate": 6.92588785046729e-05,
|
| 141 |
+
"loss": 0.3877,
|
| 142 |
+
"step": 13500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 30.82,
|
| 146 |
+
"eval_cer": 0.7305844836173977,
|
| 147 |
+
"eval_loss": 0.44864970445632935,
|
| 148 |
+
"eval_runtime": 156.7014,
|
| 149 |
+
"eval_samples_per_second": 23.299,
|
| 150 |
+
"eval_steps_per_second": 1.461,
|
| 151 |
+
"eval_wer": 0.7626823988604627,
|
| 152 |
+
"step": 13500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 34.25,
|
| 156 |
+
"learning_rate": 6.78570093457944e-05,
|
| 157 |
+
"loss": 0.3734,
|
| 158 |
+
"step": 15000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 34.25,
|
| 162 |
+
"eval_cer": 0.716420388510709,
|
| 163 |
+
"eval_loss": 0.4351600408554077,
|
| 164 |
+
"eval_runtime": 156.7512,
|
| 165 |
+
"eval_samples_per_second": 23.292,
|
| 166 |
+
"eval_steps_per_second": 1.461,
|
| 167 |
+
"eval_wer": 0.7498626050090288,
|
| 168 |
+
"step": 15000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 37.67,
|
| 172 |
+
"learning_rate": 6.64551401869159e-05,
|
| 173 |
+
"loss": 0.3595,
|
| 174 |
+
"step": 16500
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 37.67,
|
| 178 |
+
"eval_cer": 0.6001820607212581,
|
| 179 |
+
"eval_loss": 0.4632013738155365,
|
| 180 |
+
"eval_runtime": 156.8539,
|
| 181 |
+
"eval_samples_per_second": 23.276,
|
| 182 |
+
"eval_steps_per_second": 1.46,
|
| 183 |
+
"eval_wer": 0.6390493388216557,
|
| 184 |
+
"step": 16500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 41.1,
|
| 188 |
+
"learning_rate": 6.505327102803739e-05,
|
| 189 |
+
"loss": 0.3466,
|
| 190 |
+
"step": 18000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 41.1,
|
| 194 |
+
"eval_cer": 0.592630548414984,
|
| 195 |
+
"eval_loss": 0.4672980308532715,
|
| 196 |
+
"eval_runtime": 155.7423,
|
| 197 |
+
"eval_samples_per_second": 23.443,
|
| 198 |
+
"eval_steps_per_second": 1.47,
|
| 199 |
+
"eval_wer": 0.6305925369284088,
|
| 200 |
+
"step": 18000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 44.52,
|
| 204 |
+
"learning_rate": 6.365140186915889e-05,
|
| 205 |
+
"loss": 0.3346,
|
| 206 |
+
"step": 19500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 44.52,
|
| 210 |
+
"eval_cer": 0.5607641970148912,
|
| 211 |
+
"eval_loss": 0.48392337560653687,
|
| 212 |
+
"eval_runtime": 156.8128,
|
| 213 |
+
"eval_samples_per_second": 23.283,
|
| 214 |
+
"eval_steps_per_second": 1.46,
|
| 215 |
+
"eval_wer": 0.601790060453796,
|
| 216 |
+
"step": 19500
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 47.94,
|
| 220 |
+
"learning_rate": 6.224953271028038e-05,
|
| 221 |
+
"loss": 0.3224,
|
| 222 |
+
"step": 21000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 47.94,
|
| 226 |
+
"eval_cer": 0.5324932586006516,
|
| 227 |
+
"eval_loss": 0.47312837839126587,
|
| 228 |
+
"eval_runtime": 155.7519,
|
| 229 |
+
"eval_samples_per_second": 23.441,
|
| 230 |
+
"eval_steps_per_second": 1.47,
|
| 231 |
+
"eval_wer": 0.5755448131988918,
|
| 232 |
+
"step": 21000
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 51.37,
|
| 236 |
+
"learning_rate": 6.0849532710280374e-05,
|
| 237 |
+
"loss": 0.3115,
|
| 238 |
+
"step": 22500
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 51.37,
|
| 242 |
+
"eval_cer": 0.5504359724504343,
|
| 243 |
+
"eval_loss": 0.46229076385498047,
|
| 244 |
+
"eval_runtime": 157.1658,
|
| 245 |
+
"eval_samples_per_second": 23.23,
|
| 246 |
+
"eval_steps_per_second": 1.457,
|
| 247 |
+
"eval_wer": 0.5916060072454827,
|
| 248 |
+
"step": 22500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 54.79,
|
| 252 |
+
"learning_rate": 5.9448598130841125e-05,
|
| 253 |
+
"loss": 0.3004,
|
| 254 |
+
"step": 24000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 54.79,
|
| 258 |
+
"eval_cer": 0.4698426148041702,
|
| 259 |
+
"eval_loss": 0.4939606785774231,
|
| 260 |
+
"eval_runtime": 157.0777,
|
| 261 |
+
"eval_samples_per_second": 23.243,
|
| 262 |
+
"eval_steps_per_second": 1.458,
|
| 263 |
+
"eval_wer": 0.5165603023811393,
|
| 264 |
+
"step": 24000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 58.22,
|
| 268 |
+
"learning_rate": 5.804672897196262e-05,
|
| 269 |
+
"loss": 0.2911,
|
| 270 |
+
"step": 25500
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 58.22,
|
| 274 |
+
"eval_cer": 0.4784017587752695,
|
| 275 |
+
"eval_loss": 0.49099990725517273,
|
| 276 |
+
"eval_runtime": 155.9127,
|
| 277 |
+
"eval_samples_per_second": 23.417,
|
| 278 |
+
"eval_steps_per_second": 1.469,
|
| 279 |
+
"eval_wer": 0.5228972958422593,
|
| 280 |
+
"step": 25500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 61.64,
|
| 284 |
+
"learning_rate": 5.664579439252337e-05,
|
| 285 |
+
"loss": 0.2816,
|
| 286 |
+
"step": 27000
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 61.64,
|
| 290 |
+
"eval_cer": 0.48053152570319524,
|
| 291 |
+
"eval_loss": 0.48187437653541565,
|
| 292 |
+
"eval_runtime": 157.1059,
|
| 293 |
+
"eval_samples_per_second": 23.239,
|
| 294 |
+
"eval_steps_per_second": 1.458,
|
| 295 |
+
"eval_wer": 0.524153478616853,
|
| 296 |
+
"step": 27000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 65.07,
|
| 300 |
+
"learning_rate": 5.5244859813084115e-05,
|
| 301 |
+
"loss": 0.2753,
|
| 302 |
+
"step": 28500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 65.07,
|
| 306 |
+
"eval_cer": 0.43459840725494797,
|
| 307 |
+
"eval_loss": 0.5083651542663574,
|
| 308 |
+
"eval_runtime": 156.7879,
|
| 309 |
+
"eval_samples_per_second": 23.286,
|
| 310 |
+
"eval_steps_per_second": 1.461,
|
| 311 |
+
"eval_wer": 0.48194798057403065,
|
| 312 |
+
"step": 28500
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 68.49,
|
| 316 |
+
"learning_rate": 5.384299065420561e-05,
|
| 317 |
+
"loss": 0.2628,
|
| 318 |
+
"step": 30000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 68.49,
|
| 322 |
+
"eval_cer": 0.441067860557518,
|
| 323 |
+
"eval_loss": 0.5106444954872131,
|
| 324 |
+
"eval_runtime": 161.9652,
|
| 325 |
+
"eval_samples_per_second": 22.542,
|
| 326 |
+
"eval_steps_per_second": 1.414,
|
| 327 |
+
"eval_wer": 0.4881952466941083,
|
| 328 |
+
"step": 30000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 71.92,
|
| 332 |
+
"learning_rate": 5.2441121495327106e-05,
|
| 333 |
+
"loss": 0.2534,
|
| 334 |
+
"step": 31500
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 71.92,
|
| 338 |
+
"eval_cer": 0.40500495228062544,
|
| 339 |
+
"eval_loss": 0.555060088634491,
|
| 340 |
+
"eval_runtime": 162.1242,
|
| 341 |
+
"eval_samples_per_second": 22.52,
|
| 342 |
+
"eval_steps_per_second": 1.412,
|
| 343 |
+
"eval_wer": 0.45390818649827835,
|
| 344 |
+
"step": 31500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 75.34,
|
| 348 |
+
"learning_rate": 5.104018691588786e-05,
|
| 349 |
+
"loss": 0.2464,
|
| 350 |
+
"step": 33000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 75.34,
|
| 354 |
+
"eval_cer": 0.4066767048154488,
|
| 355 |
+
"eval_loss": 0.5396202206611633,
|
| 356 |
+
"eval_runtime": 161.0534,
|
| 357 |
+
"eval_samples_per_second": 22.669,
|
| 358 |
+
"eval_steps_per_second": 1.422,
|
| 359 |
+
"eval_wer": 0.45515315335524176,
|
| 360 |
+
"step": 33000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 78.77,
|
| 364 |
+
"learning_rate": 4.96392523364486e-05,
|
| 365 |
+
"loss": 0.2447,
|
| 366 |
+
"step": 34500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 78.77,
|
| 370 |
+
"eval_cer": 0.38928360823738883,
|
| 371 |
+
"eval_loss": 0.5427628755569458,
|
| 372 |
+
"eval_runtime": 161.5749,
|
| 373 |
+
"eval_samples_per_second": 22.596,
|
| 374 |
+
"eval_steps_per_second": 1.417,
|
| 375 |
+
"eval_wer": 0.4384190042508328,
|
| 376 |
+
"step": 34500
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 82.19,
|
| 380 |
+
"learning_rate": 4.82392523364486e-05,
|
| 381 |
+
"loss": 0.2311,
|
| 382 |
+
"step": 36000
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 82.19,
|
| 386 |
+
"eval_cer": 0.3919687176169511,
|
| 387 |
+
"eval_loss": 0.537956714630127,
|
| 388 |
+
"eval_runtime": 161.1887,
|
| 389 |
+
"eval_samples_per_second": 22.65,
|
| 390 |
+
"eval_steps_per_second": 1.421,
|
| 391 |
+
"eval_wer": 0.43954059601386286,
|
| 392 |
+
"step": 36000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 85.62,
|
| 396 |
+
"learning_rate": 4.6837383177570096e-05,
|
| 397 |
+
"loss": 0.2238,
|
| 398 |
+
"step": 37500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 85.62,
|
| 402 |
+
"eval_cer": 0.36581037059089583,
|
| 403 |
+
"eval_loss": 0.5633525848388672,
|
| 404 |
+
"eval_runtime": 160.6641,
|
| 405 |
+
"eval_samples_per_second": 22.724,
|
| 406 |
+
"eval_steps_per_second": 1.425,
|
| 407 |
+
"eval_wer": 0.4180508978342063,
|
| 408 |
+
"step": 37500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 89.04,
|
| 412 |
+
"learning_rate": 4.543644859813085e-05,
|
| 413 |
+
"loss": 0.2176,
|
| 414 |
+
"step": 39000
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 89.04,
|
| 418 |
+
"eval_cer": 0.34971688985326366,
|
| 419 |
+
"eval_loss": 0.5895515084266663,
|
| 420 |
+
"eval_runtime": 161.4249,
|
| 421 |
+
"eval_samples_per_second": 22.617,
|
| 422 |
+
"eval_steps_per_second": 1.419,
|
| 423 |
+
"eval_wer": 0.4021355107168093,
|
| 424 |
+
"step": 39000
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 92.47,
|
| 428 |
+
"learning_rate": 4.4034579439252336e-05,
|
| 429 |
+
"loss": 0.2106,
|
| 430 |
+
"step": 40500
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 92.47,
|
| 434 |
+
"eval_cer": 0.354417262562476,
|
| 435 |
+
"eval_loss": 0.5802760124206543,
|
| 436 |
+
"eval_runtime": 161.093,
|
| 437 |
+
"eval_samples_per_second": 22.664,
|
| 438 |
+
"eval_steps_per_second": 1.422,
|
| 439 |
+
"eval_wer": 0.40675646878049326,
|
| 440 |
+
"step": 40500
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 95.89,
|
| 444 |
+
"learning_rate": 4.263271028037384e-05,
|
| 445 |
+
"loss": 0.2058,
|
| 446 |
+
"step": 42000
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 95.89,
|
| 450 |
+
"eval_cer": 0.34699742939421874,
|
| 451 |
+
"eval_loss": 0.5864582657814026,
|
| 452 |
+
"eval_runtime": 161.933,
|
| 453 |
+
"eval_samples_per_second": 22.546,
|
| 454 |
+
"eval_steps_per_second": 1.414,
|
| 455 |
+
"eval_wer": 0.400744736930652,
|
| 456 |
+
"step": 42000
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 99.31,
|
| 460 |
+
"learning_rate": 4.1230841121495333e-05,
|
| 461 |
+
"loss": 0.1995,
|
| 462 |
+
"step": 43500
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 99.31,
|
| 466 |
+
"eval_cer": 0.33578752712303983,
|
| 467 |
+
"eval_loss": 0.5888575315475464,
|
| 468 |
+
"eval_runtime": 161.5039,
|
| 469 |
+
"eval_samples_per_second": 22.606,
|
| 470 |
+
"eval_steps_per_second": 1.418,
|
| 471 |
+
"eval_wer": 0.3900335355937146,
|
| 472 |
+
"step": 43500
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 102.74,
|
| 476 |
+
"learning_rate": 3.982990654205608e-05,
|
| 477 |
+
"loss": 0.1956,
|
| 478 |
+
"step": 45000
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 102.74,
|
| 482 |
+
"eval_cer": 0.3373562264194152,
|
| 483 |
+
"eval_loss": 0.5785849094390869,
|
| 484 |
+
"eval_runtime": 159.8624,
|
| 485 |
+
"eval_samples_per_second": 22.838,
|
| 486 |
+
"eval_steps_per_second": 1.432,
|
| 487 |
+
"eval_wer": 0.3900335355937146,
|
| 488 |
+
"step": 45000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 106.16,
|
| 492 |
+
"learning_rate": 3.842803738317757e-05,
|
| 493 |
+
"loss": 0.191,
|
| 494 |
+
"step": 46500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 106.16,
|
| 498 |
+
"eval_cer": 0.33468256739968055,
|
| 499 |
+
"eval_loss": 0.5981589555740356,
|
| 500 |
+
"eval_runtime": 159.9077,
|
| 501 |
+
"eval_samples_per_second": 22.832,
|
| 502 |
+
"eval_steps_per_second": 1.432,
|
| 503 |
+
"eval_wer": 0.38939422828878745,
|
| 504 |
+
"step": 46500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 109.59,
|
| 508 |
+
"learning_rate": 3.7027102803738324e-05,
|
| 509 |
+
"loss": 0.1851,
|
| 510 |
+
"step": 48000
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 109.59,
|
| 514 |
+
"eval_cer": 0.32407380901944843,
|
| 515 |
+
"eval_loss": 0.603424072265625,
|
| 516 |
+
"eval_runtime": 159.431,
|
| 517 |
+
"eval_samples_per_second": 22.9,
|
| 518 |
+
"eval_steps_per_second": 1.436,
|
| 519 |
+
"eval_wer": 0.3804551419374376,
|
| 520 |
+
"step": 48000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 113.01,
|
| 524 |
+
"learning_rate": 3.562616822429907e-05,
|
| 525 |
+
"loss": 0.1822,
|
| 526 |
+
"step": 49500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 113.01,
|
| 530 |
+
"eval_cer": 0.30877612828983153,
|
| 531 |
+
"eval_loss": 0.6061017513275146,
|
| 532 |
+
"eval_runtime": 158.5159,
|
| 533 |
+
"eval_samples_per_second": 23.032,
|
| 534 |
+
"eval_steps_per_second": 1.445,
|
| 535 |
+
"eval_wer": 0.3677250754270461,
|
| 536 |
+
"step": 49500
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 116.44,
|
| 540 |
+
"learning_rate": 3.422523364485982e-05,
|
| 541 |
+
"loss": 0.183,
|
| 542 |
+
"step": 51000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 116.44,
|
| 546 |
+
"eval_cer": 0.3065089570439751,
|
| 547 |
+
"eval_loss": 0.6075730919837952,
|
| 548 |
+
"eval_runtime": 160.2872,
|
| 549 |
+
"eval_samples_per_second": 22.778,
|
| 550 |
+
"eval_steps_per_second": 1.429,
|
| 551 |
+
"eval_wer": 0.36392287935037404,
|
| 552 |
+
"step": 51000
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 119.86,
|
| 556 |
+
"learning_rate": 3.2823364485981314e-05,
|
| 557 |
+
"loss": 0.172,
|
| 558 |
+
"step": 52500
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 119.86,
|
| 562 |
+
"eval_cer": 0.2992551540932174,
|
| 563 |
+
"eval_loss": 0.6227861642837524,
|
| 564 |
+
"eval_runtime": 160.5993,
|
| 565 |
+
"eval_samples_per_second": 22.734,
|
| 566 |
+
"eval_steps_per_second": 1.426,
|
| 567 |
+
"eval_wer": 0.3576083177245146,
|
| 568 |
+
"step": 52500
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 123.29,
|
| 572 |
+
"learning_rate": 3.14214953271028e-05,
|
| 573 |
+
"loss": 0.1677,
|
| 574 |
+
"step": 54000
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 123.29,
|
| 578 |
+
"eval_cer": 0.300377289356318,
|
| 579 |
+
"eval_loss": 0.630536675453186,
|
| 580 |
+
"eval_runtime": 158.8853,
|
| 581 |
+
"eval_samples_per_second": 22.979,
|
| 582 |
+
"eval_steps_per_second": 1.441,
|
| 583 |
+
"eval_wer": 0.35942529638062337,
|
| 584 |
+
"step": 54000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 126.71,
|
| 588 |
+
"learning_rate": 3.002056074766355e-05,
|
| 589 |
+
"loss": 0.1641,
|
| 590 |
+
"step": 55500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 126.71,
|
| 594 |
+
"eval_cer": 0.2999479008627846,
|
| 595 |
+
"eval_loss": 0.6270943284034729,
|
| 596 |
+
"eval_runtime": 160.9314,
|
| 597 |
+
"eval_samples_per_second": 22.687,
|
| 598 |
+
"eval_steps_per_second": 1.423,
|
| 599 |
+
"eval_wer": 0.3598515012505748,
|
| 600 |
+
"step": 55500
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 130.14,
|
| 604 |
+
"learning_rate": 2.861869158878505e-05,
|
| 605 |
+
"loss": 0.1596,
|
| 606 |
+
"step": 57000
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 130.14,
|
| 610 |
+
"eval_cer": 0.3011845397241608,
|
| 611 |
+
"eval_loss": 0.6379602551460266,
|
| 612 |
+
"eval_runtime": 160.483,
|
| 613 |
+
"eval_samples_per_second": 22.75,
|
| 614 |
+
"eval_steps_per_second": 1.427,
|
| 615 |
+
"eval_wer": 0.3607712064962595,
|
| 616 |
+
"step": 57000
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 133.56,
|
| 620 |
+
"learning_rate": 2.7217757009345797e-05,
|
| 621 |
+
"loss": 0.1561,
|
| 622 |
+
"step": 58500
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 133.56,
|
| 626 |
+
"eval_cer": 0.2984937051646848,
|
| 627 |
+
"eval_loss": 0.6372924447059631,
|
| 628 |
+
"eval_runtime": 149.4703,
|
| 629 |
+
"eval_samples_per_second": 24.426,
|
| 630 |
+
"eval_steps_per_second": 1.532,
|
| 631 |
+
"eval_wer": 0.3564530782085936,
|
| 632 |
+
"step": 58500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 136.99,
|
| 636 |
+
"learning_rate": 2.5815887850467292e-05,
|
| 637 |
+
"loss": 0.1531,
|
| 638 |
+
"step": 60000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 136.99,
|
| 642 |
+
"eval_cer": 0.29037540004694645,
|
| 643 |
+
"eval_loss": 0.633769154548645,
|
| 644 |
+
"eval_runtime": 149.3294,
|
| 645 |
+
"eval_samples_per_second": 24.449,
|
| 646 |
+
"eval_steps_per_second": 1.534,
|
| 647 |
+
"eval_wer": 0.35054228961742506,
|
| 648 |
+
"step": 60000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 140.41,
|
| 652 |
+
"learning_rate": 2.441495327102804e-05,
|
| 653 |
+
"loss": 0.1501,
|
| 654 |
+
"step": 61500
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 140.41,
|
| 658 |
+
"eval_cer": 0.280087251741886,
|
| 659 |
+
"eval_loss": 0.6684455275535583,
|
| 660 |
+
"eval_runtime": 148.2869,
|
| 661 |
+
"eval_samples_per_second": 24.621,
|
| 662 |
+
"eval_steps_per_second": 1.544,
|
| 663 |
+
"eval_wer": 0.34238831750019627,
|
| 664 |
+
"step": 61500
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 143.83,
|
| 668 |
+
"learning_rate": 2.3013084112149535e-05,
|
| 669 |
+
"loss": 0.1461,
|
| 670 |
+
"step": 63000
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 143.83,
|
| 674 |
+
"eval_cer": 0.2885776935540199,
|
| 675 |
+
"eval_loss": 0.653037428855896,
|
| 676 |
+
"eval_runtime": 148.0728,
|
| 677 |
+
"eval_samples_per_second": 24.657,
|
| 678 |
+
"eval_steps_per_second": 1.547,
|
| 679 |
+
"eval_wer": 0.34928610684283135,
|
| 680 |
+
"step": 63000
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 147.26,
|
| 684 |
+
"learning_rate": 2.1612149532710283e-05,
|
| 685 |
+
"loss": 0.1438,
|
| 686 |
+
"step": 64500
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 147.26,
|
| 690 |
+
"eval_cer": 0.28241739996679394,
|
| 691 |
+
"eval_loss": 0.6683825254440308,
|
| 692 |
+
"eval_runtime": 149.486,
|
| 693 |
+
"eval_samples_per_second": 24.424,
|
| 694 |
+
"eval_steps_per_second": 1.532,
|
| 695 |
+
"eval_wer": 0.3449006830493837,
|
| 696 |
+
"step": 64500
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 150.68,
|
| 700 |
+
"learning_rate": 2.0210280373831778e-05,
|
| 701 |
+
"loss": 0.1399,
|
| 702 |
+
"step": 66000
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 150.68,
|
| 706 |
+
"eval_cer": 0.2820109121929157,
|
| 707 |
+
"eval_loss": 0.6699740290641785,
|
| 708 |
+
"eval_runtime": 148.9217,
|
| 709 |
+
"eval_samples_per_second": 24.516,
|
| 710 |
+
"eval_steps_per_second": 1.538,
|
| 711 |
+
"eval_wer": 0.34547269484852905,
|
| 712 |
+
"step": 66000
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 154.11,
|
| 716 |
+
"learning_rate": 1.8810280373831778e-05,
|
| 717 |
+
"loss": 0.1384,
|
| 718 |
+
"step": 67500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 154.11,
|
| 722 |
+
"eval_cer": 0.2813067150635209,
|
| 723 |
+
"eval_loss": 0.6824755668640137,
|
| 724 |
+
"eval_runtime": 149.3793,
|
| 725 |
+
"eval_samples_per_second": 24.441,
|
| 726 |
+
"eval_steps_per_second": 1.533,
|
| 727 |
+
"eval_wer": 0.34288181787592953,
|
| 728 |
+
"step": 67500
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 157.53,
|
| 732 |
+
"learning_rate": 1.740841121495327e-05,
|
| 733 |
+
"loss": 0.1356,
|
| 734 |
+
"step": 69000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 157.53,
|
| 738 |
+
"eval_cer": 0.28230862154843217,
|
| 739 |
+
"eval_loss": 0.6769393086433411,
|
| 740 |
+
"eval_runtime": 147.7768,
|
| 741 |
+
"eval_samples_per_second": 24.706,
|
| 742 |
+
"eval_steps_per_second": 1.55,
|
| 743 |
+
"eval_wer": 0.34468758061440796,
|
| 744 |
+
"step": 69000
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 160.96,
|
| 748 |
+
"learning_rate": 1.600654205607477e-05,
|
| 749 |
+
"loss": 0.1327,
|
| 750 |
+
"step": 70500
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 160.96,
|
| 754 |
+
"eval_cer": 0.2698334545163082,
|
| 755 |
+
"eval_loss": 0.6889060735702515,
|
| 756 |
+
"eval_runtime": 148.0859,
|
| 757 |
+
"eval_samples_per_second": 24.655,
|
| 758 |
+
"eval_steps_per_second": 1.546,
|
| 759 |
+
"eval_wer": 0.3343016408887493,
|
| 760 |
+
"step": 70500
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 164.38,
|
| 764 |
+
"learning_rate": 1.4604672897196264e-05,
|
| 765 |
+
"loss": 0.1301,
|
| 766 |
+
"step": 72000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 164.38,
|
| 770 |
+
"eval_cer": 0.2741445149913836,
|
| 771 |
+
"eval_loss": 0.667121171951294,
|
| 772 |
+
"eval_runtime": 147.8905,
|
| 773 |
+
"eval_samples_per_second": 24.687,
|
| 774 |
+
"eval_steps_per_second": 1.548,
|
| 775 |
+
"eval_wer": 0.33721777947262754,
|
| 776 |
+
"step": 72000
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 167.81,
|
| 780 |
+
"learning_rate": 1.3204672897196262e-05,
|
| 781 |
+
"loss": 0.1274,
|
| 782 |
+
"step": 73500
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 167.81,
|
| 786 |
+
"eval_cer": 0.2703658962482896,
|
| 787 |
+
"eval_loss": 0.6766042113304138,
|
| 788 |
+
"eval_runtime": 157.2135,
|
| 789 |
+
"eval_samples_per_second": 23.223,
|
| 790 |
+
"eval_steps_per_second": 1.457,
|
| 791 |
+
"eval_wer": 0.33454839107661594,
|
| 792 |
+
"step": 73500
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 171.23,
|
| 796 |
+
"learning_rate": 1.180373831775701e-05,
|
| 797 |
+
"loss": 0.1254,
|
| 798 |
+
"step": 75000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 171.23,
|
| 802 |
+
"eval_cer": 0.26700521563890145,
|
| 803 |
+
"eval_loss": 0.6990349888801575,
|
| 804 |
+
"eval_runtime": 147.5703,
|
| 805 |
+
"eval_samples_per_second": 24.741,
|
| 806 |
+
"eval_steps_per_second": 1.552,
|
| 807 |
+
"eval_wer": 0.33097051335254996,
|
| 808 |
+
"step": 75000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 174.66,
|
| 812 |
+
"learning_rate": 1.0401869158878507e-05,
|
| 813 |
+
"loss": 0.1232,
|
| 814 |
+
"step": 76500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 174.66,
|
| 818 |
+
"eval_cer": 0.26890025019036223,
|
| 819 |
+
"eval_loss": 0.6996525526046753,
|
| 820 |
+
"eval_runtime": 148.6819,
|
| 821 |
+
"eval_samples_per_second": 24.556,
|
| 822 |
+
"eval_steps_per_second": 1.54,
|
| 823 |
+
"eval_wer": 0.3331015377023071,
|
| 824 |
+
"step": 76500
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 178.08,
|
| 828 |
+
"learning_rate": 9e-06,
|
| 829 |
+
"loss": 0.1214,
|
| 830 |
+
"step": 78000
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 178.08,
|
| 834 |
+
"eval_cer": 0.2654765926019225,
|
| 835 |
+
"eval_loss": 0.7195149660110474,
|
| 836 |
+
"eval_runtime": 147.4964,
|
| 837 |
+
"eval_samples_per_second": 24.753,
|
| 838 |
+
"eval_steps_per_second": 1.553,
|
| 839 |
+
"eval_wer": 0.33113875211700444,
|
| 840 |
+
"step": 78000
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 181.51,
|
| 844 |
+
"learning_rate": 7.600000000000001e-06,
|
| 845 |
+
"loss": 0.1192,
|
| 846 |
+
"step": 79500
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 181.51,
|
| 850 |
+
"eval_cer": 0.2665872775051956,
|
| 851 |
+
"eval_loss": 0.7197592854499817,
|
| 852 |
+
"eval_runtime": 146.5503,
|
| 853 |
+
"eval_samples_per_second": 24.913,
|
| 854 |
+
"eval_steps_per_second": 1.563,
|
| 855 |
+
"eval_wer": 0.3315873888222165,
|
| 856 |
+
"step": 79500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 184.93,
|
| 860 |
+
"learning_rate": 6.198130841121495e-06,
|
| 861 |
+
"loss": 0.117,
|
| 862 |
+
"step": 81000
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 184.93,
|
| 866 |
+
"eval_cer": 0.26343270337270347,
|
| 867 |
+
"eval_loss": 0.7252310514450073,
|
| 868 |
+
"eval_runtime": 147.1458,
|
| 869 |
+
"eval_samples_per_second": 24.812,
|
| 870 |
+
"eval_steps_per_second": 1.556,
|
| 871 |
+
"eval_wer": 0.3294339326371987,
|
| 872 |
+
"step": 81000
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 188.36,
|
| 876 |
+
"learning_rate": 4.796261682242991e-06,
|
| 877 |
+
"loss": 0.1155,
|
| 878 |
+
"step": 82500
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 188.36,
|
| 882 |
+
"eval_cer": 0.26273995660313626,
|
| 883 |
+
"eval_loss": 0.7319917678833008,
|
| 884 |
+
"eval_runtime": 145.3615,
|
| 885 |
+
"eval_samples_per_second": 25.117,
|
| 886 |
+
"eval_steps_per_second": 1.575,
|
| 887 |
+
"eval_wer": 0.3283684204623201,
|
| 888 |
+
"step": 82500
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 191.78,
|
| 892 |
+
"learning_rate": 3.3953271028037387e-06,
|
| 893 |
+
"loss": 0.1146,
|
| 894 |
+
"step": 84000
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 191.78,
|
| 898 |
+
"eval_cer": 0.2623048429296891,
|
| 899 |
+
"eval_loss": 0.7414848208427429,
|
| 900 |
+
"eval_runtime": 146.2983,
|
| 901 |
+
"eval_samples_per_second": 24.956,
|
| 902 |
+
"eval_steps_per_second": 1.565,
|
| 903 |
+
"eval_wer": 0.32897408001435635,
|
| 904 |
+
"step": 84000
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 195.21,
|
| 908 |
+
"learning_rate": 1.9934579439252336e-06,
|
| 909 |
+
"loss": 0.1131,
|
| 910 |
+
"step": 85500
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 195.21,
|
| 914 |
+
"eval_cer": 0.26136018824391555,
|
| 915 |
+
"eval_loss": 0.7424480319023132,
|
| 916 |
+
"eval_runtime": 146.0724,
|
| 917 |
+
"eval_samples_per_second": 24.994,
|
| 918 |
+
"eval_steps_per_second": 1.568,
|
| 919 |
+
"eval_wer": 0.328177749862605,
|
| 920 |
+
"step": 85500
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 198.63,
|
| 924 |
+
"learning_rate": 5.925233644859814e-07,
|
| 925 |
+
"loss": 0.1127,
|
| 926 |
+
"step": 87000
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 198.63,
|
| 930 |
+
"eval_cer": 0.261194158026416,
|
| 931 |
+
"eval_loss": 0.7422526478767395,
|
| 932 |
+
"eval_runtime": 146.2208,
|
| 933 |
+
"eval_samples_per_second": 24.969,
|
| 934 |
+
"eval_steps_per_second": 1.566,
|
| 935 |
+
"eval_wer": 0.3272580446169203,
|
| 936 |
+
"step": 87000
|
| 937 |
+
}
|
| 938 |
+
],
|
| 939 |
+
"max_steps": 87600,
|
| 940 |
+
"num_train_epochs": 200,
|
| 941 |
+
"total_flos": 9.618016300986444e+20,
|
| 942 |
+
"trial_name": null,
|
| 943 |
+
"trial_params": null
|
| 944 |
+
}
|