End of training
Browse files- README.md +11 -7
- all_results.json +16 -0
- eval_results.json +10 -0
- train_results.json +9 -0
- trainer_state.json +1316 -0
README.md
CHANGED
|
@@ -1,8 +1,12 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
|
|
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: rinna/japanese-hubert-base
|
| 5 |
tags:
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- common_voice_13_0
|
|
@@ -15,15 +19,15 @@ model-index:
|
|
| 15 |
name: Automatic Speech Recognition
|
| 16 |
type: automatic-speech-recognition
|
| 17 |
dataset:
|
| 18 |
-
name:
|
| 19 |
type: common_voice_13_0
|
| 20 |
config: ja
|
| 21 |
split: test
|
| 22 |
-
args: ja
|
| 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
|
|
@@ -31,11 +35,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 31 |
|
| 32 |
# Hubert-common_voice-phonemes-debug
|
| 33 |
|
| 34 |
-
This model is a fine-tuned version of [rinna/japanese-hubert-base](https://huggingface.co/rinna/japanese-hubert-base) on the
|
| 35 |
It achieves the following results on the evaluation set:
|
| 36 |
-
- Loss: 0.
|
| 37 |
-
- Wer: 0.
|
| 38 |
-
- Cer: 0.
|
| 39 |
|
| 40 |
## Model description
|
| 41 |
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
+
language:
|
| 4 |
+
- ja
|
| 5 |
license: apache-2.0
|
| 6 |
base_model: rinna/japanese-hubert-base
|
| 7 |
tags:
|
| 8 |
+
- automatic-speech-recognition
|
| 9 |
+
- mozilla-foundation/common_voice_13_0
|
| 10 |
- generated_from_trainer
|
| 11 |
datasets:
|
| 12 |
- common_voice_13_0
|
|
|
|
| 19 |
name: Automatic Speech Recognition
|
| 20 |
type: automatic-speech-recognition
|
| 21 |
dataset:
|
| 22 |
+
name: MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - JA
|
| 23 |
type: common_voice_13_0
|
| 24 |
config: ja
|
| 25 |
split: test
|
| 26 |
+
args: 'Config: ja, Training split: train+validation, Eval split: test'
|
| 27 |
metrics:
|
| 28 |
- name: Wer
|
| 29 |
type: wer
|
| 30 |
+
value: 0.9844573966139328
|
| 31 |
---
|
| 32 |
|
| 33 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 35 |
|
| 36 |
# Hubert-common_voice-phonemes-debug
|
| 37 |
|
| 38 |
+
This model is a fine-tuned version of [rinna/japanese-hubert-base](https://huggingface.co/rinna/japanese-hubert-base) on the MOZILLA-FOUNDATION/COMMON_VOICE_13_0 - JA dataset.
|
| 39 |
It achieves the following results on the evaluation set:
|
| 40 |
+
- Loss: 0.4214
|
| 41 |
+
- Wer: 0.9845
|
| 42 |
+
- Cer: 0.1934
|
| 43 |
|
| 44 |
## Model description
|
| 45 |
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_cer": 0.1934020965549006,
|
| 4 |
+
"eval_loss": 0.42138829827308655,
|
| 5 |
+
"eval_runtime": 214.0793,
|
| 6 |
+
"eval_samples": 4959,
|
| 7 |
+
"eval_samples_per_second": 23.164,
|
| 8 |
+
"eval_steps_per_second": 2.896,
|
| 9 |
+
"eval_wer": 0.9844573966139328,
|
| 10 |
+
"total_flos": 1.5765326864839811e+19,
|
| 11 |
+
"train_loss": 0.910362657587579,
|
| 12 |
+
"train_runtime": 61619.102,
|
| 13 |
+
"train_samples": 12031,
|
| 14 |
+
"train_samples_per_second": 5.857,
|
| 15 |
+
"train_steps_per_second": 0.183
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_cer": 0.1934020965549006,
|
| 4 |
+
"eval_loss": 0.42138829827308655,
|
| 5 |
+
"eval_runtime": 214.0793,
|
| 6 |
+
"eval_samples": 4959,
|
| 7 |
+
"eval_samples_per_second": 23.164,
|
| 8 |
+
"eval_steps_per_second": 2.896,
|
| 9 |
+
"eval_wer": 0.9844573966139328
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"total_flos": 1.5765326864839811e+19,
|
| 4 |
+
"train_loss": 0.910362657587579,
|
| 5 |
+
"train_runtime": 61619.102,
|
| 6 |
+
"train_samples": 12031,
|
| 7 |
+
"train_samples_per_second": 5.857,
|
| 8 |
+
"train_steps_per_second": 0.183
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3207731246948242,
|
| 3 |
+
"best_model_checkpoint": "./Hubert-common_voice-phonemes-debug/checkpoint-4500",
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 11280,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.26595744680851063,
|
| 13 |
+
"eval_cer": 1.829171076931253,
|
| 14 |
+
"eval_loss": 18.536373138427734,
|
| 15 |
+
"eval_runtime": 213.5728,
|
| 16 |
+
"eval_samples_per_second": 23.219,
|
| 17 |
+
"eval_steps_per_second": 2.903,
|
| 18 |
+
"eval_wer": 1.0645295587010823,
|
| 19 |
+
"step": 100
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.5319148936170213,
|
| 23 |
+
"eval_cer": 0.9812975256754901,
|
| 24 |
+
"eval_loss": 8.279097557067871,
|
| 25 |
+
"eval_runtime": 218.6029,
|
| 26 |
+
"eval_samples_per_second": 22.685,
|
| 27 |
+
"eval_steps_per_second": 2.836,
|
| 28 |
+
"eval_wer": 1.0,
|
| 29 |
+
"step": 200
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.7978723404255319,
|
| 33 |
+
"eval_cer": 0.9813063831106427,
|
| 34 |
+
"eval_loss": 7.022369384765625,
|
| 35 |
+
"eval_runtime": 215.2791,
|
| 36 |
+
"eval_samples_per_second": 23.035,
|
| 37 |
+
"eval_steps_per_second": 2.88,
|
| 38 |
+
"eval_wer": 1.0,
|
| 39 |
+
"step": 300
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.0638297872340425,
|
| 43 |
+
"eval_cer": 0.9813019543930664,
|
| 44 |
+
"eval_loss": 6.310593128204346,
|
| 45 |
+
"eval_runtime": 219.0435,
|
| 46 |
+
"eval_samples_per_second": 22.639,
|
| 47 |
+
"eval_steps_per_second": 2.83,
|
| 48 |
+
"eval_wer": 1.0,
|
| 49 |
+
"step": 400
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 1.3297872340425532,
|
| 53 |
+
"grad_norm": 28.102890014648438,
|
| 54 |
+
"learning_rate": 1.1903999999999998e-05,
|
| 55 |
+
"loss": 8.9892,
|
| 56 |
+
"step": 500
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 1.3297872340425532,
|
| 60 |
+
"eval_cer": 0.9813063831106427,
|
| 61 |
+
"eval_loss": 5.522346496582031,
|
| 62 |
+
"eval_runtime": 213.327,
|
| 63 |
+
"eval_samples_per_second": 23.246,
|
| 64 |
+
"eval_steps_per_second": 2.906,
|
| 65 |
+
"eval_wer": 1.0,
|
| 66 |
+
"step": 500
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 1.5957446808510638,
|
| 70 |
+
"eval_cer": 0.9813108118282189,
|
| 71 |
+
"eval_loss": 4.712064266204834,
|
| 72 |
+
"eval_runtime": 215.4766,
|
| 73 |
+
"eval_samples_per_second": 23.014,
|
| 74 |
+
"eval_steps_per_second": 2.877,
|
| 75 |
+
"eval_wer": 1.0,
|
| 76 |
+
"step": 600
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 1.8617021276595744,
|
| 80 |
+
"eval_cer": 0.9812975256754901,
|
| 81 |
+
"eval_loss": 4.002798080444336,
|
| 82 |
+
"eval_runtime": 215.7516,
|
| 83 |
+
"eval_samples_per_second": 22.985,
|
| 84 |
+
"eval_steps_per_second": 2.874,
|
| 85 |
+
"eval_wer": 1.0,
|
| 86 |
+
"step": 700
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.127659574468085,
|
| 90 |
+
"eval_cer": 0.9813108118282189,
|
| 91 |
+
"eval_loss": 3.475456953048706,
|
| 92 |
+
"eval_runtime": 214.2882,
|
| 93 |
+
"eval_samples_per_second": 23.142,
|
| 94 |
+
"eval_steps_per_second": 2.893,
|
| 95 |
+
"eval_wer": 1.0,
|
| 96 |
+
"step": 800
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 2.393617021276596,
|
| 100 |
+
"eval_cer": 0.9813019543930664,
|
| 101 |
+
"eval_loss": 3.1987791061401367,
|
| 102 |
+
"eval_runtime": 216.1589,
|
| 103 |
+
"eval_samples_per_second": 22.941,
|
| 104 |
+
"eval_steps_per_second": 2.868,
|
| 105 |
+
"eval_wer": 1.0,
|
| 106 |
+
"step": 900
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 2.6595744680851063,
|
| 110 |
+
"grad_norm": 2.371589422225952,
|
| 111 |
+
"learning_rate": 2.3903999999999997e-05,
|
| 112 |
+
"loss": 3.7187,
|
| 113 |
+
"step": 1000
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 2.6595744680851063,
|
| 117 |
+
"eval_cer": 0.9813108118282189,
|
| 118 |
+
"eval_loss": 3.0792152881622314,
|
| 119 |
+
"eval_runtime": 213.8516,
|
| 120 |
+
"eval_samples_per_second": 23.189,
|
| 121 |
+
"eval_steps_per_second": 2.899,
|
| 122 |
+
"eval_wer": 1.0,
|
| 123 |
+
"step": 1000
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 2.925531914893617,
|
| 127 |
+
"eval_cer": 0.9812930969579139,
|
| 128 |
+
"eval_loss": 3.0459136962890625,
|
| 129 |
+
"eval_runtime": 216.2299,
|
| 130 |
+
"eval_samples_per_second": 22.934,
|
| 131 |
+
"eval_steps_per_second": 2.867,
|
| 132 |
+
"eval_wer": 1.0,
|
| 133 |
+
"step": 1100
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 3.1914893617021276,
|
| 137 |
+
"eval_cer": 0.9813152405457951,
|
| 138 |
+
"eval_loss": 3.035966396331787,
|
| 139 |
+
"eval_runtime": 217.4243,
|
| 140 |
+
"eval_samples_per_second": 22.808,
|
| 141 |
+
"eval_steps_per_second": 2.852,
|
| 142 |
+
"eval_wer": 1.0,
|
| 143 |
+
"step": 1200
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 3.4574468085106385,
|
| 147 |
+
"eval_cer": 0.9813108118282189,
|
| 148 |
+
"eval_loss": 3.008437395095825,
|
| 149 |
+
"eval_runtime": 215.1294,
|
| 150 |
+
"eval_samples_per_second": 23.051,
|
| 151 |
+
"eval_steps_per_second": 2.882,
|
| 152 |
+
"eval_wer": 1.0,
|
| 153 |
+
"step": 1300
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 3.723404255319149,
|
| 157 |
+
"eval_cer": 0.9343265470617673,
|
| 158 |
+
"eval_loss": 2.495579481124878,
|
| 159 |
+
"eval_runtime": 217.105,
|
| 160 |
+
"eval_samples_per_second": 22.841,
|
| 161 |
+
"eval_steps_per_second": 2.856,
|
| 162 |
+
"eval_wer": 1.0,
|
| 163 |
+
"step": 1400
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 3.9893617021276597,
|
| 167 |
+
"grad_norm": 3.707914352416992,
|
| 168 |
+
"learning_rate": 3.5903999999999994e-05,
|
| 169 |
+
"loss": 2.783,
|
| 170 |
+
"step": 1500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 3.9893617021276597,
|
| 174 |
+
"eval_cer": 0.33314142223836246,
|
| 175 |
+
"eval_loss": 1.4417579174041748,
|
| 176 |
+
"eval_runtime": 217.5545,
|
| 177 |
+
"eval_samples_per_second": 22.794,
|
| 178 |
+
"eval_steps_per_second": 2.85,
|
| 179 |
+
"eval_wer": 1.0,
|
| 180 |
+
"step": 1500
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 4.25531914893617,
|
| 184 |
+
"eval_cer": 0.27530679941009484,
|
| 185 |
+
"eval_loss": 1.0227618217468262,
|
| 186 |
+
"eval_runtime": 217.2546,
|
| 187 |
+
"eval_samples_per_second": 22.826,
|
| 188 |
+
"eval_steps_per_second": 2.854,
|
| 189 |
+
"eval_wer": 1.0,
|
| 190 |
+
"step": 1600
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 4.5212765957446805,
|
| 194 |
+
"eval_cer": 0.2531632115288376,
|
| 195 |
+
"eval_loss": 0.8218135237693787,
|
| 196 |
+
"eval_runtime": 219.2044,
|
| 197 |
+
"eval_samples_per_second": 22.623,
|
| 198 |
+
"eval_steps_per_second": 2.828,
|
| 199 |
+
"eval_wer": 1.0,
|
| 200 |
+
"step": 1700
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 4.787234042553192,
|
| 204 |
+
"eval_cer": 0.24330045748652562,
|
| 205 |
+
"eval_loss": 0.7084201574325562,
|
| 206 |
+
"eval_runtime": 216.8198,
|
| 207 |
+
"eval_samples_per_second": 22.872,
|
| 208 |
+
"eval_steps_per_second": 2.86,
|
| 209 |
+
"eval_wer": 1.0,
|
| 210 |
+
"step": 1800
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 5.053191489361702,
|
| 214 |
+
"eval_cer": 0.23373885623939875,
|
| 215 |
+
"eval_loss": 0.6305696368217468,
|
| 216 |
+
"eval_runtime": 218.0682,
|
| 217 |
+
"eval_samples_per_second": 22.741,
|
| 218 |
+
"eval_steps_per_second": 2.843,
|
| 219 |
+
"eval_wer": 1.0,
|
| 220 |
+
"step": 1900
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 5.319148936170213,
|
| 224 |
+
"grad_norm": 2.717346429824829,
|
| 225 |
+
"learning_rate": 4.7903999999999994e-05,
|
| 226 |
+
"loss": 0.8659,
|
| 227 |
+
"step": 2000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 5.319148936170213,
|
| 231 |
+
"eval_cer": 0.23099305134212286,
|
| 232 |
+
"eval_loss": 0.5933591723442078,
|
| 233 |
+
"eval_runtime": 218.7287,
|
| 234 |
+
"eval_samples_per_second": 22.672,
|
| 235 |
+
"eval_steps_per_second": 2.835,
|
| 236 |
+
"eval_wer": 1.0,
|
| 237 |
+
"step": 2000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 5.585106382978723,
|
| 241 |
+
"eval_cer": 0.228406680277592,
|
| 242 |
+
"eval_loss": 0.5647820234298706,
|
| 243 |
+
"eval_runtime": 215.722,
|
| 244 |
+
"eval_samples_per_second": 22.988,
|
| 245 |
+
"eval_steps_per_second": 2.874,
|
| 246 |
+
"eval_wer": 1.0,
|
| 247 |
+
"step": 2100
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 5.851063829787234,
|
| 251 |
+
"eval_cer": 0.22140930650711474,
|
| 252 |
+
"eval_loss": 0.5329573750495911,
|
| 253 |
+
"eval_runtime": 218.357,
|
| 254 |
+
"eval_samples_per_second": 22.711,
|
| 255 |
+
"eval_steps_per_second": 2.839,
|
| 256 |
+
"eval_wer": 1.0,
|
| 257 |
+
"step": 2200
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 6.117021276595745,
|
| 261 |
+
"eval_cer": 0.22092214757372708,
|
| 262 |
+
"eval_loss": 0.5139152407646179,
|
| 263 |
+
"eval_runtime": 217.5924,
|
| 264 |
+
"eval_samples_per_second": 22.79,
|
| 265 |
+
"eval_steps_per_second": 2.849,
|
| 266 |
+
"eval_wer": 1.0,
|
| 267 |
+
"step": 2300
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 6.382978723404255,
|
| 271 |
+
"eval_cer": 0.21591769671256295,
|
| 272 |
+
"eval_loss": 0.4906846582889557,
|
| 273 |
+
"eval_runtime": 217.5534,
|
| 274 |
+
"eval_samples_per_second": 22.794,
|
| 275 |
+
"eval_steps_per_second": 2.85,
|
| 276 |
+
"eval_wer": 1.0,
|
| 277 |
+
"step": 2400
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 6.648936170212766,
|
| 281 |
+
"grad_norm": 6.231712341308594,
|
| 282 |
+
"learning_rate": 5.9903999999999994e-05,
|
| 283 |
+
"loss": 0.5271,
|
| 284 |
+
"step": 2500
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 6.648936170212766,
|
| 288 |
+
"eval_cer": 0.2159929849113592,
|
| 289 |
+
"eval_loss": 0.46404561400413513,
|
| 290 |
+
"eval_runtime": 217.7416,
|
| 291 |
+
"eval_samples_per_second": 22.775,
|
| 292 |
+
"eval_steps_per_second": 2.847,
|
| 293 |
+
"eval_wer": 1.0,
|
| 294 |
+
"step": 2500
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 6.914893617021277,
|
| 298 |
+
"eval_cer": 0.21120111249385515,
|
| 299 |
+
"eval_loss": 0.4608772397041321,
|
| 300 |
+
"eval_runtime": 216.2932,
|
| 301 |
+
"eval_samples_per_second": 22.927,
|
| 302 |
+
"eval_steps_per_second": 2.866,
|
| 303 |
+
"eval_wer": 1.0,
|
| 304 |
+
"step": 2600
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 7.180851063829787,
|
| 308 |
+
"eval_cer": 0.2096732049300484,
|
| 309 |
+
"eval_loss": 0.4550062417984009,
|
| 310 |
+
"eval_runtime": 219.5448,
|
| 311 |
+
"eval_samples_per_second": 22.588,
|
| 312 |
+
"eval_steps_per_second": 2.824,
|
| 313 |
+
"eval_wer": 1.0001387732445184,
|
| 314 |
+
"step": 2700
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 7.446808510638298,
|
| 318 |
+
"eval_cer": 0.210000930030691,
|
| 319 |
+
"eval_loss": 0.4601040184497833,
|
| 320 |
+
"eval_runtime": 218.3231,
|
| 321 |
+
"eval_samples_per_second": 22.714,
|
| 322 |
+
"eval_steps_per_second": 2.84,
|
| 323 |
+
"eval_wer": 0.9991673605328892,
|
| 324 |
+
"step": 2800
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 7.712765957446808,
|
| 328 |
+
"eval_cer": 0.20508948223862816,
|
| 329 |
+
"eval_loss": 0.4290345013141632,
|
| 330 |
+
"eval_runtime": 218.312,
|
| 331 |
+
"eval_samples_per_second": 22.715,
|
| 332 |
+
"eval_steps_per_second": 2.84,
|
| 333 |
+
"eval_wer": 0.9952817096863724,
|
| 334 |
+
"step": 2900
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 7.9787234042553195,
|
| 338 |
+
"grad_norm": 2.573235273361206,
|
| 339 |
+
"learning_rate": 7.1904e-05,
|
| 340 |
+
"loss": 0.4244,
|
| 341 |
+
"step": 3000
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 7.9787234042553195,
|
| 345 |
+
"eval_cer": 0.20240125066984355,
|
| 346 |
+
"eval_loss": 0.42562517523765564,
|
| 347 |
+
"eval_runtime": 219.9498,
|
| 348 |
+
"eval_samples_per_second": 22.546,
|
| 349 |
+
"eval_steps_per_second": 2.819,
|
| 350 |
+
"eval_wer": 0.9970857618651124,
|
| 351 |
+
"step": 3000
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 8.24468085106383,
|
| 355 |
+
"eval_cer": 0.2014180753679157,
|
| 356 |
+
"eval_loss": 0.4134505093097687,
|
| 357 |
+
"eval_runtime": 218.0911,
|
| 358 |
+
"eval_samples_per_second": 22.738,
|
| 359 |
+
"eval_steps_per_second": 2.843,
|
| 360 |
+
"eval_wer": 0.9998612267554815,
|
| 361 |
+
"step": 3100
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 8.51063829787234,
|
| 365 |
+
"eval_cer": 0.1998945965216852,
|
| 366 |
+
"eval_loss": 0.41247421503067017,
|
| 367 |
+
"eval_runtime": 220.4438,
|
| 368 |
+
"eval_samples_per_second": 22.496,
|
| 369 |
+
"eval_steps_per_second": 2.813,
|
| 370 |
+
"eval_wer": 0.9955592561754094,
|
| 371 |
+
"step": 3200
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 8.77659574468085,
|
| 375 |
+
"eval_cer": 0.1926979304602766,
|
| 376 |
+
"eval_loss": 0.3886178731918335,
|
| 377 |
+
"eval_runtime": 219.8802,
|
| 378 |
+
"eval_samples_per_second": 22.553,
|
| 379 |
+
"eval_steps_per_second": 2.82,
|
| 380 |
+
"eval_wer": 0.9941715237302248,
|
| 381 |
+
"step": 3300
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 9.042553191489361,
|
| 385 |
+
"eval_cer": 0.1911124495679786,
|
| 386 |
+
"eval_loss": 0.38326719403266907,
|
| 387 |
+
"eval_runtime": 219.6869,
|
| 388 |
+
"eval_samples_per_second": 22.573,
|
| 389 |
+
"eval_steps_per_second": 2.822,
|
| 390 |
+
"eval_wer": 1.0005550929780738,
|
| 391 |
+
"step": 3400
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 9.308510638297872,
|
| 395 |
+
"grad_norm": 2.091050148010254,
|
| 396 |
+
"learning_rate": 8.390399999999999e-05,
|
| 397 |
+
"loss": 0.3373,
|
| 398 |
+
"step": 3500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 9.308510638297872,
|
| 402 |
+
"eval_cer": 0.1886855123361928,
|
| 403 |
+
"eval_loss": 0.36107370257377625,
|
| 404 |
+
"eval_runtime": 222.4868,
|
| 405 |
+
"eval_samples_per_second": 22.289,
|
| 406 |
+
"eval_steps_per_second": 2.787,
|
| 407 |
+
"eval_wer": 1.0363585900638357,
|
| 408 |
+
"step": 3500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 9.574468085106384,
|
| 412 |
+
"eval_cer": 0.18428336706539888,
|
| 413 |
+
"eval_loss": 0.3584689497947693,
|
| 414 |
+
"eval_runtime": 217.5818,
|
| 415 |
+
"eval_samples_per_second": 22.791,
|
| 416 |
+
"eval_steps_per_second": 2.85,
|
| 417 |
+
"eval_wer": 1.0080488481820704,
|
| 418 |
+
"step": 3600
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 9.840425531914894,
|
| 422 |
+
"eval_cer": 0.18549683568129177,
|
| 423 |
+
"eval_loss": 0.3562237024307251,
|
| 424 |
+
"eval_runtime": 221.8447,
|
| 425 |
+
"eval_samples_per_second": 22.353,
|
| 426 |
+
"eval_steps_per_second": 2.795,
|
| 427 |
+
"eval_wer": 0.9980571745767416,
|
| 428 |
+
"step": 3700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 10.106382978723405,
|
| 432 |
+
"eval_cer": 0.1799387951230962,
|
| 433 |
+
"eval_loss": 0.3412463068962097,
|
| 434 |
+
"eval_runtime": 217.8756,
|
| 435 |
+
"eval_samples_per_second": 22.761,
|
| 436 |
+
"eval_steps_per_second": 2.846,
|
| 437 |
+
"eval_wer": 0.9883430474604497,
|
| 438 |
+
"step": 3800
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 10.372340425531915,
|
| 442 |
+
"eval_cer": 0.18462437831877024,
|
| 443 |
+
"eval_loss": 0.35612642765045166,
|
| 444 |
+
"eval_runtime": 218.9988,
|
| 445 |
+
"eval_samples_per_second": 22.644,
|
| 446 |
+
"eval_steps_per_second": 2.831,
|
| 447 |
+
"eval_wer": 0.9834859839023037,
|
| 448 |
+
"step": 3900
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 10.638297872340425,
|
| 452 |
+
"grad_norm": 2.690404176712036,
|
| 453 |
+
"learning_rate": 9.5904e-05,
|
| 454 |
+
"loss": 0.2779,
|
| 455 |
+
"step": 4000
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 10.638297872340425,
|
| 459 |
+
"eval_cer": 0.17977050385519866,
|
| 460 |
+
"eval_loss": 0.34815147519111633,
|
| 461 |
+
"eval_runtime": 221.1151,
|
| 462 |
+
"eval_samples_per_second": 22.427,
|
| 463 |
+
"eval_steps_per_second": 2.804,
|
| 464 |
+
"eval_wer": 0.9772411878989731,
|
| 465 |
+
"step": 4000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 10.904255319148936,
|
| 469 |
+
"eval_cer": 0.17927891620423475,
|
| 470 |
+
"eval_loss": 0.3266104757785797,
|
| 471 |
+
"eval_runtime": 217.4603,
|
| 472 |
+
"eval_samples_per_second": 22.804,
|
| 473 |
+
"eval_steps_per_second": 2.851,
|
| 474 |
+
"eval_wer": 0.9794615598112684,
|
| 475 |
+
"step": 4100
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 11.170212765957446,
|
| 479 |
+
"eval_cer": 0.17892019008055837,
|
| 480 |
+
"eval_loss": 0.34835320711135864,
|
| 481 |
+
"eval_runtime": 220.4292,
|
| 482 |
+
"eval_samples_per_second": 22.497,
|
| 483 |
+
"eval_steps_per_second": 2.813,
|
| 484 |
+
"eval_wer": 0.9791840133222315,
|
| 485 |
+
"step": 4200
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 11.436170212765958,
|
| 489 |
+
"eval_cer": 0.17989007922975744,
|
| 490 |
+
"eval_loss": 0.33775731921195984,
|
| 491 |
+
"eval_runtime": 218.1639,
|
| 492 |
+
"eval_samples_per_second": 22.731,
|
| 493 |
+
"eval_steps_per_second": 2.842,
|
| 494 |
+
"eval_wer": 0.9991673605328892,
|
| 495 |
+
"step": 4300
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 11.702127659574469,
|
| 499 |
+
"eval_cer": 0.17952249567092857,
|
| 500 |
+
"eval_loss": 0.33295243978500366,
|
| 501 |
+
"eval_runtime": 218.1937,
|
| 502 |
+
"eval_samples_per_second": 22.728,
|
| 503 |
+
"eval_steps_per_second": 2.842,
|
| 504 |
+
"eval_wer": 0.9764085484318623,
|
| 505 |
+
"step": 4400
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 11.96808510638298,
|
| 509 |
+
"grad_norm": 2.2231483459472656,
|
| 510 |
+
"learning_rate": 0.00010790399999999999,
|
| 511 |
+
"loss": 0.2409,
|
| 512 |
+
"step": 4500
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 11.96808510638298,
|
| 516 |
+
"eval_cer": 0.17923020031089598,
|
| 517 |
+
"eval_loss": 0.3207731246948242,
|
| 518 |
+
"eval_runtime": 218.6698,
|
| 519 |
+
"eval_samples_per_second": 22.678,
|
| 520 |
+
"eval_steps_per_second": 2.835,
|
| 521 |
+
"eval_wer": 0.9780738273660838,
|
| 522 |
+
"step": 4500
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 12.23404255319149,
|
| 526 |
+
"eval_cer": 0.18048352738497514,
|
| 527 |
+
"eval_loss": 0.3602401614189148,
|
| 528 |
+
"eval_runtime": 217.2746,
|
| 529 |
+
"eval_samples_per_second": 22.824,
|
| 530 |
+
"eval_steps_per_second": 2.854,
|
| 531 |
+
"eval_wer": 0.97571468220927,
|
| 532 |
+
"step": 4600
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 12.5,
|
| 536 |
+
"eval_cer": 0.17876961368296582,
|
| 537 |
+
"eval_loss": 0.3362875282764435,
|
| 538 |
+
"eval_runtime": 219.2626,
|
| 539 |
+
"eval_samples_per_second": 22.617,
|
| 540 |
+
"eval_steps_per_second": 2.828,
|
| 541 |
+
"eval_wer": 0.9938939772411879,
|
| 542 |
+
"step": 4700
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 12.76595744680851,
|
| 546 |
+
"eval_cer": 0.17953135310608107,
|
| 547 |
+
"eval_loss": 0.32531023025512695,
|
| 548 |
+
"eval_runtime": 217.8058,
|
| 549 |
+
"eval_samples_per_second": 22.768,
|
| 550 |
+
"eval_steps_per_second": 2.847,
|
| 551 |
+
"eval_wer": 0.9732167638079379,
|
| 552 |
+
"step": 4800
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 13.03191489361702,
|
| 556 |
+
"eval_cer": 0.17615224159540122,
|
| 557 |
+
"eval_loss": 0.3284818232059479,
|
| 558 |
+
"eval_runtime": 218.6415,
|
| 559 |
+
"eval_samples_per_second": 22.681,
|
| 560 |
+
"eval_steps_per_second": 2.836,
|
| 561 |
+
"eval_wer": 0.971135165140161,
|
| 562 |
+
"step": 4900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 13.297872340425531,
|
| 566 |
+
"grad_norm": 2.2732036113739014,
|
| 567 |
+
"learning_rate": 0.00011990399999999998,
|
| 568 |
+
"loss": 0.2104,
|
| 569 |
+
"step": 5000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 13.297872340425531,
|
| 573 |
+
"eval_cer": 0.17691398101851646,
|
| 574 |
+
"eval_loss": 0.32327184081077576,
|
| 575 |
+
"eval_runtime": 218.7534,
|
| 576 |
+
"eval_samples_per_second": 22.669,
|
| 577 |
+
"eval_steps_per_second": 2.834,
|
| 578 |
+
"eval_wer": 0.9729392173189009,
|
| 579 |
+
"step": 5000
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 13.563829787234042,
|
| 583 |
+
"eval_cer": 0.18268017130279585,
|
| 584 |
+
"eval_loss": 0.3362744450569153,
|
| 585 |
+
"eval_runtime": 216.3239,
|
| 586 |
+
"eval_samples_per_second": 22.924,
|
| 587 |
+
"eval_steps_per_second": 2.866,
|
| 588 |
+
"eval_wer": 0.97751873438801,
|
| 589 |
+
"step": 5100
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 13.829787234042554,
|
| 593 |
+
"eval_cer": 0.17587766110567363,
|
| 594 |
+
"eval_loss": 0.33709821105003357,
|
| 595 |
+
"eval_runtime": 219.8171,
|
| 596 |
+
"eval_samples_per_second": 22.56,
|
| 597 |
+
"eval_steps_per_second": 2.821,
|
| 598 |
+
"eval_wer": 0.9683597002497918,
|
| 599 |
+
"step": 5200
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 14.095744680851064,
|
| 603 |
+
"eval_cer": 0.17777315222830925,
|
| 604 |
+
"eval_loss": 0.3463989496231079,
|
| 605 |
+
"eval_runtime": 216.1817,
|
| 606 |
+
"eval_samples_per_second": 22.939,
|
| 607 |
+
"eval_steps_per_second": 2.868,
|
| 608 |
+
"eval_wer": 0.9730779905634194,
|
| 609 |
+
"step": 5300
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 14.361702127659575,
|
| 613 |
+
"eval_cer": 0.17830459833745943,
|
| 614 |
+
"eval_loss": 0.3450472950935364,
|
| 615 |
+
"eval_runtime": 217.8272,
|
| 616 |
+
"eval_samples_per_second": 22.766,
|
| 617 |
+
"eval_steps_per_second": 2.846,
|
| 618 |
+
"eval_wer": 0.9776575076325285,
|
| 619 |
+
"step": 5400
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 14.627659574468085,
|
| 623 |
+
"grad_norm": 3.842231273651123,
|
| 624 |
+
"learning_rate": 0.000131904,
|
| 625 |
+
"loss": 0.1947,
|
| 626 |
+
"step": 5500
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 14.627659574468085,
|
| 630 |
+
"eval_cer": 0.1773037081652266,
|
| 631 |
+
"eval_loss": 0.3442366421222687,
|
| 632 |
+
"eval_runtime": 219.014,
|
| 633 |
+
"eval_samples_per_second": 22.642,
|
| 634 |
+
"eval_steps_per_second": 2.831,
|
| 635 |
+
"eval_wer": 0.9680821537607549,
|
| 636 |
+
"step": 5500
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 14.893617021276595,
|
| 640 |
+
"eval_cer": 0.17800344554227432,
|
| 641 |
+
"eval_loss": 0.3346150517463684,
|
| 642 |
+
"eval_runtime": 216.342,
|
| 643 |
+
"eval_samples_per_second": 22.922,
|
| 644 |
+
"eval_steps_per_second": 2.866,
|
| 645 |
+
"eval_wer": 0.9858451290591174,
|
| 646 |
+
"step": 5600
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 15.159574468085106,
|
| 650 |
+
"eval_cer": 0.17705569998095652,
|
| 651 |
+
"eval_loss": 0.35239535570144653,
|
| 652 |
+
"eval_runtime": 219.7446,
|
| 653 |
+
"eval_samples_per_second": 22.567,
|
| 654 |
+
"eval_steps_per_second": 2.821,
|
| 655 |
+
"eval_wer": 0.9732167638079379,
|
| 656 |
+
"step": 5700
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 15.425531914893616,
|
| 660 |
+
"eval_cer": 0.17744542712766664,
|
| 661 |
+
"eval_loss": 0.3413674235343933,
|
| 662 |
+
"eval_runtime": 216.7948,
|
| 663 |
+
"eval_samples_per_second": 22.874,
|
| 664 |
+
"eval_steps_per_second": 2.86,
|
| 665 |
+
"eval_wer": 0.9782126006106022,
|
| 666 |
+
"step": 5800
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 15.691489361702128,
|
| 670 |
+
"eval_cer": 0.1765641123299926,
|
| 671 |
+
"eval_loss": 0.3437710702419281,
|
| 672 |
+
"eval_runtime": 217.8111,
|
| 673 |
+
"eval_samples_per_second": 22.767,
|
| 674 |
+
"eval_steps_per_second": 2.847,
|
| 675 |
+
"eval_wer": 1.0019428254232583,
|
| 676 |
+
"step": 5900
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 15.957446808510639,
|
| 680 |
+
"grad_norm": 1.6587754487991333,
|
| 681 |
+
"learning_rate": 0.00014390399999999998,
|
| 682 |
+
"loss": 0.1892,
|
| 683 |
+
"step": 6000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 15.957446808510639,
|
| 687 |
+
"eval_cer": 0.18023551920070505,
|
| 688 |
+
"eval_loss": 0.3391115367412567,
|
| 689 |
+
"eval_runtime": 220.8887,
|
| 690 |
+
"eval_samples_per_second": 22.45,
|
| 691 |
+
"eval_steps_per_second": 2.807,
|
| 692 |
+
"eval_wer": 0.9705800721620872,
|
| 693 |
+
"step": 6000
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"epoch": 16.22340425531915,
|
| 697 |
+
"eval_cer": 0.18025766278858632,
|
| 698 |
+
"eval_loss": 0.35049712657928467,
|
| 699 |
+
"eval_runtime": 218.4181,
|
| 700 |
+
"eval_samples_per_second": 22.704,
|
| 701 |
+
"eval_steps_per_second": 2.839,
|
| 702 |
+
"eval_wer": 0.9782126006106022,
|
| 703 |
+
"step": 6100
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 16.48936170212766,
|
| 707 |
+
"eval_cer": 0.17667040155182265,
|
| 708 |
+
"eval_loss": 0.34669336676597595,
|
| 709 |
+
"eval_runtime": 220.5862,
|
| 710 |
+
"eval_samples_per_second": 22.481,
|
| 711 |
+
"eval_steps_per_second": 2.811,
|
| 712 |
+
"eval_wer": 0.9736330835414932,
|
| 713 |
+
"step": 6200
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 16.75531914893617,
|
| 717 |
+
"eval_cer": 0.17923462902847223,
|
| 718 |
+
"eval_loss": 0.36808422207832336,
|
| 719 |
+
"eval_runtime": 219.3778,
|
| 720 |
+
"eval_samples_per_second": 22.605,
|
| 721 |
+
"eval_steps_per_second": 2.826,
|
| 722 |
+
"eval_wer": 0.9945878434637802,
|
| 723 |
+
"step": 6300
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 17.02127659574468,
|
| 727 |
+
"eval_cer": 0.17693612460639774,
|
| 728 |
+
"eval_loss": 0.35571783781051636,
|
| 729 |
+
"eval_runtime": 218.8646,
|
| 730 |
+
"eval_samples_per_second": 22.658,
|
| 731 |
+
"eval_steps_per_second": 2.833,
|
| 732 |
+
"eval_wer": 1.0104079933388843,
|
| 733 |
+
"step": 6400
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 17.28723404255319,
|
| 737 |
+
"grad_norm": 1.837514877319336,
|
| 738 |
+
"learning_rate": 0.000155904,
|
| 739 |
+
"loss": 0.1749,
|
| 740 |
+
"step": 6500
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 17.28723404255319,
|
| 744 |
+
"eval_cer": 0.17874747009508457,
|
| 745 |
+
"eval_loss": 0.3445756435394287,
|
| 746 |
+
"eval_runtime": 220.0286,
|
| 747 |
+
"eval_samples_per_second": 22.538,
|
| 748 |
+
"eval_steps_per_second": 2.818,
|
| 749 |
+
"eval_wer": 0.9769636414099362,
|
| 750 |
+
"step": 6500
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 17.5531914893617,
|
| 754 |
+
"eval_cer": 0.18034180842253508,
|
| 755 |
+
"eval_loss": 0.34961050748825073,
|
| 756 |
+
"eval_runtime": 216.6863,
|
| 757 |
+
"eval_samples_per_second": 22.886,
|
| 758 |
+
"eval_steps_per_second": 2.861,
|
| 759 |
+
"eval_wer": 0.983902303635859,
|
| 760 |
+
"step": 6600
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 17.819148936170212,
|
| 764 |
+
"eval_cer": 0.18056324430134765,
|
| 765 |
+
"eval_loss": 0.3584659695625305,
|
| 766 |
+
"eval_runtime": 217.6982,
|
| 767 |
+
"eval_samples_per_second": 22.779,
|
| 768 |
+
"eval_steps_per_second": 2.848,
|
| 769 |
+
"eval_wer": 1.0012489592006661,
|
| 770 |
+
"step": 6700
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 18.085106382978722,
|
| 774 |
+
"eval_cer": 0.17987236435945245,
|
| 775 |
+
"eval_loss": 0.35618793964385986,
|
| 776 |
+
"eval_runtime": 217.2896,
|
| 777 |
+
"eval_samples_per_second": 22.822,
|
| 778 |
+
"eval_steps_per_second": 2.853,
|
| 779 |
+
"eval_wer": 0.9716902581182348,
|
| 780 |
+
"step": 6800
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 18.351063829787233,
|
| 784 |
+
"eval_cer": 0.18354819994774113,
|
| 785 |
+
"eval_loss": 0.3722291588783264,
|
| 786 |
+
"eval_runtime": 217.7923,
|
| 787 |
+
"eval_samples_per_second": 22.769,
|
| 788 |
+
"eval_steps_per_second": 2.847,
|
| 789 |
+
"eval_wer": 1.0503746877601998,
|
| 790 |
+
"step": 6900
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 18.617021276595743,
|
| 794 |
+
"grad_norm": 5.107634544372559,
|
| 795 |
+
"learning_rate": 0.00016790399999999997,
|
| 796 |
+
"loss": 0.1717,
|
| 797 |
+
"step": 7000
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 18.617021276595743,
|
| 801 |
+
"eval_cer": 0.18089982683714276,
|
| 802 |
+
"eval_loss": 0.3554377853870392,
|
| 803 |
+
"eval_runtime": 218.5289,
|
| 804 |
+
"eval_samples_per_second": 22.693,
|
| 805 |
+
"eval_steps_per_second": 2.837,
|
| 806 |
+
"eval_wer": 0.9772411878989731,
|
| 807 |
+
"step": 7000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 18.882978723404257,
|
| 811 |
+
"eval_cer": 0.1788094721411521,
|
| 812 |
+
"eval_loss": 0.36780402064323425,
|
| 813 |
+
"eval_runtime": 216.2085,
|
| 814 |
+
"eval_samples_per_second": 22.936,
|
| 815 |
+
"eval_steps_per_second": 2.868,
|
| 816 |
+
"eval_wer": 0.9683597002497918,
|
| 817 |
+
"step": 7100
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 19.148936170212767,
|
| 821 |
+
"eval_cer": 0.18539497517703799,
|
| 822 |
+
"eval_loss": 0.49383628368377686,
|
| 823 |
+
"eval_runtime": 219.4656,
|
| 824 |
+
"eval_samples_per_second": 22.596,
|
| 825 |
+
"eval_steps_per_second": 2.825,
|
| 826 |
+
"eval_wer": 1.041909519844574,
|
| 827 |
+
"step": 7200
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 19.414893617021278,
|
| 831 |
+
"eval_cer": 0.18054552943104266,
|
| 832 |
+
"eval_loss": 0.3926149308681488,
|
| 833 |
+
"eval_runtime": 216.1405,
|
| 834 |
+
"eval_samples_per_second": 22.943,
|
| 835 |
+
"eval_steps_per_second": 2.869,
|
| 836 |
+
"eval_wer": 0.9826533444351929,
|
| 837 |
+
"step": 7300
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 19.680851063829788,
|
| 841 |
+
"eval_cer": 0.1819007170093756,
|
| 842 |
+
"eval_loss": 0.3581140339374542,
|
| 843 |
+
"eval_runtime": 217.0297,
|
| 844 |
+
"eval_samples_per_second": 22.849,
|
| 845 |
+
"eval_steps_per_second": 2.857,
|
| 846 |
+
"eval_wer": 1.0001387732445184,
|
| 847 |
+
"step": 7400
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 19.9468085106383,
|
| 851 |
+
"grad_norm": 1.7489858865737915,
|
| 852 |
+
"learning_rate": 0.000179904,
|
| 853 |
+
"loss": 0.1715,
|
| 854 |
+
"step": 7500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 19.9468085106383,
|
| 858 |
+
"eval_cer": 0.18397778555263752,
|
| 859 |
+
"eval_loss": 0.35685011744499207,
|
| 860 |
+
"eval_runtime": 219.4728,
|
| 861 |
+
"eval_samples_per_second": 22.595,
|
| 862 |
+
"eval_steps_per_second": 2.825,
|
| 863 |
+
"eval_wer": 0.9929225645295587,
|
| 864 |
+
"step": 7500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 20.21276595744681,
|
| 868 |
+
"eval_cer": 0.18136927090022542,
|
| 869 |
+
"eval_loss": 0.3910757303237915,
|
| 870 |
+
"eval_runtime": 217.6276,
|
| 871 |
+
"eval_samples_per_second": 22.787,
|
| 872 |
+
"eval_steps_per_second": 2.849,
|
| 873 |
+
"eval_wer": 0.9969469886205939,
|
| 874 |
+
"step": 7600
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 20.47872340425532,
|
| 878 |
+
"eval_cer": 0.18082896735592274,
|
| 879 |
+
"eval_loss": 0.39733073115348816,
|
| 880 |
+
"eval_runtime": 218.1385,
|
| 881 |
+
"eval_samples_per_second": 22.733,
|
| 882 |
+
"eval_steps_per_second": 2.842,
|
| 883 |
+
"eval_wer": 1.0016652789342215,
|
| 884 |
+
"step": 7700
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 20.74468085106383,
|
| 888 |
+
"eval_cer": 0.18392906965929876,
|
| 889 |
+
"eval_loss": 0.3943112790584564,
|
| 890 |
+
"eval_runtime": 217.3912,
|
| 891 |
+
"eval_samples_per_second": 22.811,
|
| 892 |
+
"eval_steps_per_second": 2.852,
|
| 893 |
+
"eval_wer": 0.9723841243408271,
|
| 894 |
+
"step": 7800
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 21.01063829787234,
|
| 898 |
+
"eval_cer": 0.18227715800335698,
|
| 899 |
+
"eval_loss": 0.3984341025352478,
|
| 900 |
+
"eval_runtime": 216.5247,
|
| 901 |
+
"eval_samples_per_second": 22.903,
|
| 902 |
+
"eval_steps_per_second": 2.863,
|
| 903 |
+
"eval_wer": 0.9764085484318623,
|
| 904 |
+
"step": 7900
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 21.27659574468085,
|
| 908 |
+
"grad_norm": 1.9202663898468018,
|
| 909 |
+
"learning_rate": 0.00019190399999999998,
|
| 910 |
+
"loss": 0.1667,
|
| 911 |
+
"step": 8000
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 21.27659574468085,
|
| 915 |
+
"eval_cer": 0.18395121324718,
|
| 916 |
+
"eval_loss": 0.4305620491504669,
|
| 917 |
+
"eval_runtime": 219.6412,
|
| 918 |
+
"eval_samples_per_second": 22.578,
|
| 919 |
+
"eval_steps_per_second": 2.823,
|
| 920 |
+
"eval_wer": 1.0499583680266444,
|
| 921 |
+
"step": 8000
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 21.54255319148936,
|
| 925 |
+
"eval_cer": 0.1881762098149239,
|
| 926 |
+
"eval_loss": 0.37944892048835754,
|
| 927 |
+
"eval_runtime": 214.5629,
|
| 928 |
+
"eval_samples_per_second": 23.112,
|
| 929 |
+
"eval_steps_per_second": 2.89,
|
| 930 |
+
"eval_wer": 0.9728004440743825,
|
| 931 |
+
"step": 8100
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 21.80851063829787,
|
| 935 |
+
"eval_cer": 0.18339762355014858,
|
| 936 |
+
"eval_loss": 0.3965916037559509,
|
| 937 |
+
"eval_runtime": 214.8413,
|
| 938 |
+
"eval_samples_per_second": 23.082,
|
| 939 |
+
"eval_steps_per_second": 2.886,
|
| 940 |
+
"eval_wer": 0.9912572855953372,
|
| 941 |
+
"step": 8200
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 22.074468085106382,
|
| 945 |
+
"eval_cer": 0.18381392300231622,
|
| 946 |
+
"eval_loss": 0.39811334013938904,
|
| 947 |
+
"eval_runtime": 215.0561,
|
| 948 |
+
"eval_samples_per_second": 23.059,
|
| 949 |
+
"eval_steps_per_second": 2.883,
|
| 950 |
+
"eval_wer": 0.9744657230086039,
|
| 951 |
+
"step": 8300
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 22.340425531914892,
|
| 955 |
+
"eval_cer": 0.18262702669188083,
|
| 956 |
+
"eval_loss": 0.4328286349773407,
|
| 957 |
+
"eval_runtime": 214.1041,
|
| 958 |
+
"eval_samples_per_second": 23.162,
|
| 959 |
+
"eval_steps_per_second": 2.896,
|
| 960 |
+
"eval_wer": 0.9926450180405217,
|
| 961 |
+
"step": 8400
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 22.606382978723403,
|
| 965 |
+
"grad_norm": 2.1760590076446533,
|
| 966 |
+
"learning_rate": 0.00020387999999999998,
|
| 967 |
+
"loss": 0.1625,
|
| 968 |
+
"step": 8500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 22.606382978723403,
|
| 972 |
+
"eval_cer": 0.18350834148955486,
|
| 973 |
+
"eval_loss": 0.4087058901786804,
|
| 974 |
+
"eval_runtime": 215.1826,
|
| 975 |
+
"eval_samples_per_second": 23.046,
|
| 976 |
+
"eval_steps_per_second": 2.881,
|
| 977 |
+
"eval_wer": 0.9709963918956426,
|
| 978 |
+
"step": 8500
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 22.872340425531917,
|
| 982 |
+
"eval_cer": 0.18611685614196696,
|
| 983 |
+
"eval_loss": 0.4149414300918579,
|
| 984 |
+
"eval_runtime": 214.9057,
|
| 985 |
+
"eval_samples_per_second": 23.075,
|
| 986 |
+
"eval_steps_per_second": 2.885,
|
| 987 |
+
"eval_wer": 1.0062447960033305,
|
| 988 |
+
"step": 8600
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 23.138297872340427,
|
| 992 |
+
"eval_cer": 0.18750304474333368,
|
| 993 |
+
"eval_loss": 0.4106564223766327,
|
| 994 |
+
"eval_runtime": 214.2659,
|
| 995 |
+
"eval_samples_per_second": 23.144,
|
| 996 |
+
"eval_steps_per_second": 2.894,
|
| 997 |
+
"eval_wer": 0.992089925062448,
|
| 998 |
+
"step": 8700
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 23.404255319148938,
|
| 1002 |
+
"eval_cer": 0.18692731145842098,
|
| 1003 |
+
"eval_loss": 0.41398727893829346,
|
| 1004 |
+
"eval_runtime": 214.1566,
|
| 1005 |
+
"eval_samples_per_second": 23.156,
|
| 1006 |
+
"eval_steps_per_second": 2.895,
|
| 1007 |
+
"eval_wer": 0.9834859839023037,
|
| 1008 |
+
"step": 8800
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 23.670212765957448,
|
| 1012 |
+
"eval_cer": 0.18904423845986917,
|
| 1013 |
+
"eval_loss": 0.40873560309410095,
|
| 1014 |
+
"eval_runtime": 213.9242,
|
| 1015 |
+
"eval_samples_per_second": 23.181,
|
| 1016 |
+
"eval_steps_per_second": 2.898,
|
| 1017 |
+
"eval_wer": 0.9918123785734111,
|
| 1018 |
+
"step": 8900
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 23.93617021276596,
|
| 1022 |
+
"grad_norm": 1.4986417293548584,
|
| 1023 |
+
"learning_rate": 0.00021588,
|
| 1024 |
+
"loss": 0.1647,
|
| 1025 |
+
"step": 9000
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 23.93617021276596,
|
| 1029 |
+
"eval_cer": 0.1869538837638785,
|
| 1030 |
+
"eval_loss": 0.4083107113838196,
|
| 1031 |
+
"eval_runtime": 229.7373,
|
| 1032 |
+
"eval_samples_per_second": 21.586,
|
| 1033 |
+
"eval_steps_per_second": 2.699,
|
| 1034 |
+
"eval_wer": 0.984179850124896,
|
| 1035 |
+
"step": 9000
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 24.20212765957447,
|
| 1039 |
+
"eval_cer": 0.18472181010544778,
|
| 1040 |
+
"eval_loss": 0.40058091282844543,
|
| 1041 |
+
"eval_runtime": 215.0023,
|
| 1042 |
+
"eval_samples_per_second": 23.065,
|
| 1043 |
+
"eval_steps_per_second": 2.884,
|
| 1044 |
+
"eval_wer": 0.9858451290591174,
|
| 1045 |
+
"step": 9100
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 24.46808510638298,
|
| 1049 |
+
"eval_cer": 0.18496096085456534,
|
| 1050 |
+
"eval_loss": 0.4137335419654846,
|
| 1051 |
+
"eval_runtime": 233.293,
|
| 1052 |
+
"eval_samples_per_second": 21.257,
|
| 1053 |
+
"eval_steps_per_second": 2.658,
|
| 1054 |
+
"eval_wer": 1.001526505689703,
|
| 1055 |
+
"step": 9200
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 24.73404255319149,
|
| 1059 |
+
"eval_cer": 0.1906208619170147,
|
| 1060 |
+
"eval_loss": 0.4106745719909668,
|
| 1061 |
+
"eval_runtime": 218.151,
|
| 1062 |
+
"eval_samples_per_second": 22.732,
|
| 1063 |
+
"eval_steps_per_second": 2.842,
|
| 1064 |
+
"eval_wer": 0.9994449070219261,
|
| 1065 |
+
"step": 9300
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 25.0,
|
| 1069 |
+
"eval_cer": 0.19122759622496113,
|
| 1070 |
+
"eval_loss": 0.42086583375930786,
|
| 1071 |
+
"eval_runtime": 214.8363,
|
| 1072 |
+
"eval_samples_per_second": 23.083,
|
| 1073 |
+
"eval_steps_per_second": 2.886,
|
| 1074 |
+
"eval_wer": 0.9843186233694143,
|
| 1075 |
+
"step": 9400
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 25.26595744680851,
|
| 1079 |
+
"grad_norm": 6.019426345825195,
|
| 1080 |
+
"learning_rate": 0.00022785599999999997,
|
| 1081 |
+
"loss": 0.1667,
|
| 1082 |
+
"step": 9500
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 25.26595744680851,
|
| 1086 |
+
"eval_cer": 0.18928338920898674,
|
| 1087 |
+
"eval_loss": 0.4373093247413635,
|
| 1088 |
+
"eval_runtime": 214.1648,
|
| 1089 |
+
"eval_samples_per_second": 23.155,
|
| 1090 |
+
"eval_steps_per_second": 2.895,
|
| 1091 |
+
"eval_wer": 0.9956980294199278,
|
| 1092 |
+
"step": 9500
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 25.53191489361702,
|
| 1096 |
+
"eval_cer": 0.18901766615441168,
|
| 1097 |
+
"eval_loss": 0.43900319933891296,
|
| 1098 |
+
"eval_runtime": 216.3618,
|
| 1099 |
+
"eval_samples_per_second": 22.92,
|
| 1100 |
+
"eval_steps_per_second": 2.866,
|
| 1101 |
+
"eval_wer": 0.9822370247016375,
|
| 1102 |
+
"step": 9600
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 25.79787234042553,
|
| 1106 |
+
"eval_cer": 0.19637376604856532,
|
| 1107 |
+
"eval_loss": 0.4538724422454834,
|
| 1108 |
+
"eval_runtime": 220.7821,
|
| 1109 |
+
"eval_samples_per_second": 22.461,
|
| 1110 |
+
"eval_steps_per_second": 2.808,
|
| 1111 |
+
"eval_wer": 0.985706355814599,
|
| 1112 |
+
"step": 9700
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 26.06382978723404,
|
| 1116 |
+
"eval_cer": 0.19334895194398558,
|
| 1117 |
+
"eval_loss": 0.43809235095977783,
|
| 1118 |
+
"eval_runtime": 213.8497,
|
| 1119 |
+
"eval_samples_per_second": 23.189,
|
| 1120 |
+
"eval_steps_per_second": 2.899,
|
| 1121 |
+
"eval_wer": 1.0037468776019984,
|
| 1122 |
+
"step": 9800
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 26.329787234042552,
|
| 1126 |
+
"eval_cer": 0.18651101200625336,
|
| 1127 |
+
"eval_loss": 0.4227481782436371,
|
| 1128 |
+
"eval_runtime": 219.4755,
|
| 1129 |
+
"eval_samples_per_second": 22.595,
|
| 1130 |
+
"eval_steps_per_second": 2.825,
|
| 1131 |
+
"eval_wer": 0.9875104079933389,
|
| 1132 |
+
"step": 9900
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 26.595744680851062,
|
| 1136 |
+
"grad_norm": 3.2743945121765137,
|
| 1137 |
+
"learning_rate": 0.000239856,
|
| 1138 |
+
"loss": 0.1644,
|
| 1139 |
+
"step": 10000
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 26.595744680851062,
|
| 1143 |
+
"eval_cer": 0.18837107338827896,
|
| 1144 |
+
"eval_loss": 0.4802379012107849,
|
| 1145 |
+
"eval_runtime": 214.7366,
|
| 1146 |
+
"eval_samples_per_second": 23.093,
|
| 1147 |
+
"eval_steps_per_second": 2.887,
|
| 1148 |
+
"eval_wer": 1.0266444629475437,
|
| 1149 |
+
"step": 10000
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 26.861702127659573,
|
| 1153 |
+
"eval_cer": 0.19581131891638137,
|
| 1154 |
+
"eval_loss": 0.4389278292655945,
|
| 1155 |
+
"eval_runtime": 217.9698,
|
| 1156 |
+
"eval_samples_per_second": 22.751,
|
| 1157 |
+
"eval_steps_per_second": 2.844,
|
| 1158 |
+
"eval_wer": 0.9950041631973355,
|
| 1159 |
+
"step": 10100
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 27.127659574468087,
|
| 1163 |
+
"eval_cer": 0.19391139907616953,
|
| 1164 |
+
"eval_loss": 0.4744134247303009,
|
| 1165 |
+
"eval_runtime": 217.5599,
|
| 1166 |
+
"eval_samples_per_second": 22.794,
|
| 1167 |
+
"eval_steps_per_second": 2.85,
|
| 1168 |
+
"eval_wer": 0.9827921176797113,
|
| 1169 |
+
"step": 10200
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 27.393617021276597,
|
| 1173 |
+
"eval_cer": 0.1983356879348447,
|
| 1174 |
+
"eval_loss": 0.4494173526763916,
|
| 1175 |
+
"eval_runtime": 214.0697,
|
| 1176 |
+
"eval_samples_per_second": 23.165,
|
| 1177 |
+
"eval_steps_per_second": 2.896,
|
| 1178 |
+
"eval_wer": 1.0005550929780738,
|
| 1179 |
+
"step": 10300
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 27.659574468085108,
|
| 1183 |
+
"eval_cer": 0.19612132914671898,
|
| 1184 |
+
"eval_loss": 0.44136643409729004,
|
| 1185 |
+
"eval_runtime": 214.2461,
|
| 1186 |
+
"eval_samples_per_second": 23.146,
|
| 1187 |
+
"eval_steps_per_second": 2.894,
|
| 1188 |
+
"eval_wer": 0.9962531223980017,
|
| 1189 |
+
"step": 10400
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 27.925531914893618,
|
| 1193 |
+
"grad_norm": 5.488828659057617,
|
| 1194 |
+
"learning_rate": 0.000251856,
|
| 1195 |
+
"loss": 0.1742,
|
| 1196 |
+
"step": 10500
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 27.925531914893618,
|
| 1200 |
+
"eval_cer": 0.19315851708820678,
|
| 1201 |
+
"eval_loss": 0.46681222319602966,
|
| 1202 |
+
"eval_runtime": 214.6029,
|
| 1203 |
+
"eval_samples_per_second": 23.108,
|
| 1204 |
+
"eval_steps_per_second": 2.889,
|
| 1205 |
+
"eval_wer": 0.9764085484318623,
|
| 1206 |
+
"step": 10500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 28.19148936170213,
|
| 1210 |
+
"eval_cer": 0.18782634112640004,
|
| 1211 |
+
"eval_loss": 0.42840632796287537,
|
| 1212 |
+
"eval_runtime": 214.4792,
|
| 1213 |
+
"eval_samples_per_second": 23.121,
|
| 1214 |
+
"eval_steps_per_second": 2.891,
|
| 1215 |
+
"eval_wer": 0.9719678046072717,
|
| 1216 |
+
"step": 10600
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 28.45744680851064,
|
| 1220 |
+
"eval_cer": 0.1943764144216759,
|
| 1221 |
+
"eval_loss": 0.4257587492465973,
|
| 1222 |
+
"eval_runtime": 214.0589,
|
| 1223 |
+
"eval_samples_per_second": 23.167,
|
| 1224 |
+
"eval_steps_per_second": 2.896,
|
| 1225 |
+
"eval_wer": 1.02789342214821,
|
| 1226 |
+
"step": 10700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 28.72340425531915,
|
| 1230 |
+
"eval_cer": 0.18916381383442796,
|
| 1231 |
+
"eval_loss": 0.4250844120979309,
|
| 1232 |
+
"eval_runtime": 228.7764,
|
| 1233 |
+
"eval_samples_per_second": 21.676,
|
| 1234 |
+
"eval_steps_per_second": 2.71,
|
| 1235 |
+
"eval_wer": 1.0023591451568137,
|
| 1236 |
+
"step": 10800
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 28.98936170212766,
|
| 1240 |
+
"eval_cer": 0.19779538439054203,
|
| 1241 |
+
"eval_loss": 0.4596838653087616,
|
| 1242 |
+
"eval_runtime": 214.1002,
|
| 1243 |
+
"eval_samples_per_second": 23.162,
|
| 1244 |
+
"eval_steps_per_second": 2.896,
|
| 1245 |
+
"eval_wer": 1.0201221204551763,
|
| 1246 |
+
"step": 10900
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 29.25531914893617,
|
| 1250 |
+
"grad_norm": 4.931338787078857,
|
| 1251 |
+
"learning_rate": 0.000263832,
|
| 1252 |
+
"loss": 0.1669,
|
| 1253 |
+
"step": 11000
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 29.25531914893617,
|
| 1257 |
+
"eval_cer": 0.1918786177086701,
|
| 1258 |
+
"eval_loss": 0.4414420425891876,
|
| 1259 |
+
"eval_runtime": 214.9531,
|
| 1260 |
+
"eval_samples_per_second": 23.07,
|
| 1261 |
+
"eval_steps_per_second": 2.884,
|
| 1262 |
+
"eval_wer": 0.9879267277268943,
|
| 1263 |
+
"step": 11000
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 29.52127659574468,
|
| 1267 |
+
"eval_cer": 0.19090429984189478,
|
| 1268 |
+
"eval_loss": 0.4472629725933075,
|
| 1269 |
+
"eval_runtime": 220.3712,
|
| 1270 |
+
"eval_samples_per_second": 22.503,
|
| 1271 |
+
"eval_steps_per_second": 2.813,
|
| 1272 |
+
"eval_wer": 0.9772411878989731,
|
| 1273 |
+
"step": 11100
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 29.78723404255319,
|
| 1277 |
+
"eval_cer": 0.19330466476822306,
|
| 1278 |
+
"eval_loss": 0.45273834466934204,
|
| 1279 |
+
"eval_runtime": 214.7179,
|
| 1280 |
+
"eval_samples_per_second": 23.095,
|
| 1281 |
+
"eval_steps_per_second": 2.888,
|
| 1282 |
+
"eval_wer": 0.9944490702192618,
|
| 1283 |
+
"step": 11200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 30.0,
|
| 1287 |
+
"step": 11280,
|
| 1288 |
+
"total_flos": 1.5765326864839811e+19,
|
| 1289 |
+
"train_loss": 0.910362657587579,
|
| 1290 |
+
"train_runtime": 61619.102,
|
| 1291 |
+
"train_samples_per_second": 5.857,
|
| 1292 |
+
"train_steps_per_second": 0.183
|
| 1293 |
+
}
|
| 1294 |
+
],
|
| 1295 |
+
"logging_steps": 500,
|
| 1296 |
+
"max_steps": 11280,
|
| 1297 |
+
"num_input_tokens_seen": 0,
|
| 1298 |
+
"num_train_epochs": 30,
|
| 1299 |
+
"save_steps": 400,
|
| 1300 |
+
"stateful_callbacks": {
|
| 1301 |
+
"TrainerControl": {
|
| 1302 |
+
"args": {
|
| 1303 |
+
"should_epoch_stop": false,
|
| 1304 |
+
"should_evaluate": false,
|
| 1305 |
+
"should_log": false,
|
| 1306 |
+
"should_save": true,
|
| 1307 |
+
"should_training_stop": true
|
| 1308 |
+
},
|
| 1309 |
+
"attributes": {}
|
| 1310 |
+
}
|
| 1311 |
+
},
|
| 1312 |
+
"total_flos": 1.5765326864839811e+19,
|
| 1313 |
+
"train_batch_size": 16,
|
| 1314 |
+
"trial_name": null,
|
| 1315 |
+
"trial_params": null
|
| 1316 |
+
}
|