End of training
Browse files- README.md +4 -4
- all_results.json +12 -12
- eval_results.json +7 -7
- train_results.json +6 -6
- trainer_state.json +440 -12
README.md
CHANGED
|
@@ -23,7 +23,7 @@ model-index:
|
|
| 23 |
metrics:
|
| 24 |
- name: Wer
|
| 25 |
type: wer
|
| 26 |
-
value: 0.
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -33,9 +33,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [mouseyy/result_data-1](https://huggingface.co/mouseyy/result_data-1) on the common_voice_17_0 dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Wer: 0.
|
| 38 |
-
- Cer: 0.
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
|
|
|
| 23 |
metrics:
|
| 24 |
- name: Wer
|
| 25 |
type: wer
|
| 26 |
+
value: 0.3390885932635154
|
| 27 |
---
|
| 28 |
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [mouseyy/result_data-1](https://huggingface.co/mouseyy/result_data-1) on the common_voice_17_0 dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
+
- Loss: 0.2363
|
| 37 |
+
- Wer: 0.3391
|
| 38 |
+
- Cer: 0.1637
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
all_results.json
CHANGED
|
@@ -1,16 +1,16 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 0
|
| 3 |
-
"eval_cer": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 16.
|
| 6 |
"eval_samples": 500,
|
| 7 |
-
"eval_samples_per_second": 30.
|
| 8 |
-
"eval_steps_per_second": 0.
|
| 9 |
-
"eval_wer": 0.
|
| 10 |
-
"total_flos":
|
| 11 |
-
"train_loss": 0.
|
| 12 |
-
"train_runtime":
|
| 13 |
"train_samples": 35144,
|
| 14 |
-
"train_samples_per_second":
|
| 15 |
-
"train_steps_per_second": 0.
|
| 16 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_cer": 0.16366974008093788,
|
| 4 |
+
"eval_loss": 0.23625396192073822,
|
| 5 |
+
"eval_runtime": 16.2931,
|
| 6 |
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 30.688,
|
| 8 |
+
"eval_steps_per_second": 0.982,
|
| 9 |
+
"eval_wer": 0.3390885932635154,
|
| 10 |
+
"total_flos": 4.052184710714386e+19,
|
| 11 |
+
"train_loss": 0.1293755126064533,
|
| 12 |
+
"train_runtime": 8966.103,
|
| 13 |
"train_samples": 35144,
|
| 14 |
+
"train_samples_per_second": 19.598,
|
| 15 |
+
"train_steps_per_second": 0.613
|
| 16 |
}
|
eval_results.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 0
|
| 3 |
-
"eval_cer": 0.
|
| 4 |
-
"eval_loss": 0.
|
| 5 |
-
"eval_runtime": 16.
|
| 6 |
"eval_samples": 500,
|
| 7 |
-
"eval_samples_per_second": 30.
|
| 8 |
-
"eval_steps_per_second": 0.
|
| 9 |
-
"eval_wer": 0.
|
| 10 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_cer": 0.16366974008093788,
|
| 4 |
+
"eval_loss": 0.23625396192073822,
|
| 5 |
+
"eval_runtime": 16.2931,
|
| 6 |
"eval_samples": 500,
|
| 7 |
+
"eval_samples_per_second": 30.688,
|
| 8 |
+
"eval_steps_per_second": 0.982,
|
| 9 |
+
"eval_wer": 0.3390885932635154
|
| 10 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch": 0
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 35144,
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second": 0.
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 4.052184710714386e+19,
|
| 4 |
+
"train_loss": 0.1293755126064533,
|
| 5 |
+
"train_runtime": 8966.103,
|
| 6 |
"train_samples": 35144,
|
| 7 |
+
"train_samples_per_second": 19.598,
|
| 8 |
+
"train_steps_per_second": 0.613
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -1,27 +1,455 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0
|
| 5 |
"eval_steps": 1000.0,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
|
| 18 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 19 |
}
|
| 20 |
],
|
| 21 |
"logging_steps": 100,
|
| 22 |
-
"max_steps":
|
| 23 |
"num_input_tokens_seen": 0,
|
| 24 |
-
"num_train_epochs":
|
| 25 |
"save_steps": 1000,
|
| 26 |
"stateful_callbacks": {
|
| 27 |
"TrainerControl": {
|
|
@@ -35,7 +463,7 @@
|
|
| 35 |
"attributes": {}
|
| 36 |
}
|
| 37 |
},
|
| 38 |
-
"total_flos":
|
| 39 |
"train_batch_size": 16,
|
| 40 |
"trial_name": null,
|
| 41 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 5.0,
|
| 5 |
"eval_steps": 1000.0,
|
| 6 |
+
"global_step": 5495,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.09099181073703366,
|
| 13 |
+
"grad_norm": 1.2852333784103394,
|
| 14 |
+
"learning_rate": 3.080335788799027e-05,
|
| 15 |
+
"loss": 0.1592,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.18198362147406733,
|
| 20 |
+
"grad_norm": 1.0739330053329468,
|
| 21 |
+
"learning_rate": 4.1080243101273595e-05,
|
| 22 |
+
"loss": 0.1603,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.272975432211101,
|
| 27 |
+
"grad_norm": 0.9484991431236267,
|
| 28 |
+
"learning_rate": 4.030441225894548e-05,
|
| 29 |
+
"loss": 0.1797,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.36396724294813465,
|
| 34 |
+
"grad_norm": 1.8888756036758423,
|
| 35 |
+
"learning_rate": 3.9528581416617364e-05,
|
| 36 |
+
"loss": 0.1754,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.4549590536851683,
|
| 41 |
+
"grad_norm": 1.1603392362594604,
|
| 42 |
+
"learning_rate": 3.8752750574289256e-05,
|
| 43 |
+
"loss": 0.1818,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.545950864422202,
|
| 48 |
+
"grad_norm": 1.3092777729034424,
|
| 49 |
+
"learning_rate": 3.797691973196114e-05,
|
| 50 |
+
"loss": 0.1805,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.6369426751592356,
|
| 55 |
+
"grad_norm": 1.2778502702713013,
|
| 56 |
+
"learning_rate": 3.7201088889633025e-05,
|
| 57 |
+
"loss": 0.1642,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.7279344858962693,
|
| 62 |
+
"grad_norm": 3.04357647895813,
|
| 63 |
+
"learning_rate": 3.642525804730491e-05,
|
| 64 |
+
"loss": 0.1626,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.818926296633303,
|
| 69 |
+
"grad_norm": 0.8114387392997742,
|
| 70 |
+
"learning_rate": 3.56494272049768e-05,
|
| 71 |
+
"loss": 0.1625,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.9099181073703366,
|
| 76 |
+
"grad_norm": 1.6360591650009155,
|
| 77 |
+
"learning_rate": 3.4873596362648686e-05,
|
| 78 |
+
"loss": 0.1644,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.9099181073703366,
|
| 83 |
+
"eval_cer": 0.17117944011014227,
|
| 84 |
+
"eval_loss": 0.24234530329704285,
|
| 85 |
+
"eval_runtime": 16.3456,
|
| 86 |
+
"eval_samples_per_second": 30.589,
|
| 87 |
+
"eval_steps_per_second": 0.979,
|
| 88 |
+
"eval_wer": 0.36909142371921877,
|
| 89 |
+
"step": 1000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 1.0009099181073704,
|
| 93 |
+
"grad_norm": 1.0669325590133667,
|
| 94 |
+
"learning_rate": 3.409776552032057e-05,
|
| 95 |
+
"loss": 0.1556,
|
| 96 |
+
"step": 1100
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.091901728844404,
|
| 100 |
+
"grad_norm": 0.9532307386398315,
|
| 101 |
+
"learning_rate": 3.332193467799246e-05,
|
| 102 |
+
"loss": 0.1368,
|
| 103 |
+
"step": 1200
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 1.1828935395814377,
|
| 107 |
+
"grad_norm": 2.0215086936950684,
|
| 108 |
+
"learning_rate": 3.254610383566435e-05,
|
| 109 |
+
"loss": 0.1458,
|
| 110 |
+
"step": 1300
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 1.2738853503184713,
|
| 114 |
+
"grad_norm": 1.452370285987854,
|
| 115 |
+
"learning_rate": 3.177027299333623e-05,
|
| 116 |
+
"loss": 0.1433,
|
| 117 |
+
"step": 1400
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 1.364877161055505,
|
| 121 |
+
"grad_norm": 0.7807812690734863,
|
| 122 |
+
"learning_rate": 3.099444215100812e-05,
|
| 123 |
+
"loss": 0.1433,
|
| 124 |
+
"step": 1500
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.4558689717925386,
|
| 128 |
+
"grad_norm": 0.5693336129188538,
|
| 129 |
+
"learning_rate": 3.0218611308680008e-05,
|
| 130 |
+
"loss": 0.1372,
|
| 131 |
+
"step": 1600
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 1.5468607825295724,
|
| 135 |
+
"grad_norm": 1.0194613933563232,
|
| 136 |
+
"learning_rate": 2.9442780466351896e-05,
|
| 137 |
+
"loss": 0.1429,
|
| 138 |
+
"step": 1700
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 1.6378525932666061,
|
| 142 |
+
"grad_norm": 1.2318494319915771,
|
| 143 |
+
"learning_rate": 2.866694962402378e-05,
|
| 144 |
+
"loss": 0.1403,
|
| 145 |
+
"step": 1800
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 1.7288444040036397,
|
| 149 |
+
"grad_norm": 0.6520742774009705,
|
| 150 |
+
"learning_rate": 2.789111878169567e-05,
|
| 151 |
+
"loss": 0.1442,
|
| 152 |
+
"step": 1900
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.8198362147406733,
|
| 156 |
+
"grad_norm": 0.8180395364761353,
|
| 157 |
+
"learning_rate": 2.7115287939367557e-05,
|
| 158 |
+
"loss": 0.1373,
|
| 159 |
+
"step": 2000
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 1.8198362147406733,
|
| 163 |
+
"eval_cer": 0.1699278234386082,
|
| 164 |
+
"eval_loss": 0.23588787019252777,
|
| 165 |
+
"eval_runtime": 16.2786,
|
| 166 |
+
"eval_samples_per_second": 30.715,
|
| 167 |
+
"eval_steps_per_second": 0.983,
|
| 168 |
+
"eval_wer": 0.3586187376167563,
|
| 169 |
+
"step": 2000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.910828025477707,
|
| 173 |
+
"grad_norm": 0.929227352142334,
|
| 174 |
+
"learning_rate": 2.6339457097039442e-05,
|
| 175 |
+
"loss": 0.1407,
|
| 176 |
+
"step": 2100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 2.001819836214741,
|
| 180 |
+
"grad_norm": 1.137143850326538,
|
| 181 |
+
"learning_rate": 2.556362625471133e-05,
|
| 182 |
+
"loss": 0.1363,
|
| 183 |
+
"step": 2200
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 2.092811646951774,
|
| 187 |
+
"grad_norm": 1.959381341934204,
|
| 188 |
+
"learning_rate": 2.4787795412383218e-05,
|
| 189 |
+
"loss": 0.1274,
|
| 190 |
+
"step": 2300
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 2.183803457688808,
|
| 194 |
+
"grad_norm": 0.8095592856407166,
|
| 195 |
+
"learning_rate": 2.4011964570055103e-05,
|
| 196 |
+
"loss": 0.1288,
|
| 197 |
+
"step": 2400
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 2.2747952684258417,
|
| 201 |
+
"grad_norm": 0.717786431312561,
|
| 202 |
+
"learning_rate": 2.323613372772699e-05,
|
| 203 |
+
"loss": 0.1314,
|
| 204 |
+
"step": 2500
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.3657870791628755,
|
| 208 |
+
"grad_norm": 1.3463408946990967,
|
| 209 |
+
"learning_rate": 2.246030288539888e-05,
|
| 210 |
+
"loss": 0.1243,
|
| 211 |
+
"step": 2600
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 2.4567788898999092,
|
| 215 |
+
"grad_norm": 1.703278660774231,
|
| 216 |
+
"learning_rate": 2.1684472043070764e-05,
|
| 217 |
+
"loss": 0.1226,
|
| 218 |
+
"step": 2700
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.5477707006369426,
|
| 222 |
+
"grad_norm": 1.0250530242919922,
|
| 223 |
+
"learning_rate": 2.0908641200742652e-05,
|
| 224 |
+
"loss": 0.1217,
|
| 225 |
+
"step": 2800
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 2.6387625113739763,
|
| 229 |
+
"grad_norm": 2.0595955848693848,
|
| 230 |
+
"learning_rate": 2.0132810358414537e-05,
|
| 231 |
+
"loss": 0.1162,
|
| 232 |
+
"step": 2900
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 2.72975432211101,
|
| 236 |
+
"grad_norm": 0.5484445691108704,
|
| 237 |
+
"learning_rate": 1.9356979516086425e-05,
|
| 238 |
+
"loss": 0.1219,
|
| 239 |
+
"step": 3000
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 2.72975432211101,
|
| 243 |
+
"eval_cer": 0.16654845842546623,
|
| 244 |
+
"eval_loss": 0.2350655198097229,
|
| 245 |
+
"eval_runtime": 16.364,
|
| 246 |
+
"eval_samples_per_second": 30.555,
|
| 247 |
+
"eval_steps_per_second": 0.978,
|
| 248 |
+
"eval_wer": 0.34871214265496747,
|
| 249 |
+
"step": 3000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.8207461328480434,
|
| 253 |
+
"grad_norm": 0.8254183530807495,
|
| 254 |
+
"learning_rate": 1.858114867375831e-05,
|
| 255 |
+
"loss": 0.1302,
|
| 256 |
+
"step": 3100
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.911737943585077,
|
| 260 |
+
"grad_norm": 1.559329867362976,
|
| 261 |
+
"learning_rate": 1.7805317831430197e-05,
|
| 262 |
+
"loss": 0.1224,
|
| 263 |
+
"step": 3200
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 3.002729754322111,
|
| 267 |
+
"grad_norm": 0.9372493624687195,
|
| 268 |
+
"learning_rate": 1.7029486989102082e-05,
|
| 269 |
+
"loss": 0.1186,
|
| 270 |
+
"step": 3300
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 3.0937215650591448,
|
| 274 |
+
"grad_norm": 0.5558441877365112,
|
| 275 |
+
"learning_rate": 1.625365614677397e-05,
|
| 276 |
+
"loss": 0.1084,
|
| 277 |
+
"step": 3400
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 3.1847133757961785,
|
| 281 |
+
"grad_norm": 0.9594938158988953,
|
| 282 |
+
"learning_rate": 1.547782530444586e-05,
|
| 283 |
+
"loss": 0.114,
|
| 284 |
+
"step": 3500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 3.275705186533212,
|
| 288 |
+
"grad_norm": 0.835418164730072,
|
| 289 |
+
"learning_rate": 1.4701994462117745e-05,
|
| 290 |
+
"loss": 0.1099,
|
| 291 |
+
"step": 3600
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 3.3666969972702456,
|
| 295 |
+
"grad_norm": 1.4973269701004028,
|
| 296 |
+
"learning_rate": 1.3926163619789631e-05,
|
| 297 |
+
"loss": 0.1111,
|
| 298 |
+
"step": 3700
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 3.4576888080072794,
|
| 302 |
+
"grad_norm": 0.5722721219062805,
|
| 303 |
+
"learning_rate": 1.3150332777461518e-05,
|
| 304 |
+
"loss": 0.1117,
|
| 305 |
+
"step": 3800
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 3.548680618744313,
|
| 309 |
+
"grad_norm": 1.0630409717559814,
|
| 310 |
+
"learning_rate": 1.2374501935133406e-05,
|
| 311 |
+
"loss": 0.1171,
|
| 312 |
+
"step": 3900
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 3.6396724294813465,
|
| 316 |
+
"grad_norm": 4.837808609008789,
|
| 317 |
+
"learning_rate": 1.1598671092805292e-05,
|
| 318 |
+
"loss": 0.1106,
|
| 319 |
+
"step": 4000
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 3.6396724294813465,
|
| 323 |
+
"eval_cer": 0.16671534064833743,
|
| 324 |
+
"eval_loss": 0.23994748294353485,
|
| 325 |
+
"eval_runtime": 16.3865,
|
| 326 |
+
"eval_samples_per_second": 30.513,
|
| 327 |
+
"eval_steps_per_second": 0.976,
|
| 328 |
+
"eval_wer": 0.3447495046702519,
|
| 329 |
+
"step": 4000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 3.7306642402183803,
|
| 333 |
+
"grad_norm": 1.5471532344818115,
|
| 334 |
+
"learning_rate": 1.0822840250477179e-05,
|
| 335 |
+
"loss": 0.1061,
|
| 336 |
+
"step": 4100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 3.821656050955414,
|
| 340 |
+
"grad_norm": 0.7752039432525635,
|
| 341 |
+
"learning_rate": 1.0047009408149065e-05,
|
| 342 |
+
"loss": 0.1032,
|
| 343 |
+
"step": 4200
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 3.912647861692448,
|
| 347 |
+
"grad_norm": 1.4473958015441895,
|
| 348 |
+
"learning_rate": 9.271178565820951e-06,
|
| 349 |
+
"loss": 0.1117,
|
| 350 |
+
"step": 4300
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 4.003639672429482,
|
| 354 |
+
"grad_norm": 1.3828502893447876,
|
| 355 |
+
"learning_rate": 8.495347723492838e-06,
|
| 356 |
+
"loss": 0.1089,
|
| 357 |
+
"step": 4400
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 4.094631483166515,
|
| 361 |
+
"grad_norm": 0.6974540948867798,
|
| 362 |
+
"learning_rate": 7.719516881164726e-06,
|
| 363 |
+
"loss": 0.1072,
|
| 364 |
+
"step": 4500
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 4.185623293903548,
|
| 368 |
+
"grad_norm": 0.8870043158531189,
|
| 369 |
+
"learning_rate": 6.943686038836612e-06,
|
| 370 |
+
"loss": 0.1028,
|
| 371 |
+
"step": 4600
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 4.276615104640582,
|
| 375 |
+
"grad_norm": 0.685329020023346,
|
| 376 |
+
"learning_rate": 6.1678551965085e-06,
|
| 377 |
+
"loss": 0.1063,
|
| 378 |
+
"step": 4700
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 4.367606915377616,
|
| 382 |
+
"grad_norm": 0.7879564762115479,
|
| 383 |
+
"learning_rate": 5.392024354180386e-06,
|
| 384 |
+
"loss": 0.0988,
|
| 385 |
+
"step": 4800
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 4.45859872611465,
|
| 389 |
+
"grad_norm": 0.7849826216697693,
|
| 390 |
+
"learning_rate": 4.616193511852273e-06,
|
| 391 |
+
"loss": 0.1055,
|
| 392 |
+
"step": 4900
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 4.549590536851683,
|
| 396 |
+
"grad_norm": 0.9511623382568359,
|
| 397 |
+
"learning_rate": 3.840362669524161e-06,
|
| 398 |
+
"loss": 0.0986,
|
| 399 |
+
"step": 5000
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 4.549590536851683,
|
| 403 |
+
"eval_cer": 0.16387834285952688,
|
| 404 |
+
"eval_loss": 0.24175503849983215,
|
| 405 |
+
"eval_runtime": 16.3966,
|
| 406 |
+
"eval_samples_per_second": 30.494,
|
| 407 |
+
"eval_steps_per_second": 0.976,
|
| 408 |
+
"eval_wer": 0.3388055476931786,
|
| 409 |
+
"step": 5000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 4.640582347588717,
|
| 413 |
+
"grad_norm": 1.7066700458526611,
|
| 414 |
+
"learning_rate": 3.064531827196047e-06,
|
| 415 |
+
"loss": 0.1091,
|
| 416 |
+
"step": 5100
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.731574158325751,
|
| 420 |
+
"grad_norm": 2.1451282501220703,
|
| 421 |
+
"learning_rate": 2.288700984867934e-06,
|
| 422 |
+
"loss": 0.1034,
|
| 423 |
+
"step": 5200
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 4.822565969062785,
|
| 427 |
+
"grad_norm": 1.3042198419570923,
|
| 428 |
+
"learning_rate": 1.5128701425398207e-06,
|
| 429 |
+
"loss": 0.1025,
|
| 430 |
+
"step": 5300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 4.9135577797998184,
|
| 434 |
+
"grad_norm": 0.7029935717582703,
|
| 435 |
+
"learning_rate": 7.370393002117075e-07,
|
| 436 |
+
"loss": 0.1028,
|
| 437 |
+
"step": 5400
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 5.0,
|
| 441 |
+
"step": 5495,
|
| 442 |
+
"total_flos": 4.052184710714386e+19,
|
| 443 |
+
"train_loss": 0.1293755126064533,
|
| 444 |
+
"train_runtime": 8966.103,
|
| 445 |
+
"train_samples_per_second": 19.598,
|
| 446 |
+
"train_steps_per_second": 0.613
|
| 447 |
}
|
| 448 |
],
|
| 449 |
"logging_steps": 100,
|
| 450 |
+
"max_steps": 5495,
|
| 451 |
"num_input_tokens_seen": 0,
|
| 452 |
+
"num_train_epochs": 5,
|
| 453 |
"save_steps": 1000,
|
| 454 |
"stateful_callbacks": {
|
| 455 |
"TrainerControl": {
|
|
|
|
| 463 |
"attributes": {}
|
| 464 |
}
|
| 465 |
},
|
| 466 |
+
"total_flos": 4.052184710714386e+19,
|
| 467 |
"train_batch_size": 16,
|
| 468 |
"trial_name": null,
|
| 469 |
"trial_params": null
|