End of training
Browse files- README.md +8 -17
- all_results.json +16 -0
- eval_results.json +10 -0
- runs/May10_16-01-02_mypc/events.out.tfevents.1715363551.mypc.12097.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +1131 -0
README.md
CHANGED
|
@@ -1,7 +1,11 @@
|
|
| 1 |
---
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
base_model: facebook/wav2vec2-xls-r-300m
|
| 4 |
tags:
|
|
|
|
|
|
|
| 5 |
- generated_from_trainer
|
| 6 |
datasets:
|
| 7 |
- common_voice_8_0
|
|
@@ -9,20 +13,7 @@ metrics:
|
|
| 9 |
- wer
|
| 10 |
model-index:
|
| 11 |
- name: xls-r-uzbek-cv8
|
| 12 |
-
results:
|
| 13 |
-
- task:
|
| 14 |
-
name: Automatic Speech Recognition
|
| 15 |
-
type: automatic-speech-recognition
|
| 16 |
-
dataset:
|
| 17 |
-
name: common_voice_8_0
|
| 18 |
-
type: common_voice_8_0
|
| 19 |
-
config: uz
|
| 20 |
-
split: validation[50%:]
|
| 21 |
-
args: uz
|
| 22 |
-
metrics:
|
| 23 |
-
- name: Wer
|
| 24 |
-
type: wer
|
| 25 |
-
value: 0.3785223774567843
|
| 26 |
---
|
| 27 |
|
| 28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -30,10 +21,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 30 |
|
| 31 |
# xls-r-uzbek-cv8
|
| 32 |
|
| 33 |
-
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the
|
| 34 |
It achieves the following results on the evaluation set:
|
| 35 |
-
- Loss: 0.
|
| 36 |
-
- Wer: 0.
|
| 37 |
- Cer: 0.0760
|
| 38 |
|
| 39 |
## Model description
|
|
|
|
| 1 |
---
|
| 2 |
+
language:
|
| 3 |
+
- uz
|
| 4 |
license: apache-2.0
|
| 5 |
base_model: facebook/wav2vec2-xls-r-300m
|
| 6 |
tags:
|
| 7 |
+
- automatic-speech-recognition
|
| 8 |
+
- mozilla-foundation/common_voice_8_0
|
| 9 |
- generated_from_trainer
|
| 10 |
datasets:
|
| 11 |
- common_voice_8_0
|
|
|
|
| 13 |
- wer
|
| 14 |
model-index:
|
| 15 |
- name: xls-r-uzbek-cv8
|
| 16 |
+
results: []
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
---
|
| 18 |
|
| 19 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 21 |
|
| 22 |
# xls-r-uzbek-cv8
|
| 23 |
|
| 24 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - UZ dataset.
|
| 25 |
It achieves the following results on the evaluation set:
|
| 26 |
+
- Loss: 0.2924
|
| 27 |
+
- Wer: 0.3780
|
| 28 |
- Cer: 0.0760
|
| 29 |
|
| 30 |
## Model description
|
all_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_cer": 0.07595390150143673,
|
| 4 |
+
"eval_loss": 0.29243385791778564,
|
| 5 |
+
"eval_runtime": 275.1252,
|
| 6 |
+
"eval_samples": 5425,
|
| 7 |
+
"eval_samples_per_second": 19.718,
|
| 8 |
+
"eval_steps_per_second": 4.932,
|
| 9 |
+
"eval_wer": 0.37798958086668244,
|
| 10 |
+
"total_flos": 2.6616913888307843e+19,
|
| 11 |
+
"train_loss": 1.1441452417737183,
|
| 12 |
+
"train_runtime": 24098.893,
|
| 13 |
+
"train_samples": 19726,
|
| 14 |
+
"train_samples_per_second": 8.185,
|
| 15 |
+
"train_steps_per_second": 0.512
|
| 16 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_cer": 0.07595390150143673,
|
| 4 |
+
"eval_loss": 0.29243385791778564,
|
| 5 |
+
"eval_runtime": 275.1252,
|
| 6 |
+
"eval_samples": 5425,
|
| 7 |
+
"eval_samples_per_second": 19.718,
|
| 8 |
+
"eval_steps_per_second": 4.932,
|
| 9 |
+
"eval_wer": 0.37798958086668244
|
| 10 |
+
}
|
runs/May10_16-01-02_mypc/events.out.tfevents.1715363551.mypc.12097.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d19709b7d014080061e00d5b79d7ddb18c9c366253fcd623bdedde4b303a4e5a
|
| 3 |
+
size 453
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 2.6616913888307843e+19,
|
| 4 |
+
"train_loss": 1.1441452417737183,
|
| 5 |
+
"train_runtime": 24098.893,
|
| 6 |
+
"train_samples": 19726,
|
| 7 |
+
"train_samples_per_second": 8.185,
|
| 8 |
+
"train_steps_per_second": 0.512
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1131 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 12330,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.08110300081103,
|
| 13 |
+
"grad_norm": 28.603330612182617,
|
| 14 |
+
"learning_rate": 5.8800000000000005e-06,
|
| 15 |
+
"loss": 12.3758,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.16220600162206,
|
| 20 |
+
"grad_norm": 19.174470901489258,
|
| 21 |
+
"learning_rate": 1.1880000000000001e-05,
|
| 22 |
+
"loss": 5.5109,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.24330900243309003,
|
| 27 |
+
"grad_norm": 13.463099479675293,
|
| 28 |
+
"learning_rate": 1.7879999999999998e-05,
|
| 29 |
+
"loss": 3.8591,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.32441200324412,
|
| 34 |
+
"grad_norm": 7.847727298736572,
|
| 35 |
+
"learning_rate": 2.3880000000000002e-05,
|
| 36 |
+
"loss": 3.3292,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.40551500405515006,
|
| 41 |
+
"grad_norm": 7.480207443237305,
|
| 42 |
+
"learning_rate": 2.9880000000000002e-05,
|
| 43 |
+
"loss": 3.1444,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.40551500405515006,
|
| 48 |
+
"eval_cer": 1.0,
|
| 49 |
+
"eval_loss": 3.1199822425842285,
|
| 50 |
+
"eval_runtime": 244.3979,
|
| 51 |
+
"eval_samples_per_second": 22.197,
|
| 52 |
+
"eval_steps_per_second": 5.552,
|
| 53 |
+
"eval_wer": 1.0,
|
| 54 |
+
"step": 500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.48661800486618007,
|
| 58 |
+
"grad_norm": 5.100943565368652,
|
| 59 |
+
"learning_rate": 2.9751479289940832e-05,
|
| 60 |
+
"loss": 3.0927,
|
| 61 |
+
"step": 600
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.5677210056772101,
|
| 65 |
+
"grad_norm": 4.977197647094727,
|
| 66 |
+
"learning_rate": 2.9497886728655962e-05,
|
| 67 |
+
"loss": 3.0491,
|
| 68 |
+
"step": 700
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.64882400648824,
|
| 72 |
+
"grad_norm": 4.224981307983398,
|
| 73 |
+
"learning_rate": 2.9244294167371092e-05,
|
| 74 |
+
"loss": 3.0201,
|
| 75 |
+
"step": 800
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.7299270072992701,
|
| 79 |
+
"grad_norm": 5.425148010253906,
|
| 80 |
+
"learning_rate": 2.8990701606086222e-05,
|
| 81 |
+
"loss": 2.9859,
|
| 82 |
+
"step": 900
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.8110300081103001,
|
| 86 |
+
"grad_norm": 1.351231336593628,
|
| 87 |
+
"learning_rate": 2.8737109044801352e-05,
|
| 88 |
+
"loss": 2.9488,
|
| 89 |
+
"step": 1000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.8110300081103001,
|
| 93 |
+
"eval_cer": 0.9806716021231193,
|
| 94 |
+
"eval_loss": 2.9561665058135986,
|
| 95 |
+
"eval_runtime": 243.1917,
|
| 96 |
+
"eval_samples_per_second": 22.308,
|
| 97 |
+
"eval_steps_per_second": 5.58,
|
| 98 |
+
"eval_wer": 1.0,
|
| 99 |
+
"step": 1000
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.8921330089213301,
|
| 103 |
+
"grad_norm": 1.8469974994659424,
|
| 104 |
+
"learning_rate": 2.8483516483516482e-05,
|
| 105 |
+
"loss": 2.8852,
|
| 106 |
+
"step": 1100
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.9732360097323601,
|
| 110 |
+
"grad_norm": 2.408888101577759,
|
| 111 |
+
"learning_rate": 2.8229923922231613e-05,
|
| 112 |
+
"loss": 2.542,
|
| 113 |
+
"step": 1200
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 1.05433901054339,
|
| 117 |
+
"grad_norm": 1.82163405418396,
|
| 118 |
+
"learning_rate": 2.7976331360946746e-05,
|
| 119 |
+
"loss": 1.978,
|
| 120 |
+
"step": 1300
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.1354420113544201,
|
| 124 |
+
"grad_norm": 1.5221664905548096,
|
| 125 |
+
"learning_rate": 2.7722738799661876e-05,
|
| 126 |
+
"loss": 1.6172,
|
| 127 |
+
"step": 1400
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 1.2165450121654502,
|
| 131 |
+
"grad_norm": 1.6060534715652466,
|
| 132 |
+
"learning_rate": 2.7469146238377006e-05,
|
| 133 |
+
"loss": 1.4553,
|
| 134 |
+
"step": 1500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.2165450121654502,
|
| 138 |
+
"eval_cer": 0.16442234359713118,
|
| 139 |
+
"eval_loss": 0.7867947816848755,
|
| 140 |
+
"eval_runtime": 245.9044,
|
| 141 |
+
"eval_samples_per_second": 22.061,
|
| 142 |
+
"eval_steps_per_second": 5.518,
|
| 143 |
+
"eval_wer": 0.7034394979872128,
|
| 144 |
+
"step": 1500
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.29764801297648,
|
| 148 |
+
"grad_norm": 1.677599310874939,
|
| 149 |
+
"learning_rate": 2.721555367709214e-05,
|
| 150 |
+
"loss": 1.3456,
|
| 151 |
+
"step": 1600
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.37875101378751,
|
| 155 |
+
"grad_norm": 1.7622677087783813,
|
| 156 |
+
"learning_rate": 2.696196111580727e-05,
|
| 157 |
+
"loss": 1.2794,
|
| 158 |
+
"step": 1700
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.4598540145985401,
|
| 162 |
+
"grad_norm": 2.301548480987549,
|
| 163 |
+
"learning_rate": 2.67083685545224e-05,
|
| 164 |
+
"loss": 1.2401,
|
| 165 |
+
"step": 1800
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.5409570154095702,
|
| 169 |
+
"grad_norm": 1.7648967504501343,
|
| 170 |
+
"learning_rate": 2.6454775993237533e-05,
|
| 171 |
+
"loss": 1.2073,
|
| 172 |
+
"step": 1900
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 1.6220600162206003,
|
| 176 |
+
"grad_norm": 1.7196375131607056,
|
| 177 |
+
"learning_rate": 2.6201183431952664e-05,
|
| 178 |
+
"loss": 1.1495,
|
| 179 |
+
"step": 2000
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.6220600162206003,
|
| 183 |
+
"eval_cer": 0.13369643091002936,
|
| 184 |
+
"eval_loss": 0.559805154800415,
|
| 185 |
+
"eval_runtime": 246.7531,
|
| 186 |
+
"eval_samples_per_second": 21.986,
|
| 187 |
+
"eval_steps_per_second": 5.499,
|
| 188 |
+
"eval_wer": 0.6075953113900071,
|
| 189 |
+
"step": 2000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.7031630170316303,
|
| 193 |
+
"grad_norm": 1.7879103422164917,
|
| 194 |
+
"learning_rate": 2.5947590870667794e-05,
|
| 195 |
+
"loss": 1.1087,
|
| 196 |
+
"step": 2100
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.7842660178426601,
|
| 200 |
+
"grad_norm": 1.6160575151443481,
|
| 201 |
+
"learning_rate": 2.5693998309382927e-05,
|
| 202 |
+
"loss": 1.0974,
|
| 203 |
+
"step": 2200
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.8653690186536902,
|
| 207 |
+
"grad_norm": 2.434413194656372,
|
| 208 |
+
"learning_rate": 2.5440405748098057e-05,
|
| 209 |
+
"loss": 1.0571,
|
| 210 |
+
"step": 2300
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.94647201946472,
|
| 214 |
+
"grad_norm": 1.7195013761520386,
|
| 215 |
+
"learning_rate": 2.5186813186813187e-05,
|
| 216 |
+
"loss": 1.0384,
|
| 217 |
+
"step": 2400
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 2.02757502027575,
|
| 221 |
+
"grad_norm": 2.289520263671875,
|
| 222 |
+
"learning_rate": 2.4933220625528317e-05,
|
| 223 |
+
"loss": 1.041,
|
| 224 |
+
"step": 2500
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 2.02757502027575,
|
| 228 |
+
"eval_cer": 0.11740326171528939,
|
| 229 |
+
"eval_loss": 0.46495679020881653,
|
| 230 |
+
"eval_runtime": 246.6192,
|
| 231 |
+
"eval_samples_per_second": 21.997,
|
| 232 |
+
"eval_steps_per_second": 5.502,
|
| 233 |
+
"eval_wer": 0.5536644565474781,
|
| 234 |
+
"step": 2500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.10867802108678,
|
| 238 |
+
"grad_norm": 2.3239097595214844,
|
| 239 |
+
"learning_rate": 2.467962806424345e-05,
|
| 240 |
+
"loss": 1.002,
|
| 241 |
+
"step": 2600
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 2.18978102189781,
|
| 245 |
+
"grad_norm": 2.725745677947998,
|
| 246 |
+
"learning_rate": 2.442603550295858e-05,
|
| 247 |
+
"loss": 0.9909,
|
| 248 |
+
"step": 2700
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.2708840227088403,
|
| 252 |
+
"grad_norm": 2.6716995239257812,
|
| 253 |
+
"learning_rate": 2.417244294167371e-05,
|
| 254 |
+
"loss": 0.9665,
|
| 255 |
+
"step": 2800
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 2.3519870235198703,
|
| 259 |
+
"grad_norm": 2.5717599391937256,
|
| 260 |
+
"learning_rate": 2.3918850380388845e-05,
|
| 261 |
+
"loss": 0.9682,
|
| 262 |
+
"step": 2900
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 2.4330900243309004,
|
| 266 |
+
"grad_norm": 2.675999164581299,
|
| 267 |
+
"learning_rate": 2.3665257819103975e-05,
|
| 268 |
+
"loss": 0.9524,
|
| 269 |
+
"step": 3000
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 2.4330900243309004,
|
| 273 |
+
"eval_cer": 0.10609819041822986,
|
| 274 |
+
"eval_loss": 0.420441210269928,
|
| 275 |
+
"eval_runtime": 247.0264,
|
| 276 |
+
"eval_samples_per_second": 21.961,
|
| 277 |
+
"eval_steps_per_second": 5.493,
|
| 278 |
+
"eval_wer": 0.5097679374852001,
|
| 279 |
+
"step": 3000
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 2.51419302514193,
|
| 283 |
+
"grad_norm": 2.483025550842285,
|
| 284 |
+
"learning_rate": 2.3411665257819105e-05,
|
| 285 |
+
"loss": 0.9359,
|
| 286 |
+
"step": 3100
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 2.59529602595296,
|
| 290 |
+
"grad_norm": 3.9523444175720215,
|
| 291 |
+
"learning_rate": 2.315807269653424e-05,
|
| 292 |
+
"loss": 0.9349,
|
| 293 |
+
"step": 3200
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 2.67639902676399,
|
| 297 |
+
"grad_norm": 1.923336148262024,
|
| 298 |
+
"learning_rate": 2.290448013524937e-05,
|
| 299 |
+
"loss": 0.915,
|
| 300 |
+
"step": 3300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.75750202757502,
|
| 304 |
+
"grad_norm": 2.7855656147003174,
|
| 305 |
+
"learning_rate": 2.2650887573964495e-05,
|
| 306 |
+
"loss": 0.902,
|
| 307 |
+
"step": 3400
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.83860502838605,
|
| 311 |
+
"grad_norm": 2.1685218811035156,
|
| 312 |
+
"learning_rate": 2.239729501267963e-05,
|
| 313 |
+
"loss": 0.902,
|
| 314 |
+
"step": 3500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.83860502838605,
|
| 318 |
+
"eval_cer": 0.10258533691808734,
|
| 319 |
+
"eval_loss": 0.3919403553009033,
|
| 320 |
+
"eval_runtime": 247.0433,
|
| 321 |
+
"eval_samples_per_second": 21.96,
|
| 322 |
+
"eval_steps_per_second": 5.493,
|
| 323 |
+
"eval_wer": 0.49840161022969454,
|
| 324 |
+
"step": 3500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.9197080291970803,
|
| 328 |
+
"grad_norm": 2.623002052307129,
|
| 329 |
+
"learning_rate": 2.214370245139476e-05,
|
| 330 |
+
"loss": 0.886,
|
| 331 |
+
"step": 3600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 3.0008110300081103,
|
| 335 |
+
"grad_norm": 2.3057801723480225,
|
| 336 |
+
"learning_rate": 2.189010989010989e-05,
|
| 337 |
+
"loss": 0.8985,
|
| 338 |
+
"step": 3700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 3.0819140308191404,
|
| 342 |
+
"grad_norm": 1.5028676986694336,
|
| 343 |
+
"learning_rate": 2.163651732882502e-05,
|
| 344 |
+
"loss": 0.8698,
|
| 345 |
+
"step": 3800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.1630170316301705,
|
| 349 |
+
"grad_norm": 1.727243423461914,
|
| 350 |
+
"learning_rate": 2.1382924767540153e-05,
|
| 351 |
+
"loss": 0.8694,
|
| 352 |
+
"step": 3900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.2441200324412005,
|
| 356 |
+
"grad_norm": 1.664138913154602,
|
| 357 |
+
"learning_rate": 2.1129332206255283e-05,
|
| 358 |
+
"loss": 0.8505,
|
| 359 |
+
"step": 4000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 3.2441200324412005,
|
| 363 |
+
"eval_cer": 0.09652258317990278,
|
| 364 |
+
"eval_loss": 0.3688036799430847,
|
| 365 |
+
"eval_runtime": 255.621,
|
| 366 |
+
"eval_samples_per_second": 21.223,
|
| 367 |
+
"eval_steps_per_second": 5.309,
|
| 368 |
+
"eval_wer": 0.4677954061094009,
|
| 369 |
+
"step": 4000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 3.3252230332522306,
|
| 373 |
+
"grad_norm": 1.814882755279541,
|
| 374 |
+
"learning_rate": 2.0875739644970413e-05,
|
| 375 |
+
"loss": 0.848,
|
| 376 |
+
"step": 4100
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 3.40632603406326,
|
| 380 |
+
"grad_norm": 1.9209064245224,
|
| 381 |
+
"learning_rate": 2.0622147083685546e-05,
|
| 382 |
+
"loss": 0.8697,
|
| 383 |
+
"step": 4200
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.4874290348742902,
|
| 387 |
+
"grad_norm": 2.795137882232666,
|
| 388 |
+
"learning_rate": 2.0368554522400676e-05,
|
| 389 |
+
"loss": 0.8474,
|
| 390 |
+
"step": 4300
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 3.5685320356853203,
|
| 394 |
+
"grad_norm": 1.8658416271209717,
|
| 395 |
+
"learning_rate": 2.0114961961115806e-05,
|
| 396 |
+
"loss": 0.8287,
|
| 397 |
+
"step": 4400
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 3.6496350364963503,
|
| 401 |
+
"grad_norm": 2.143341541290283,
|
| 402 |
+
"learning_rate": 1.986136939983094e-05,
|
| 403 |
+
"loss": 0.8353,
|
| 404 |
+
"step": 4500
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 3.6496350364963503,
|
| 408 |
+
"eval_cer": 0.09146900446039952,
|
| 409 |
+
"eval_loss": 0.34911593794822693,
|
| 410 |
+
"eval_runtime": 247.7614,
|
| 411 |
+
"eval_samples_per_second": 21.896,
|
| 412 |
+
"eval_steps_per_second": 5.477,
|
| 413 |
+
"eval_wer": 0.44882192753966377,
|
| 414 |
+
"step": 4500
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 3.7307380373073804,
|
| 418 |
+
"grad_norm": 1.6942418813705444,
|
| 419 |
+
"learning_rate": 1.960777683854607e-05,
|
| 420 |
+
"loss": 0.8344,
|
| 421 |
+
"step": 4600
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 3.8118410381184105,
|
| 425 |
+
"grad_norm": 2.338500499725342,
|
| 426 |
+
"learning_rate": 1.93541842772612e-05,
|
| 427 |
+
"loss": 0.8232,
|
| 428 |
+
"step": 4700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 3.8929440389294405,
|
| 432 |
+
"grad_norm": 2.209710121154785,
|
| 433 |
+
"learning_rate": 1.9100591715976334e-05,
|
| 434 |
+
"loss": 0.8208,
|
| 435 |
+
"step": 4800
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 3.97404703974047,
|
| 439 |
+
"grad_norm": 2.0351734161376953,
|
| 440 |
+
"learning_rate": 1.8846999154691464e-05,
|
| 441 |
+
"loss": 0.8382,
|
| 442 |
+
"step": 4900
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 4.0551500405515,
|
| 446 |
+
"grad_norm": 2.32840633392334,
|
| 447 |
+
"learning_rate": 1.8593406593406594e-05,
|
| 448 |
+
"loss": 0.8015,
|
| 449 |
+
"step": 5000
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 4.0551500405515,
|
| 453 |
+
"eval_cer": 0.0895970233188762,
|
| 454 |
+
"eval_loss": 0.3409828543663025,
|
| 455 |
+
"eval_runtime": 247.0784,
|
| 456 |
+
"eval_samples_per_second": 21.957,
|
| 457 |
+
"eval_steps_per_second": 5.492,
|
| 458 |
+
"eval_wer": 0.435620412029363,
|
| 459 |
+
"step": 5000
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 4.13625304136253,
|
| 463 |
+
"grad_norm": 3.6171586513519287,
|
| 464 |
+
"learning_rate": 1.8339814032121724e-05,
|
| 465 |
+
"loss": 0.8046,
|
| 466 |
+
"step": 5100
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 4.21735604217356,
|
| 470 |
+
"grad_norm": 2.958981513977051,
|
| 471 |
+
"learning_rate": 1.8086221470836857e-05,
|
| 472 |
+
"loss": 0.7952,
|
| 473 |
+
"step": 5200
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 4.29845904298459,
|
| 477 |
+
"grad_norm": 2.3776564598083496,
|
| 478 |
+
"learning_rate": 1.7832628909551988e-05,
|
| 479 |
+
"loss": 0.8001,
|
| 480 |
+
"step": 5300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 4.37956204379562,
|
| 484 |
+
"grad_norm": 1.985517978668213,
|
| 485 |
+
"learning_rate": 1.7579036348267118e-05,
|
| 486 |
+
"loss": 0.8129,
|
| 487 |
+
"step": 5400
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 4.4606650446066505,
|
| 491 |
+
"grad_norm": 2.3717586994171143,
|
| 492 |
+
"learning_rate": 1.732544378698225e-05,
|
| 493 |
+
"loss": 0.7771,
|
| 494 |
+
"step": 5500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 4.4606650446066505,
|
| 498 |
+
"eval_cer": 0.08826429600412915,
|
| 499 |
+
"eval_loss": 0.3367164433002472,
|
| 500 |
+
"eval_runtime": 248.3247,
|
| 501 |
+
"eval_samples_per_second": 21.846,
|
| 502 |
+
"eval_steps_per_second": 5.465,
|
| 503 |
+
"eval_wer": 0.4330452285105375,
|
| 504 |
+
"step": 5500
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 4.5417680454176805,
|
| 508 |
+
"grad_norm": 4.7962188720703125,
|
| 509 |
+
"learning_rate": 1.707185122569738e-05,
|
| 510 |
+
"loss": 0.7874,
|
| 511 |
+
"step": 5600
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 4.622871046228711,
|
| 515 |
+
"grad_norm": 2.154824733734131,
|
| 516 |
+
"learning_rate": 1.682079459002536e-05,
|
| 517 |
+
"loss": 0.7793,
|
| 518 |
+
"step": 5700
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 4.703974047039741,
|
| 522 |
+
"grad_norm": 1.78102445602417,
|
| 523 |
+
"learning_rate": 1.656720202874049e-05,
|
| 524 |
+
"loss": 0.79,
|
| 525 |
+
"step": 5800
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 4.785077047850771,
|
| 529 |
+
"grad_norm": 5.108635902404785,
|
| 530 |
+
"learning_rate": 1.6313609467455624e-05,
|
| 531 |
+
"loss": 0.7702,
|
| 532 |
+
"step": 5900
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 4.866180048661801,
|
| 536 |
+
"grad_norm": 2.097984552383423,
|
| 537 |
+
"learning_rate": 1.6060016906170754e-05,
|
| 538 |
+
"loss": 0.7894,
|
| 539 |
+
"step": 6000
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 4.866180048661801,
|
| 543 |
+
"eval_cer": 0.08582609834449074,
|
| 544 |
+
"eval_loss": 0.32740598917007446,
|
| 545 |
+
"eval_runtime": 259.0642,
|
| 546 |
+
"eval_samples_per_second": 20.941,
|
| 547 |
+
"eval_steps_per_second": 5.238,
|
| 548 |
+
"eval_wer": 0.42008051148472647,
|
| 549 |
+
"step": 6000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 4.947283049472831,
|
| 553 |
+
"grad_norm": 3.253312349319458,
|
| 554 |
+
"learning_rate": 1.5806424344885884e-05,
|
| 555 |
+
"loss": 0.7845,
|
| 556 |
+
"step": 6100
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 5.028386050283861,
|
| 560 |
+
"grad_norm": 2.6536951065063477,
|
| 561 |
+
"learning_rate": 1.5552831783601017e-05,
|
| 562 |
+
"loss": 0.7613,
|
| 563 |
+
"step": 6200
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 5.109489051094891,
|
| 567 |
+
"grad_norm": 3.970766544342041,
|
| 568 |
+
"learning_rate": 1.5299239222316148e-05,
|
| 569 |
+
"loss": 0.7546,
|
| 570 |
+
"step": 6300
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 5.19059205190592,
|
| 574 |
+
"grad_norm": 3.2884013652801514,
|
| 575 |
+
"learning_rate": 1.5045646661031278e-05,
|
| 576 |
+
"loss": 0.7666,
|
| 577 |
+
"step": 6400
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 5.27169505271695,
|
| 581 |
+
"grad_norm": 2.420640707015991,
|
| 582 |
+
"learning_rate": 1.4792054099746408e-05,
|
| 583 |
+
"loss": 0.7624,
|
| 584 |
+
"step": 6500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 5.27169505271695,
|
| 588 |
+
"eval_cer": 0.08352271414154643,
|
| 589 |
+
"eval_loss": 0.3266396224498749,
|
| 590 |
+
"eval_runtime": 275.6132,
|
| 591 |
+
"eval_samples_per_second": 19.683,
|
| 592 |
+
"eval_steps_per_second": 4.924,
|
| 593 |
+
"eval_wer": 0.41152616623253613,
|
| 594 |
+
"step": 6500
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 5.35279805352798,
|
| 598 |
+
"grad_norm": 3.38496470451355,
|
| 599 |
+
"learning_rate": 1.453846153846154e-05,
|
| 600 |
+
"loss": 0.7502,
|
| 601 |
+
"step": 6600
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 5.43390105433901,
|
| 605 |
+
"grad_norm": 4.515685081481934,
|
| 606 |
+
"learning_rate": 1.428486897717667e-05,
|
| 607 |
+
"loss": 0.7602,
|
| 608 |
+
"step": 6700
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 5.51500405515004,
|
| 612 |
+
"grad_norm": 4.857451915740967,
|
| 613 |
+
"learning_rate": 1.4031276415891802e-05,
|
| 614 |
+
"loss": 0.7547,
|
| 615 |
+
"step": 6800
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 5.59610705596107,
|
| 619 |
+
"grad_norm": 2.685760974884033,
|
| 620 |
+
"learning_rate": 1.3777683854606932e-05,
|
| 621 |
+
"loss": 0.7619,
|
| 622 |
+
"step": 6900
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 5.6772100567721,
|
| 626 |
+
"grad_norm": 2.401392698287964,
|
| 627 |
+
"learning_rate": 1.3524091293322063e-05,
|
| 628 |
+
"loss": 0.7522,
|
| 629 |
+
"step": 7000
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 5.6772100567721,
|
| 633 |
+
"eval_cer": 0.0824557619271391,
|
| 634 |
+
"eval_loss": 0.31717103719711304,
|
| 635 |
+
"eval_runtime": 275.5677,
|
| 636 |
+
"eval_samples_per_second": 19.687,
|
| 637 |
+
"eval_steps_per_second": 4.924,
|
| 638 |
+
"eval_wer": 0.4072341937011603,
|
| 639 |
+
"step": 7000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.7583130575831305,
|
| 643 |
+
"grad_norm": 6.463571548461914,
|
| 644 |
+
"learning_rate": 1.3270498732037195e-05,
|
| 645 |
+
"loss": 0.7555,
|
| 646 |
+
"step": 7100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 5.839416058394161,
|
| 650 |
+
"grad_norm": 2.7044007778167725,
|
| 651 |
+
"learning_rate": 1.3016906170752325e-05,
|
| 652 |
+
"loss": 0.7544,
|
| 653 |
+
"step": 7200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 5.920519059205191,
|
| 657 |
+
"grad_norm": 2.792142152786255,
|
| 658 |
+
"learning_rate": 1.2763313609467455e-05,
|
| 659 |
+
"loss": 0.7479,
|
| 660 |
+
"step": 7300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 6.001622060016221,
|
| 664 |
+
"grad_norm": 2.3005964756011963,
|
| 665 |
+
"learning_rate": 1.2509721048182586e-05,
|
| 666 |
+
"loss": 0.7378,
|
| 667 |
+
"step": 7400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 6.082725060827251,
|
| 671 |
+
"grad_norm": 1.4675862789154053,
|
| 672 |
+
"learning_rate": 1.2256128486897717e-05,
|
| 673 |
+
"loss": 0.7545,
|
| 674 |
+
"step": 7500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 6.082725060827251,
|
| 678 |
+
"eval_cer": 0.08170851019574914,
|
| 679 |
+
"eval_loss": 0.30963605642318726,
|
| 680 |
+
"eval_runtime": 275.6684,
|
| 681 |
+
"eval_samples_per_second": 19.679,
|
| 682 |
+
"eval_steps_per_second": 4.923,
|
| 683 |
+
"eval_wer": 0.4033862183282027,
|
| 684 |
+
"step": 7500
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 6.163828061638281,
|
| 688 |
+
"grad_norm": 1.5494076013565063,
|
| 689 |
+
"learning_rate": 1.2002535925612849e-05,
|
| 690 |
+
"loss": 0.7299,
|
| 691 |
+
"step": 7600
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 6.244931062449311,
|
| 695 |
+
"grad_norm": 1.931067705154419,
|
| 696 |
+
"learning_rate": 1.175147928994083e-05,
|
| 697 |
+
"loss": 0.7178,
|
| 698 |
+
"step": 7700
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 6.326034063260341,
|
| 702 |
+
"grad_norm": 2.1837286949157715,
|
| 703 |
+
"learning_rate": 1.149788672865596e-05,
|
| 704 |
+
"loss": 0.7326,
|
| 705 |
+
"step": 7800
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 6.407137064071371,
|
| 709 |
+
"grad_norm": 2.1483843326568604,
|
| 710 |
+
"learning_rate": 1.1244294167371092e-05,
|
| 711 |
+
"loss": 0.7401,
|
| 712 |
+
"step": 7900
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 6.488240064882401,
|
| 716 |
+
"grad_norm": 1.7043359279632568,
|
| 717 |
+
"learning_rate": 1.0990701606086222e-05,
|
| 718 |
+
"loss": 0.7412,
|
| 719 |
+
"step": 8000
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 6.488240064882401,
|
| 723 |
+
"eval_cer": 0.08098436934264959,
|
| 724 |
+
"eval_loss": 0.3062187731266022,
|
| 725 |
+
"eval_runtime": 275.8064,
|
| 726 |
+
"eval_samples_per_second": 19.67,
|
| 727 |
+
"eval_steps_per_second": 4.92,
|
| 728 |
+
"eval_wer": 0.40137343121004027,
|
| 729 |
+
"step": 8000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 6.569343065693431,
|
| 733 |
+
"grad_norm": 2.9203195571899414,
|
| 734 |
+
"learning_rate": 1.0737109044801352e-05,
|
| 735 |
+
"loss": 0.7424,
|
| 736 |
+
"step": 8100
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 6.650446066504461,
|
| 740 |
+
"grad_norm": 2.0401434898376465,
|
| 741 |
+
"learning_rate": 1.0483516483516484e-05,
|
| 742 |
+
"loss": 0.7338,
|
| 743 |
+
"step": 8200
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 6.73154906731549,
|
| 747 |
+
"grad_norm": 3.5595827102661133,
|
| 748 |
+
"learning_rate": 1.0229923922231614e-05,
|
| 749 |
+
"loss": 0.7345,
|
| 750 |
+
"step": 8300
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 6.81265206812652,
|
| 754 |
+
"grad_norm": 2.4295144081115723,
|
| 755 |
+
"learning_rate": 9.976331360946746e-06,
|
| 756 |
+
"loss": 0.7453,
|
| 757 |
+
"step": 8400
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 6.89375506893755,
|
| 761 |
+
"grad_norm": 2.3278117179870605,
|
| 762 |
+
"learning_rate": 9.722738799661877e-06,
|
| 763 |
+
"loss": 0.7405,
|
| 764 |
+
"step": 8500
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 6.89375506893755,
|
| 768 |
+
"eval_cer": 0.0795900130191281,
|
| 769 |
+
"eval_loss": 0.3056836724281311,
|
| 770 |
+
"eval_runtime": 277.6736,
|
| 771 |
+
"eval_samples_per_second": 19.537,
|
| 772 |
+
"eval_steps_per_second": 4.887,
|
| 773 |
+
"eval_wer": 0.39326308311626806,
|
| 774 |
+
"step": 8500
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 6.9748580697485805,
|
| 778 |
+
"grad_norm": 2.005995273590088,
|
| 779 |
+
"learning_rate": 9.469146238377007e-06,
|
| 780 |
+
"loss": 0.7305,
|
| 781 |
+
"step": 8600
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 7.0559610705596105,
|
| 785 |
+
"grad_norm": 2.017148017883301,
|
| 786 |
+
"learning_rate": 9.21555367709214e-06,
|
| 787 |
+
"loss": 0.717,
|
| 788 |
+
"step": 8700
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 7.137064071370641,
|
| 792 |
+
"grad_norm": 5.9526143074035645,
|
| 793 |
+
"learning_rate": 8.96196111580727e-06,
|
| 794 |
+
"loss": 0.7096,
|
| 795 |
+
"step": 8800
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 7.218167072181671,
|
| 799 |
+
"grad_norm": 2.648564338684082,
|
| 800 |
+
"learning_rate": 8.708368554522401e-06,
|
| 801 |
+
"loss": 0.7068,
|
| 802 |
+
"step": 8900
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 7.299270072992701,
|
| 806 |
+
"grad_norm": 3.1904115676879883,
|
| 807 |
+
"learning_rate": 8.454775993237533e-06,
|
| 808 |
+
"loss": 0.703,
|
| 809 |
+
"step": 9000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 7.299270072992701,
|
| 813 |
+
"eval_cer": 0.07842291366546233,
|
| 814 |
+
"eval_loss": 0.2966497540473938,
|
| 815 |
+
"eval_runtime": 276.0376,
|
| 816 |
+
"eval_samples_per_second": 19.653,
|
| 817 |
+
"eval_steps_per_second": 4.916,
|
| 818 |
+
"eval_wer": 0.38938550793274923,
|
| 819 |
+
"step": 9000
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 7.380373073803731,
|
| 823 |
+
"grad_norm": 2.2572691440582275,
|
| 824 |
+
"learning_rate": 8.203719357565512e-06,
|
| 825 |
+
"loss": 0.7074,
|
| 826 |
+
"step": 9100
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 7.461476074614761,
|
| 830 |
+
"grad_norm": 2.1065661907196045,
|
| 831 |
+
"learning_rate": 7.950126796280642e-06,
|
| 832 |
+
"loss": 0.7178,
|
| 833 |
+
"step": 9200
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 7.542579075425791,
|
| 837 |
+
"grad_norm": 2.131349802017212,
|
| 838 |
+
"learning_rate": 7.696534234995774e-06,
|
| 839 |
+
"loss": 0.7223,
|
| 840 |
+
"step": 9300
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 7.623682076236821,
|
| 844 |
+
"grad_norm": 2.2301273345947266,
|
| 845 |
+
"learning_rate": 7.442941673710905e-06,
|
| 846 |
+
"loss": 0.7221,
|
| 847 |
+
"step": 9400
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 7.704785077047851,
|
| 851 |
+
"grad_norm": 2.2693982124328613,
|
| 852 |
+
"learning_rate": 7.189349112426036e-06,
|
| 853 |
+
"loss": 0.7091,
|
| 854 |
+
"step": 9500
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 7.704785077047851,
|
| 858 |
+
"eval_cer": 0.07844217273070435,
|
| 859 |
+
"eval_loss": 0.2999822795391083,
|
| 860 |
+
"eval_runtime": 276.8928,
|
| 861 |
+
"eval_samples_per_second": 19.592,
|
| 862 |
+
"eval_steps_per_second": 4.901,
|
| 863 |
+
"eval_wer": 0.3894743073644329,
|
| 864 |
+
"step": 9500
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 7.785888077858881,
|
| 868 |
+
"grad_norm": 2.934065818786621,
|
| 869 |
+
"learning_rate": 6.935756551141167e-06,
|
| 870 |
+
"loss": 0.7104,
|
| 871 |
+
"step": 9600
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 7.866991078669911,
|
| 875 |
+
"grad_norm": 3.0223405361175537,
|
| 876 |
+
"learning_rate": 6.682163989856298e-06,
|
| 877 |
+
"loss": 0.6909,
|
| 878 |
+
"step": 9700
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 7.948094079480941,
|
| 882 |
+
"grad_norm": 1.906209945678711,
|
| 883 |
+
"learning_rate": 6.428571428571429e-06,
|
| 884 |
+
"loss": 0.7272,
|
| 885 |
+
"step": 9800
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 8.02919708029197,
|
| 889 |
+
"grad_norm": 3.197634220123291,
|
| 890 |
+
"learning_rate": 6.1749788672865596e-06,
|
| 891 |
+
"loss": 0.6958,
|
| 892 |
+
"step": 9900
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 8.110300081103,
|
| 896 |
+
"grad_norm": 2.091019868850708,
|
| 897 |
+
"learning_rate": 5.9213863060016905e-06,
|
| 898 |
+
"loss": 0.7117,
|
| 899 |
+
"step": 10000
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 8.110300081103,
|
| 903 |
+
"eval_cer": 0.0781494349390258,
|
| 904 |
+
"eval_loss": 0.2987510859966278,
|
| 905 |
+
"eval_runtime": 276.3074,
|
| 906 |
+
"eval_samples_per_second": 19.634,
|
| 907 |
+
"eval_steps_per_second": 4.911,
|
| 908 |
+
"eval_wer": 0.38808311626805586,
|
| 909 |
+
"step": 10000
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 8.19140308191403,
|
| 913 |
+
"grad_norm": 2.5339033603668213,
|
| 914 |
+
"learning_rate": 5.6677937447168215e-06,
|
| 915 |
+
"loss": 0.7112,
|
| 916 |
+
"step": 10100
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 8.27250608272506,
|
| 920 |
+
"grad_norm": 2.70039701461792,
|
| 921 |
+
"learning_rate": 5.414201183431953e-06,
|
| 922 |
+
"loss": 0.7088,
|
| 923 |
+
"step": 10200
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 8.35360908353609,
|
| 927 |
+
"grad_norm": 2.5597479343414307,
|
| 928 |
+
"learning_rate": 5.160608622147084e-06,
|
| 929 |
+
"loss": 0.6943,
|
| 930 |
+
"step": 10300
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 8.43471208434712,
|
| 934 |
+
"grad_norm": 3.5557029247283936,
|
| 935 |
+
"learning_rate": 4.907016060862214e-06,
|
| 936 |
+
"loss": 0.6885,
|
| 937 |
+
"step": 10400
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 8.51581508515815,
|
| 941 |
+
"grad_norm": 5.567288875579834,
|
| 942 |
+
"learning_rate": 4.653423499577345e-06,
|
| 943 |
+
"loss": 0.6871,
|
| 944 |
+
"step": 10500
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 8.51581508515815,
|
| 948 |
+
"eval_cer": 0.07708633453766688,
|
| 949 |
+
"eval_loss": 0.29385408759117126,
|
| 950 |
+
"eval_runtime": 278.0557,
|
| 951 |
+
"eval_samples_per_second": 19.51,
|
| 952 |
+
"eval_steps_per_second": 4.88,
|
| 953 |
+
"eval_wer": 0.38316954771489464,
|
| 954 |
+
"step": 10500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 8.59691808596918,
|
| 958 |
+
"grad_norm": 3.0736474990844727,
|
| 959 |
+
"learning_rate": 4.399830938292477e-06,
|
| 960 |
+
"loss": 0.7077,
|
| 961 |
+
"step": 10600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 8.67802108678021,
|
| 965 |
+
"grad_norm": 3.527780532836914,
|
| 966 |
+
"learning_rate": 4.146238377007608e-06,
|
| 967 |
+
"loss": 0.6993,
|
| 968 |
+
"step": 10700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 8.75912408759124,
|
| 972 |
+
"grad_norm": 5.08736515045166,
|
| 973 |
+
"learning_rate": 3.892645815722739e-06,
|
| 974 |
+
"loss": 0.6987,
|
| 975 |
+
"step": 10800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 8.840227088402271,
|
| 979 |
+
"grad_norm": 2.081068515777588,
|
| 980 |
+
"learning_rate": 3.63905325443787e-06,
|
| 981 |
+
"loss": 0.7117,
|
| 982 |
+
"step": 10900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 8.921330089213301,
|
| 986 |
+
"grad_norm": 3.8801658153533936,
|
| 987 |
+
"learning_rate": 3.385460693153001e-06,
|
| 988 |
+
"loss": 0.6942,
|
| 989 |
+
"step": 11000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 8.921330089213301,
|
| 993 |
+
"eval_cer": 0.07657404340222942,
|
| 994 |
+
"eval_loss": 0.29502201080322266,
|
| 995 |
+
"eval_runtime": 276.8481,
|
| 996 |
+
"eval_samples_per_second": 19.596,
|
| 997 |
+
"eval_steps_per_second": 4.902,
|
| 998 |
+
"eval_wer": 0.38160075775515034,
|
| 999 |
+
"step": 11000
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 9.002433090024331,
|
| 1003 |
+
"grad_norm": 2.015519857406616,
|
| 1004 |
+
"learning_rate": 3.1318681318681323e-06,
|
| 1005 |
+
"loss": 0.7185,
|
| 1006 |
+
"step": 11100
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 9.083536090835361,
|
| 1010 |
+
"grad_norm": 2.2934184074401855,
|
| 1011 |
+
"learning_rate": 2.878275570583263e-06,
|
| 1012 |
+
"loss": 0.6961,
|
| 1013 |
+
"step": 11200
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 9.164639091646391,
|
| 1017 |
+
"grad_norm": 1.9660650491714478,
|
| 1018 |
+
"learning_rate": 2.627218934911243e-06,
|
| 1019 |
+
"loss": 0.6899,
|
| 1020 |
+
"step": 11300
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 9.245742092457421,
|
| 1024 |
+
"grad_norm": 2.4155807495117188,
|
| 1025 |
+
"learning_rate": 2.3736263736263735e-06,
|
| 1026 |
+
"loss": 0.7156,
|
| 1027 |
+
"step": 11400
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 9.326845093268451,
|
| 1031 |
+
"grad_norm": 4.100208759307861,
|
| 1032 |
+
"learning_rate": 2.120033812341505e-06,
|
| 1033 |
+
"loss": 0.6919,
|
| 1034 |
+
"step": 11500
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 9.326845093268451,
|
| 1038 |
+
"eval_cer": 0.07598471600582395,
|
| 1039 |
+
"eval_loss": 0.29096654057502747,
|
| 1040 |
+
"eval_runtime": 277.9639,
|
| 1041 |
+
"eval_samples_per_second": 19.517,
|
| 1042 |
+
"eval_steps_per_second": 4.882,
|
| 1043 |
+
"eval_wer": 0.37813757991948854,
|
| 1044 |
+
"step": 11500
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 9.407948094079481,
|
| 1048 |
+
"grad_norm": 1.8773695230484009,
|
| 1049 |
+
"learning_rate": 1.8664412510566357e-06,
|
| 1050 |
+
"loss": 0.6747,
|
| 1051 |
+
"step": 11600
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 9.489051094890511,
|
| 1055 |
+
"grad_norm": 6.90272855758667,
|
| 1056 |
+
"learning_rate": 1.6128486897717668e-06,
|
| 1057 |
+
"loss": 0.6839,
|
| 1058 |
+
"step": 11700
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 9.570154095701541,
|
| 1062 |
+
"grad_norm": 2.4050042629241943,
|
| 1063 |
+
"learning_rate": 1.3592561284868978e-06,
|
| 1064 |
+
"loss": 0.6812,
|
| 1065 |
+
"step": 11800
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 9.651257096512571,
|
| 1069 |
+
"grad_norm": 2.382669687271118,
|
| 1070 |
+
"learning_rate": 1.1056635672020287e-06,
|
| 1071 |
+
"loss": 0.6902,
|
| 1072 |
+
"step": 11900
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 9.732360097323602,
|
| 1076 |
+
"grad_norm": 4.156976222991943,
|
| 1077 |
+
"learning_rate": 8.520710059171598e-07,
|
| 1078 |
+
"loss": 0.6756,
|
| 1079 |
+
"step": 12000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 9.732360097323602,
|
| 1083 |
+
"eval_cer": 0.07602323413630796,
|
| 1084 |
+
"eval_loss": 0.292733758687973,
|
| 1085 |
+
"eval_runtime": 276.5457,
|
| 1086 |
+
"eval_samples_per_second": 19.617,
|
| 1087 |
+
"eval_steps_per_second": 4.907,
|
| 1088 |
+
"eval_wer": 0.3785223774567843,
|
| 1089 |
+
"step": 12000
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 9.813463098134632,
|
| 1093 |
+
"grad_norm": 3.4550888538360596,
|
| 1094 |
+
"learning_rate": 5.984784446322908e-07,
|
| 1095 |
+
"loss": 0.7084,
|
| 1096 |
+
"step": 12100
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 9.894566098945662,
|
| 1100 |
+
"grad_norm": 2.9977035522460938,
|
| 1101 |
+
"learning_rate": 3.4488588334742185e-07,
|
| 1102 |
+
"loss": 0.6843,
|
| 1103 |
+
"step": 12200
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 9.975669099756692,
|
| 1107 |
+
"grad_norm": 3.0470476150512695,
|
| 1108 |
+
"learning_rate": 9.129332206255284e-08,
|
| 1109 |
+
"loss": 0.702,
|
| 1110 |
+
"step": 12300
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 10.0,
|
| 1114 |
+
"step": 12330,
|
| 1115 |
+
"total_flos": 2.6616913888307843e+19,
|
| 1116 |
+
"train_loss": 1.1441452417737183,
|
| 1117 |
+
"train_runtime": 24098.893,
|
| 1118 |
+
"train_samples_per_second": 8.185,
|
| 1119 |
+
"train_steps_per_second": 0.512
|
| 1120 |
+
}
|
| 1121 |
+
],
|
| 1122 |
+
"logging_steps": 100,
|
| 1123 |
+
"max_steps": 12330,
|
| 1124 |
+
"num_input_tokens_seen": 0,
|
| 1125 |
+
"num_train_epochs": 10,
|
| 1126 |
+
"save_steps": 500,
|
| 1127 |
+
"total_flos": 2.6616913888307843e+19,
|
| 1128 |
+
"train_batch_size": 4,
|
| 1129 |
+
"trial_name": null,
|
| 1130 |
+
"trial_params": null
|
| 1131 |
+
}
|