End of training
Browse files- README.md +11 -7
- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +509 -0
README.md
CHANGED
|
@@ -1,6 +1,10 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
tags:
|
|
|
|
|
|
|
| 4 |
- generated_from_trainer
|
| 5 |
datasets:
|
| 6 |
- common_voice_13_0
|
|
@@ -13,15 +17,15 @@ model-index:
|
|
| 13 |
name: Automatic Speech Recognition
|
| 14 |
type: automatic-speech-recognition
|
| 15 |
dataset:
|
| 16 |
-
name:
|
| 17 |
type: common_voice_13_0
|
| 18 |
config: eo
|
| 19 |
split: validation
|
| 20 |
-
args: eo
|
| 21 |
metrics:
|
| 22 |
- name: Wer
|
| 23 |
type: wer
|
| 24 |
-
value: 0.
|
| 25 |
---
|
| 26 |
|
| 27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -29,11 +33,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 29 |
|
| 30 |
# wav2vec2-common_voice_13_0-eo-10
|
| 31 |
|
| 32 |
-
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the
|
| 33 |
It achieves the following results on the evaluation set:
|
| 34 |
-
- Loss: 0.
|
| 35 |
-
- Cer: 0.
|
| 36 |
-
- Wer: 0.
|
| 37 |
|
| 38 |
## Model description
|
| 39 |
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- eo
|
| 4 |
license: apache-2.0
|
| 5 |
tags:
|
| 6 |
+
- automatic-speech-recognition
|
| 7 |
+
- mozilla-foundation/common_voice_13_0
|
| 8 |
- generated_from_trainer
|
| 9 |
datasets:
|
| 10 |
- common_voice_13_0
|
|
|
|
| 17 |
name: Automatic Speech Recognition
|
| 18 |
type: automatic-speech-recognition
|
| 19 |
dataset:
|
| 20 |
+
name: MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - EO
|
| 21 |
type: common_voice_13_0
|
| 22 |
config: eo
|
| 23 |
split: validation
|
| 24 |
+
args: 'Config: eo, Training split: train, Eval split: validation'
|
| 25 |
metrics:
|
| 26 |
- name: Wer
|
| 27 |
type: wer
|
| 28 |
+
value: 0.0656526475637132
|
| 29 |
---
|
| 30 |
|
| 31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 33 |
|
| 34 |
# wav2vec2-common_voice_13_0-eo-10
|
| 35 |
|
| 36 |
+
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - EO dataset.
|
| 37 |
It achieves the following results on the evaluation set:
|
| 38 |
+
- Loss: 0.0453
|
| 39 |
+
- Cer: 0.0118
|
| 40 |
+
- Wer: 0.0657
|
| 41 |
|
| 42 |
## Model description
|
| 43 |
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_cer": 0.011829732593934161,
|
| 4 |
+
"eval_loss": 0.04534716159105301,
|
| 5 |
+
"eval_runtime": 1308.5745,
|
| 6 |
+
"eval_samples": 14892,
|
| 7 |
+
"eval_samples_per_second": 11.38,
|
| 8 |
+
"eval_steps_per_second": 1.423,
|
| 9 |
+
"eval_wer": 0.0656526475637132,
|
| 10 |
+
"train_loss": 0.05346228662136532,
|
| 11 |
+
"train_runtime": 59652.5991,
|
| 12 |
+
"train_samples": 143849,
|
| 13 |
+
"train_samples_per_second": 12.057,
|
| 14 |
+
"train_steps_per_second": 0.377
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_cer": 0.011829732593934161,
|
| 4 |
+
"eval_loss": 0.04534716159105301,
|
| 5 |
+
"eval_runtime": 1308.5745,
|
| 6 |
+
"eval_samples": 14892,
|
| 7 |
+
"eval_samples_per_second": 11.38,
|
| 8 |
+
"eval_steps_per_second": 1.423,
|
| 9 |
+
"eval_wer": 0.0656526475637132
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"train_loss": 0.05346228662136532,
|
| 4 |
+
"train_runtime": 59652.5991,
|
| 5 |
+
"train_samples": 143849,
|
| 6 |
+
"train_samples_per_second": 12.057,
|
| 7 |
+
"train_steps_per_second": 0.377
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,509 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.9996663329996665,
|
| 5 |
+
"global_step": 22475,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.11,
|
| 12 |
+
"learning_rate": 2.982e-05,
|
| 13 |
+
"loss": 10.9521,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.22,
|
| 18 |
+
"learning_rate": 2.9866994413407824e-05,
|
| 19 |
+
"loss": 2.9894,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.22,
|
| 24 |
+
"eval_cer": 1.0,
|
| 25 |
+
"eval_loss": 2.925715684890747,
|
| 26 |
+
"eval_runtime": 1316.2549,
|
| 27 |
+
"eval_samples_per_second": 11.314,
|
| 28 |
+
"eval_steps_per_second": 1.415,
|
| 29 |
+
"eval_wer": 1.0,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.33,
|
| 34 |
+
"learning_rate": 2.973291620111732e-05,
|
| 35 |
+
"loss": 2.6003,
|
| 36 |
+
"step": 1500
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.44,
|
| 40 |
+
"learning_rate": 2.9598837988826818e-05,
|
| 41 |
+
"loss": 0.7104,
|
| 42 |
+
"step": 2000
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.44,
|
| 46 |
+
"eval_cer": 0.045692850294096206,
|
| 47 |
+
"eval_loss": 0.2129213511943817,
|
| 48 |
+
"eval_runtime": 1315.3965,
|
| 49 |
+
"eval_samples_per_second": 11.321,
|
| 50 |
+
"eval_steps_per_second": 1.416,
|
| 51 |
+
"eval_wer": 0.25382939389937936,
|
| 52 |
+
"step": 2000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.56,
|
| 56 |
+
"learning_rate": 2.9464759776536317e-05,
|
| 57 |
+
"loss": 0.3546,
|
| 58 |
+
"step": 2500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.67,
|
| 62 |
+
"learning_rate": 2.933068156424581e-05,
|
| 63 |
+
"loss": 0.2853,
|
| 64 |
+
"step": 3000
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.67,
|
| 68 |
+
"eval_cer": 0.027426083445944053,
|
| 69 |
+
"eval_loss": 0.11085120588541031,
|
| 70 |
+
"eval_runtime": 1321.34,
|
| 71 |
+
"eval_samples_per_second": 11.27,
|
| 72 |
+
"eval_steps_per_second": 1.409,
|
| 73 |
+
"eval_wer": 0.15825135349267133,
|
| 74 |
+
"step": 3000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.78,
|
| 78 |
+
"learning_rate": 2.9196603351955308e-05,
|
| 79 |
+
"loss": 0.2563,
|
| 80 |
+
"step": 3500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.89,
|
| 84 |
+
"learning_rate": 2.9062525139664803e-05,
|
| 85 |
+
"loss": 0.2327,
|
| 86 |
+
"step": 4000
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.89,
|
| 90 |
+
"eval_cer": 0.02314500709307346,
|
| 91 |
+
"eval_loss": 0.09091737866401672,
|
| 92 |
+
"eval_runtime": 1319.6784,
|
| 93 |
+
"eval_samples_per_second": 11.285,
|
| 94 |
+
"eval_steps_per_second": 1.411,
|
| 95 |
+
"eval_wer": 0.1320315594876535,
|
| 96 |
+
"step": 4000
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 1.0,
|
| 100 |
+
"learning_rate": 2.8928446927374302e-05,
|
| 101 |
+
"loss": 0.2072,
|
| 102 |
+
"step": 4500
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 1.11,
|
| 106 |
+
"learning_rate": 2.8794368715083797e-05,
|
| 107 |
+
"loss": 0.1917,
|
| 108 |
+
"step": 5000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 1.11,
|
| 112 |
+
"eval_cer": 0.02055449330783939,
|
| 113 |
+
"eval_loss": 0.07748183608055115,
|
| 114 |
+
"eval_runtime": 1318.284,
|
| 115 |
+
"eval_samples_per_second": 11.297,
|
| 116 |
+
"eval_steps_per_second": 1.412,
|
| 117 |
+
"eval_wer": 0.11881024692988247,
|
| 118 |
+
"step": 5000
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 1.22,
|
| 122 |
+
"learning_rate": 2.8660290502793296e-05,
|
| 123 |
+
"loss": 0.1853,
|
| 124 |
+
"step": 5500
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.33,
|
| 128 |
+
"learning_rate": 2.8526212290502795e-05,
|
| 129 |
+
"loss": 0.1803,
|
| 130 |
+
"step": 6000
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.33,
|
| 134 |
+
"eval_cer": 0.018439187408532995,
|
| 135 |
+
"eval_loss": 0.0698164626955986,
|
| 136 |
+
"eval_runtime": 1319.3199,
|
| 137 |
+
"eval_samples_per_second": 11.288,
|
| 138 |
+
"eval_steps_per_second": 1.411,
|
| 139 |
+
"eval_wer": 0.10547339231480259,
|
| 140 |
+
"step": 6000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 1.45,
|
| 144 |
+
"learning_rate": 2.839213407821229e-05,
|
| 145 |
+
"loss": 0.1724,
|
| 146 |
+
"step": 6500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 1.56,
|
| 150 |
+
"learning_rate": 2.825805586592179e-05,
|
| 151 |
+
"loss": 0.1661,
|
| 152 |
+
"step": 7000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.56,
|
| 156 |
+
"eval_cer": 0.01694487588523239,
|
| 157 |
+
"eval_loss": 0.06446516513824463,
|
| 158 |
+
"eval_runtime": 1319.092,
|
| 159 |
+
"eval_samples_per_second": 11.29,
|
| 160 |
+
"eval_steps_per_second": 1.412,
|
| 161 |
+
"eval_wer": 0.09606496764822395,
|
| 162 |
+
"step": 7000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.67,
|
| 166 |
+
"learning_rate": 2.8123977653631284e-05,
|
| 167 |
+
"loss": 0.161,
|
| 168 |
+
"step": 7500
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.78,
|
| 172 |
+
"learning_rate": 2.7989899441340783e-05,
|
| 173 |
+
"loss": 0.1635,
|
| 174 |
+
"step": 8000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 1.78,
|
| 178 |
+
"eval_cer": 0.017010760164402304,
|
| 179 |
+
"eval_loss": 0.0639302060008049,
|
| 180 |
+
"eval_runtime": 1319.0191,
|
| 181 |
+
"eval_samples_per_second": 11.29,
|
| 182 |
+
"eval_steps_per_second": 1.412,
|
| 183 |
+
"eval_wer": 0.09642809982833751,
|
| 184 |
+
"step": 8000
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.89,
|
| 188 |
+
"learning_rate": 2.785582122905028e-05,
|
| 189 |
+
"loss": 0.1547,
|
| 190 |
+
"step": 8500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 2.0,
|
| 194 |
+
"learning_rate": 2.7721743016759777e-05,
|
| 195 |
+
"loss": 0.1555,
|
| 196 |
+
"step": 9000
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 2.0,
|
| 200 |
+
"eval_cer": 0.015554297056795053,
|
| 201 |
+
"eval_loss": 0.0592055507004261,
|
| 202 |
+
"eval_runtime": 1320.0874,
|
| 203 |
+
"eval_samples_per_second": 11.281,
|
| 204 |
+
"eval_steps_per_second": 1.411,
|
| 205 |
+
"eval_wer": 0.08808431268981909,
|
| 206 |
+
"step": 9000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 2.11,
|
| 210 |
+
"learning_rate": 2.7587664804469276e-05,
|
| 211 |
+
"loss": 0.1405,
|
| 212 |
+
"step": 9500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.22,
|
| 216 |
+
"learning_rate": 2.745358659217877e-05,
|
| 217 |
+
"loss": 0.1386,
|
| 218 |
+
"step": 10000
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 2.22,
|
| 222 |
+
"eval_cer": 0.014650140459676017,
|
| 223 |
+
"eval_loss": 0.055936481803655624,
|
| 224 |
+
"eval_runtime": 1320.1961,
|
| 225 |
+
"eval_samples_per_second": 11.28,
|
| 226 |
+
"eval_steps_per_second": 1.41,
|
| 227 |
+
"eval_wer": 0.08208437871385184,
|
| 228 |
+
"step": 10000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.34,
|
| 232 |
+
"learning_rate": 2.731950837988827e-05,
|
| 233 |
+
"loss": 0.1353,
|
| 234 |
+
"step": 10500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.45,
|
| 238 |
+
"learning_rate": 2.7185430167597765e-05,
|
| 239 |
+
"loss": 0.1338,
|
| 240 |
+
"step": 11000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.45,
|
| 244 |
+
"eval_cer": 0.014644533286980706,
|
| 245 |
+
"eval_loss": 0.05477109178900719,
|
| 246 |
+
"eval_runtime": 1321.7173,
|
| 247 |
+
"eval_samples_per_second": 11.267,
|
| 248 |
+
"eval_steps_per_second": 1.409,
|
| 249 |
+
"eval_wer": 0.08314901624191205,
|
| 250 |
+
"step": 11000
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 2.56,
|
| 254 |
+
"learning_rate": 2.7051620111731843e-05,
|
| 255 |
+
"loss": 0.1345,
|
| 256 |
+
"step": 11500
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.67,
|
| 260 |
+
"learning_rate": 2.691754189944134e-05,
|
| 261 |
+
"loss": 0.1307,
|
| 262 |
+
"step": 12000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.67,
|
| 266 |
+
"eval_cer": 0.013712340826385111,
|
| 267 |
+
"eval_loss": 0.0528746023774147,
|
| 268 |
+
"eval_runtime": 1321.5984,
|
| 269 |
+
"eval_samples_per_second": 11.268,
|
| 270 |
+
"eval_steps_per_second": 1.409,
|
| 271 |
+
"eval_wer": 0.07593589066420177,
|
| 272 |
+
"step": 12000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.78,
|
| 276 |
+
"learning_rate": 1.3624573378839592e-05,
|
| 277 |
+
"loss": 0.1264,
|
| 278 |
+
"step": 12500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 2.89,
|
| 282 |
+
"learning_rate": 1.29419795221843e-05,
|
| 283 |
+
"loss": 0.1297,
|
| 284 |
+
"step": 13000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 2.89,
|
| 288 |
+
"eval_cer": 0.013381517637361713,
|
| 289 |
+
"eval_loss": 0.05041037127375603,
|
| 290 |
+
"eval_runtime": 1303.0852,
|
| 291 |
+
"eval_samples_per_second": 11.428,
|
| 292 |
+
"eval_steps_per_second": 1.429,
|
| 293 |
+
"eval_wer": 0.07453287996830846,
|
| 294 |
+
"step": 13000
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 3.0,
|
| 298 |
+
"learning_rate": 1.225938566552901e-05,
|
| 299 |
+
"loss": 0.122,
|
| 300 |
+
"step": 13500
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 3.11,
|
| 304 |
+
"learning_rate": 1.1578156996587031e-05,
|
| 305 |
+
"loss": 0.1201,
|
| 306 |
+
"step": 14000
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 3.11,
|
| 310 |
+
"eval_cer": 0.013103962588943778,
|
| 311 |
+
"eval_loss": 0.04994073510169983,
|
| 312 |
+
"eval_runtime": 1305.505,
|
| 313 |
+
"eval_samples_per_second": 11.407,
|
| 314 |
+
"eval_steps_per_second": 1.426,
|
| 315 |
+
"eval_wer": 0.07336095338703288,
|
| 316 |
+
"step": 14000
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 3.23,
|
| 320 |
+
"learning_rate": 1.089556313993174e-05,
|
| 321 |
+
"loss": 0.1185,
|
| 322 |
+
"step": 14500
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 3.34,
|
| 326 |
+
"learning_rate": 1.0214334470989761e-05,
|
| 327 |
+
"loss": 0.1152,
|
| 328 |
+
"step": 15000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 3.34,
|
| 332 |
+
"eval_cer": 0.012759121468182099,
|
| 333 |
+
"eval_loss": 0.0484052337706089,
|
| 334 |
+
"eval_runtime": 1304.2248,
|
| 335 |
+
"eval_samples_per_second": 11.418,
|
| 336 |
+
"eval_steps_per_second": 1.428,
|
| 337 |
+
"eval_wer": 0.07121517232272548,
|
| 338 |
+
"step": 15000
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.45,
|
| 342 |
+
"learning_rate": 9.531740614334472e-06,
|
| 343 |
+
"loss": 0.1165,
|
| 344 |
+
"step": 15500
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 3.56,
|
| 348 |
+
"learning_rate": 8.849146757679181e-06,
|
| 349 |
+
"loss": 0.1144,
|
| 350 |
+
"step": 16000
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 3.56,
|
| 354 |
+
"eval_cer": 0.012485771799285646,
|
| 355 |
+
"eval_loss": 0.04772350192070007,
|
| 356 |
+
"eval_runtime": 1302.5642,
|
| 357 |
+
"eval_samples_per_second": 11.433,
|
| 358 |
+
"eval_steps_per_second": 1.429,
|
| 359 |
+
"eval_wer": 0.06949029446718606,
|
| 360 |
+
"step": 16000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 3.67,
|
| 364 |
+
"learning_rate": 8.166552901023892e-06,
|
| 365 |
+
"loss": 0.1129,
|
| 366 |
+
"step": 16500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 3.78,
|
| 370 |
+
"learning_rate": 7.483959044368601e-06,
|
| 371 |
+
"loss": 0.1179,
|
| 372 |
+
"step": 17000
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 3.78,
|
| 376 |
+
"eval_cer": 0.012212422130389193,
|
| 377 |
+
"eval_loss": 0.04678424075245857,
|
| 378 |
+
"eval_runtime": 1303.8755,
|
| 379 |
+
"eval_samples_per_second": 11.421,
|
| 380 |
+
"eval_steps_per_second": 1.428,
|
| 381 |
+
"eval_wer": 0.0678809586689555,
|
| 382 |
+
"step": 17000
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 3.89,
|
| 386 |
+
"learning_rate": 6.8040955631399314e-06,
|
| 387 |
+
"loss": 0.1118,
|
| 388 |
+
"step": 17500
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 4.0,
|
| 392 |
+
"learning_rate": 6.1215017064846416e-06,
|
| 393 |
+
"loss": 0.1112,
|
| 394 |
+
"step": 18000
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 4.0,
|
| 398 |
+
"eval_cer": 0.012117100194568893,
|
| 399 |
+
"eval_loss": 0.04677628353238106,
|
| 400 |
+
"eval_runtime": 1306.4884,
|
| 401 |
+
"eval_samples_per_second": 11.398,
|
| 402 |
+
"eval_steps_per_second": 1.425,
|
| 403 |
+
"eval_wer": 0.0676416215502443,
|
| 404 |
+
"step": 18000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 4.12,
|
| 408 |
+
"learning_rate": 5.438907849829352e-06,
|
| 409 |
+
"loss": 0.1117,
|
| 410 |
+
"step": 18500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 4.23,
|
| 414 |
+
"learning_rate": 4.756313993174062e-06,
|
| 415 |
+
"loss": 0.1141,
|
| 416 |
+
"step": 19000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 4.23,
|
| 420 |
+
"eval_cer": 0.012068037433484913,
|
| 421 |
+
"eval_loss": 0.04616771265864372,
|
| 422 |
+
"eval_runtime": 1302.5466,
|
| 423 |
+
"eval_samples_per_second": 11.433,
|
| 424 |
+
"eval_steps_per_second": 1.43,
|
| 425 |
+
"eval_wer": 0.06683282714908227,
|
| 426 |
+
"step": 19000
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 4.34,
|
| 430 |
+
"learning_rate": 4.073720136518771e-06,
|
| 431 |
+
"loss": 0.1108,
|
| 432 |
+
"step": 19500
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 4.45,
|
| 436 |
+
"learning_rate": 3.3911262798634816e-06,
|
| 437 |
+
"loss": 0.1085,
|
| 438 |
+
"step": 20000
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 4.45,
|
| 442 |
+
"eval_cer": 0.011937670668318913,
|
| 443 |
+
"eval_loss": 0.04581034556031227,
|
| 444 |
+
"eval_runtime": 1302.9709,
|
| 445 |
+
"eval_samples_per_second": 11.429,
|
| 446 |
+
"eval_steps_per_second": 1.429,
|
| 447 |
+
"eval_wer": 0.06640367093622078,
|
| 448 |
+
"step": 20000
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 4.56,
|
| 452 |
+
"learning_rate": 2.708532423208191e-06,
|
| 453 |
+
"loss": 0.1132,
|
| 454 |
+
"step": 20500
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 4.67,
|
| 458 |
+
"learning_rate": 2.025938566552901e-06,
|
| 459 |
+
"loss": 0.105,
|
| 460 |
+
"step": 21000
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 4.67,
|
| 464 |
+
"eval_cer": 0.01189281328675642,
|
| 465 |
+
"eval_loss": 0.04555269330739975,
|
| 466 |
+
"eval_runtime": 1307.8114,
|
| 467 |
+
"eval_samples_per_second": 11.387,
|
| 468 |
+
"eval_steps_per_second": 1.424,
|
| 469 |
+
"eval_wer": 0.06604053875610723,
|
| 470 |
+
"step": 21000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 4.78,
|
| 474 |
+
"learning_rate": 1.3433447098976108e-06,
|
| 475 |
+
"loss": 0.1123,
|
| 476 |
+
"step": 21500
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 4.89,
|
| 480 |
+
"learning_rate": 6.621160409556314e-07,
|
| 481 |
+
"loss": 0.1072,
|
| 482 |
+
"step": 22000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 4.89,
|
| 486 |
+
"eval_cer": 0.011854964871063065,
|
| 487 |
+
"eval_loss": 0.04538210481405258,
|
| 488 |
+
"eval_runtime": 1302.5621,
|
| 489 |
+
"eval_samples_per_second": 11.433,
|
| 490 |
+
"eval_steps_per_second": 1.429,
|
| 491 |
+
"eval_wer": 0.06575168361283507,
|
| 492 |
+
"step": 22000
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 5.0,
|
| 496 |
+
"step": 22475,
|
| 497 |
+
"total_flos": 1.3602266687003442e+20,
|
| 498 |
+
"train_loss": 0.05346228662136532,
|
| 499 |
+
"train_runtime": 59652.5991,
|
| 500 |
+
"train_samples_per_second": 12.057,
|
| 501 |
+
"train_steps_per_second": 0.377
|
| 502 |
+
}
|
| 503 |
+
],
|
| 504 |
+
"max_steps": 22475,
|
| 505 |
+
"num_train_epochs": 5,
|
| 506 |
+
"total_flos": 1.3602266687003442e+20,
|
| 507 |
+
"trial_name": null,
|
| 508 |
+
"trial_params": null
|
| 509 |
+
}
|