End of training
Browse files- README.md +7 -2
- all_results.json +14 -1
- eval_results.json +9 -0
- runs/Mar16_11-31-27_dgx-a100/events.out.tfevents.1742127702.dgx-a100.697190.1 +3 -0
- train_results.json +9 -0
- trainer_state.json +923 -0
README.md
CHANGED
|
@@ -1,15 +1,20 @@
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
|
|
|
|
|
|
| 3 |
license: apache-2.0
|
| 4 |
base_model: openai/whisper-tiny
|
| 5 |
tags:
|
|
|
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
datasets:
|
| 8 |
- jv_id_asr_split
|
| 9 |
metrics:
|
| 10 |
- wer
|
| 11 |
model-index:
|
| 12 |
-
- name:
|
| 13 |
results:
|
| 14 |
- task:
|
| 15 |
name: Automatic Speech Recognition
|
|
@@ -29,7 +34,7 @@ model-index:
|
|
| 29 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 30 |
should probably proofread and complete it, then remove this comment. -->
|
| 31 |
|
| 32 |
-
#
|
| 33 |
|
| 34 |
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the jv_id_asr_split dataset.
|
| 35 |
It achieves the following results on the evaluation set:
|
|
|
|
| 1 |
---
|
| 2 |
library_name: transformers
|
| 3 |
+
language:
|
| 4 |
+
- jv
|
| 5 |
license: apache-2.0
|
| 6 |
base_model: openai/whisper-tiny
|
| 7 |
tags:
|
| 8 |
+
- whisper
|
| 9 |
+
- javanese
|
| 10 |
+
- asr
|
| 11 |
- generated_from_trainer
|
| 12 |
datasets:
|
| 13 |
- jv_id_asr_split
|
| 14 |
metrics:
|
| 15 |
- wer
|
| 16 |
model-index:
|
| 17 |
+
- name: Whisper-Tiny-Java-v3
|
| 18 |
results:
|
| 19 |
- task:
|
| 20 |
name: Automatic Speech Recognition
|
|
|
|
| 34 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 35 |
should probably proofread and complete it, then remove this comment. -->
|
| 36 |
|
| 37 |
+
# Whisper-Tiny-Java-v3
|
| 38 |
|
| 39 |
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the jv_id_asr_split dataset.
|
| 40 |
It achieves the following results on the evaluation set:
|
all_results.json
CHANGED
|
@@ -1,8 +1,21 @@
|
|
| 1 |
{
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
"pretrained_eval_loss": 4.039339065551758,
|
| 3 |
"pretrained_eval_model_preparation_time": 0.0024,
|
| 4 |
"pretrained_eval_runtime": 5840.2863,
|
| 5 |
"pretrained_eval_samples_per_second": 3.168,
|
| 6 |
"pretrained_eval_steps_per_second": 0.792,
|
| 7 |
-
"pretrained_eval_wer": 1.301261348618616
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 1.0807262117036798,
|
| 3 |
+
"eval_loss": 0.2979792058467865,
|
| 4 |
+
"eval_runtime": 3148.6396,
|
| 5 |
+
"eval_samples": 18504,
|
| 6 |
+
"eval_samples_per_second": 5.877,
|
| 7 |
+
"eval_steps_per_second": 1.469,
|
| 8 |
+
"eval_wer": 0.2586507557925852,
|
| 9 |
"pretrained_eval_loss": 4.039339065551758,
|
| 10 |
"pretrained_eval_model_preparation_time": 0.0024,
|
| 11 |
"pretrained_eval_runtime": 5840.2863,
|
| 12 |
"pretrained_eval_samples_per_second": 3.168,
|
| 13 |
"pretrained_eval_steps_per_second": 0.792,
|
| 14 |
+
"pretrained_eval_wer": 1.301261348618616,
|
| 15 |
+
"total_flos": 3.93912009474048e+18,
|
| 16 |
+
"train_loss": 0.09938106536865235,
|
| 17 |
+
"train_runtime": 25030.863,
|
| 18 |
+
"train_samples": 148052,
|
| 19 |
+
"train_samples_per_second": 6.392,
|
| 20 |
+
"train_steps_per_second": 0.4
|
| 21 |
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0807262117036798,
|
| 3 |
+
"eval_loss": 0.2979792058467865,
|
| 4 |
+
"eval_runtime": 3148.6396,
|
| 5 |
+
"eval_samples": 18504,
|
| 6 |
+
"eval_samples_per_second": 5.877,
|
| 7 |
+
"eval_steps_per_second": 1.469,
|
| 8 |
+
"eval_wer": 0.2586507557925852
|
| 9 |
+
}
|
runs/Mar16_11-31-27_dgx-a100/events.out.tfevents.1742127702.dgx-a100.697190.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:966f22d843e92926febc2e74c35913c0b321d884675ffde7467ee6f6147e4eca
|
| 3 |
+
size 406
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0807262117036798,
|
| 3 |
+
"total_flos": 3.93912009474048e+18,
|
| 4 |
+
"train_loss": 0.09938106536865235,
|
| 5 |
+
"train_runtime": 25030.863,
|
| 6 |
+
"train_samples": 148052,
|
| 7 |
+
"train_samples_per_second": 6.392,
|
| 8 |
+
"train_steps_per_second": 0.4
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,923 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 10000,
|
| 3 |
+
"best_metric": 0.2586507557925852,
|
| 4 |
+
"best_model_checkpoint": "/home/cluster-dgxa100/slp01/bagas-fine-tune-whisper/whisper-tiny-javanese-openslr-v3/checkpoint-10000",
|
| 5 |
+
"epoch": 1.0807262117036798,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 10000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01080672178094775,
|
| 14 |
+
"grad_norm": 25.346445083618164,
|
| 15 |
+
"learning_rate": 1.94e-06,
|
| 16 |
+
"loss": 3.5433,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0216134435618955,
|
| 21 |
+
"grad_norm": 21.656307220458984,
|
| 22 |
+
"learning_rate": 3.94e-06,
|
| 23 |
+
"loss": 2.0264,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03242016534284325,
|
| 28 |
+
"grad_norm": 18.657211303710938,
|
| 29 |
+
"learning_rate": 5.94e-06,
|
| 30 |
+
"loss": 1.5688,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.043226887123791,
|
| 35 |
+
"grad_norm": 16.42237663269043,
|
| 36 |
+
"learning_rate": 7.94e-06,
|
| 37 |
+
"loss": 1.3214,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.054033608904738746,
|
| 42 |
+
"grad_norm": 18.631206512451172,
|
| 43 |
+
"learning_rate": 9.940000000000001e-06,
|
| 44 |
+
"loss": 1.1788,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.054033608904738746,
|
| 49 |
+
"eval_loss": 0.967095136642456,
|
| 50 |
+
"eval_runtime": 5770.5819,
|
| 51 |
+
"eval_samples_per_second": 3.207,
|
| 52 |
+
"eval_steps_per_second": 0.802,
|
| 53 |
+
"eval_wer": 0.6590292385770924,
|
| 54 |
+
"step": 500
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.0648403306856865,
|
| 58 |
+
"grad_norm": 15.337555885314941,
|
| 59 |
+
"learning_rate": 1.1940000000000001e-05,
|
| 60 |
+
"loss": 1.0627,
|
| 61 |
+
"step": 600
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.07564705246663425,
|
| 65 |
+
"grad_norm": 14.623177528381348,
|
| 66 |
+
"learning_rate": 1.394e-05,
|
| 67 |
+
"loss": 0.9632,
|
| 68 |
+
"step": 700
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.086453774247582,
|
| 72 |
+
"grad_norm": 17.126712799072266,
|
| 73 |
+
"learning_rate": 1.5940000000000003e-05,
|
| 74 |
+
"loss": 0.906,
|
| 75 |
+
"step": 800
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 0.09726049602852975,
|
| 79 |
+
"grad_norm": 16.75067710876465,
|
| 80 |
+
"learning_rate": 1.794e-05,
|
| 81 |
+
"loss": 0.8503,
|
| 82 |
+
"step": 900
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.10806721780947749,
|
| 86 |
+
"grad_norm": 14.265076637268066,
|
| 87 |
+
"learning_rate": 1.9940000000000002e-05,
|
| 88 |
+
"loss": 0.8015,
|
| 89 |
+
"step": 1000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.10806721780947749,
|
| 93 |
+
"eval_loss": 0.6976613402366638,
|
| 94 |
+
"eval_runtime": 5463.0331,
|
| 95 |
+
"eval_samples_per_second": 3.387,
|
| 96 |
+
"eval_steps_per_second": 0.847,
|
| 97 |
+
"eval_wer": 0.5304858499049883,
|
| 98 |
+
"step": 1000
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.11887393959042525,
|
| 102 |
+
"grad_norm": 13.737130165100098,
|
| 103 |
+
"learning_rate": 1.9784444444444446e-05,
|
| 104 |
+
"loss": 0.7589,
|
| 105 |
+
"step": 1100
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.129680661371373,
|
| 109 |
+
"grad_norm": 18.01378631591797,
|
| 110 |
+
"learning_rate": 1.9562222222222225e-05,
|
| 111 |
+
"loss": 0.7589,
|
| 112 |
+
"step": 1200
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.14048738315232073,
|
| 116 |
+
"grad_norm": 11.696120262145996,
|
| 117 |
+
"learning_rate": 1.934e-05,
|
| 118 |
+
"loss": 0.7087,
|
| 119 |
+
"step": 1300
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.1512941049332685,
|
| 123 |
+
"grad_norm": 13.419560432434082,
|
| 124 |
+
"learning_rate": 1.911777777777778e-05,
|
| 125 |
+
"loss": 0.683,
|
| 126 |
+
"step": 1400
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.16210082671421625,
|
| 130 |
+
"grad_norm": 12.753211975097656,
|
| 131 |
+
"learning_rate": 1.8895555555555557e-05,
|
| 132 |
+
"loss": 0.6498,
|
| 133 |
+
"step": 1500
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.16210082671421625,
|
| 137 |
+
"eval_loss": 0.5724753737449646,
|
| 138 |
+
"eval_runtime": 4564.4621,
|
| 139 |
+
"eval_samples_per_second": 4.054,
|
| 140 |
+
"eval_steps_per_second": 1.013,
|
| 141 |
+
"eval_wer": 0.6670133485560569,
|
| 142 |
+
"step": 1500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.172907548495164,
|
| 146 |
+
"grad_norm": 11.64907455444336,
|
| 147 |
+
"learning_rate": 1.8673333333333333e-05,
|
| 148 |
+
"loss": 0.6216,
|
| 149 |
+
"step": 1600
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.18371427027611173,
|
| 153 |
+
"grad_norm": 13.781865119934082,
|
| 154 |
+
"learning_rate": 1.8451111111111113e-05,
|
| 155 |
+
"loss": 0.6138,
|
| 156 |
+
"step": 1700
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.1945209920570595,
|
| 160 |
+
"grad_norm": 12.58388900756836,
|
| 161 |
+
"learning_rate": 1.822888888888889e-05,
|
| 162 |
+
"loss": 0.595,
|
| 163 |
+
"step": 1800
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.20532771383800724,
|
| 167 |
+
"grad_norm": 14.661055564880371,
|
| 168 |
+
"learning_rate": 1.8006666666666668e-05,
|
| 169 |
+
"loss": 0.5938,
|
| 170 |
+
"step": 1900
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.21613443561895498,
|
| 174 |
+
"grad_norm": 11.948161125183105,
|
| 175 |
+
"learning_rate": 1.7784444444444448e-05,
|
| 176 |
+
"loss": 0.5828,
|
| 177 |
+
"step": 2000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.21613443561895498,
|
| 181 |
+
"eval_loss": 0.5093731880187988,
|
| 182 |
+
"eval_runtime": 5328.6402,
|
| 183 |
+
"eval_samples_per_second": 3.473,
|
| 184 |
+
"eval_steps_per_second": 0.868,
|
| 185 |
+
"eval_wer": 0.4828939857208768,
|
| 186 |
+
"step": 2000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 0.22694115739990273,
|
| 190 |
+
"grad_norm": 12.322188377380371,
|
| 191 |
+
"learning_rate": 1.7562222222222224e-05,
|
| 192 |
+
"loss": 0.5752,
|
| 193 |
+
"step": 2100
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.2377478791808505,
|
| 197 |
+
"grad_norm": 17.046159744262695,
|
| 198 |
+
"learning_rate": 1.734e-05,
|
| 199 |
+
"loss": 0.5663,
|
| 200 |
+
"step": 2200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.24855460096179824,
|
| 204 |
+
"grad_norm": 10.154263496398926,
|
| 205 |
+
"learning_rate": 1.711777777777778e-05,
|
| 206 |
+
"loss": 0.537,
|
| 207 |
+
"step": 2300
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.259361322742746,
|
| 211 |
+
"grad_norm": 11.958285331726074,
|
| 212 |
+
"learning_rate": 1.6895555555555556e-05,
|
| 213 |
+
"loss": 0.5246,
|
| 214 |
+
"step": 2400
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.27016804452369375,
|
| 218 |
+
"grad_norm": 10.264266014099121,
|
| 219 |
+
"learning_rate": 1.6673333333333335e-05,
|
| 220 |
+
"loss": 0.5226,
|
| 221 |
+
"step": 2500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.27016804452369375,
|
| 225 |
+
"eval_loss": 0.46415480971336365,
|
| 226 |
+
"eval_runtime": 4645.745,
|
| 227 |
+
"eval_samples_per_second": 3.983,
|
| 228 |
+
"eval_steps_per_second": 0.996,
|
| 229 |
+
"eval_wer": 0.38602898052064843,
|
| 230 |
+
"step": 2500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.28097476630464147,
|
| 234 |
+
"grad_norm": 12.049257278442383,
|
| 235 |
+
"learning_rate": 1.6451111111111115e-05,
|
| 236 |
+
"loss": 0.493,
|
| 237 |
+
"step": 2600
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 0.29178148808558924,
|
| 241 |
+
"grad_norm": 9.821508407592773,
|
| 242 |
+
"learning_rate": 1.622888888888889e-05,
|
| 243 |
+
"loss": 0.5153,
|
| 244 |
+
"step": 2700
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.302588209866537,
|
| 248 |
+
"grad_norm": 10.481095314025879,
|
| 249 |
+
"learning_rate": 1.6006666666666667e-05,
|
| 250 |
+
"loss": 0.5,
|
| 251 |
+
"step": 2800
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.3133949316474847,
|
| 255 |
+
"grad_norm": 10.193309783935547,
|
| 256 |
+
"learning_rate": 1.5784444444444447e-05,
|
| 257 |
+
"loss": 0.5248,
|
| 258 |
+
"step": 2900
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.3242016534284325,
|
| 262 |
+
"grad_norm": 12.328668594360352,
|
| 263 |
+
"learning_rate": 1.5562222222222223e-05,
|
| 264 |
+
"loss": 0.4955,
|
| 265 |
+
"step": 3000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.3242016534284325,
|
| 269 |
+
"eval_loss": 0.4340818226337433,
|
| 270 |
+
"eval_runtime": 4456.9484,
|
| 271 |
+
"eval_samples_per_second": 4.152,
|
| 272 |
+
"eval_steps_per_second": 1.038,
|
| 273 |
+
"eval_wer": 0.39154200455117727,
|
| 274 |
+
"step": 3000
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.33500837520938026,
|
| 278 |
+
"grad_norm": 12.583343505859375,
|
| 279 |
+
"learning_rate": 1.5340000000000002e-05,
|
| 280 |
+
"loss": 0.5082,
|
| 281 |
+
"step": 3100
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.345815096990328,
|
| 285 |
+
"grad_norm": 8.40932846069336,
|
| 286 |
+
"learning_rate": 1.511777777777778e-05,
|
| 287 |
+
"loss": 0.4905,
|
| 288 |
+
"step": 3200
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.35662181877127574,
|
| 292 |
+
"grad_norm": 14.150980949401855,
|
| 293 |
+
"learning_rate": 1.4895555555555556e-05,
|
| 294 |
+
"loss": 0.466,
|
| 295 |
+
"step": 3300
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.36742854055222346,
|
| 299 |
+
"grad_norm": 13.014771461486816,
|
| 300 |
+
"learning_rate": 1.4673333333333336e-05,
|
| 301 |
+
"loss": 0.4788,
|
| 302 |
+
"step": 3400
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.37823526233317123,
|
| 306 |
+
"grad_norm": 11.843710899353027,
|
| 307 |
+
"learning_rate": 1.4451111111111112e-05,
|
| 308 |
+
"loss": 0.4616,
|
| 309 |
+
"step": 3500
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.37823526233317123,
|
| 313 |
+
"eval_loss": 0.4127795398235321,
|
| 314 |
+
"eval_runtime": 4528.1925,
|
| 315 |
+
"eval_samples_per_second": 4.086,
|
| 316 |
+
"eval_steps_per_second": 1.022,
|
| 317 |
+
"eval_wer": 0.35399088200564593,
|
| 318 |
+
"step": 3500
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.389041984114119,
|
| 322 |
+
"grad_norm": 11.520469665527344,
|
| 323 |
+
"learning_rate": 1.422888888888889e-05,
|
| 324 |
+
"loss": 0.4695,
|
| 325 |
+
"step": 3600
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.3998487058950667,
|
| 329 |
+
"grad_norm": 10.21032428741455,
|
| 330 |
+
"learning_rate": 1.400666666666667e-05,
|
| 331 |
+
"loss": 0.47,
|
| 332 |
+
"step": 3700
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.4106554276760145,
|
| 336 |
+
"grad_norm": 9.393896102905273,
|
| 337 |
+
"learning_rate": 1.3784444444444445e-05,
|
| 338 |
+
"loss": 0.4656,
|
| 339 |
+
"step": 3800
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.42146214945696225,
|
| 343 |
+
"grad_norm": 10.503016471862793,
|
| 344 |
+
"learning_rate": 1.3562222222222223e-05,
|
| 345 |
+
"loss": 0.4446,
|
| 346 |
+
"step": 3900
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.43226887123790997,
|
| 350 |
+
"grad_norm": 10.747596740722656,
|
| 351 |
+
"learning_rate": 1.3340000000000001e-05,
|
| 352 |
+
"loss": 0.4474,
|
| 353 |
+
"step": 4000
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.43226887123790997,
|
| 357 |
+
"eval_loss": 0.3900074064731598,
|
| 358 |
+
"eval_runtime": 4858.8536,
|
| 359 |
+
"eval_samples_per_second": 3.808,
|
| 360 |
+
"eval_steps_per_second": 0.952,
|
| 361 |
+
"eval_wer": 0.36136504038974343,
|
| 362 |
+
"step": 4000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.44307559301885774,
|
| 366 |
+
"grad_norm": 13.275285720825195,
|
| 367 |
+
"learning_rate": 1.3117777777777779e-05,
|
| 368 |
+
"loss": 0.4488,
|
| 369 |
+
"step": 4100
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.45388231479980545,
|
| 373 |
+
"grad_norm": 11.318832397460938,
|
| 374 |
+
"learning_rate": 1.2897777777777778e-05,
|
| 375 |
+
"loss": 0.4292,
|
| 376 |
+
"step": 4200
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.4646890365807532,
|
| 380 |
+
"grad_norm": 10.3064546585083,
|
| 381 |
+
"learning_rate": 1.2675555555555557e-05,
|
| 382 |
+
"loss": 0.4302,
|
| 383 |
+
"step": 4300
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.475495758361701,
|
| 387 |
+
"grad_norm": 11.634562492370605,
|
| 388 |
+
"learning_rate": 1.2453333333333335e-05,
|
| 389 |
+
"loss": 0.426,
|
| 390 |
+
"step": 4400
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.4863024801426487,
|
| 394 |
+
"grad_norm": 10.647918701171875,
|
| 395 |
+
"learning_rate": 1.2231111111111111e-05,
|
| 396 |
+
"loss": 0.4387,
|
| 397 |
+
"step": 4500
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.4863024801426487,
|
| 401 |
+
"eval_loss": 0.37359631061553955,
|
| 402 |
+
"eval_runtime": 4990.4878,
|
| 403 |
+
"eval_samples_per_second": 3.708,
|
| 404 |
+
"eval_steps_per_second": 0.927,
|
| 405 |
+
"eval_wer": 0.35633684967821144,
|
| 406 |
+
"step": 4500
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.4971092019235965,
|
| 410 |
+
"grad_norm": 9.396610260009766,
|
| 411 |
+
"learning_rate": 1.200888888888889e-05,
|
| 412 |
+
"loss": 0.4195,
|
| 413 |
+
"step": 4600
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.5079159237045442,
|
| 417 |
+
"grad_norm": 10.845105171203613,
|
| 418 |
+
"learning_rate": 1.1786666666666668e-05,
|
| 419 |
+
"loss": 0.4056,
|
| 420 |
+
"step": 4700
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.518722645485492,
|
| 424 |
+
"grad_norm": 9.404190063476562,
|
| 425 |
+
"learning_rate": 1.1564444444444445e-05,
|
| 426 |
+
"loss": 0.4306,
|
| 427 |
+
"step": 4800
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.5295293672664397,
|
| 431 |
+
"grad_norm": 9.176289558410645,
|
| 432 |
+
"learning_rate": 1.1342222222222224e-05,
|
| 433 |
+
"loss": 0.4239,
|
| 434 |
+
"step": 4900
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.5403360890473875,
|
| 438 |
+
"grad_norm": 10.088706016540527,
|
| 439 |
+
"learning_rate": 1.1120000000000002e-05,
|
| 440 |
+
"loss": 0.4154,
|
| 441 |
+
"step": 5000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.5403360890473875,
|
| 445 |
+
"eval_loss": 0.36057594418525696,
|
| 446 |
+
"eval_runtime": 5945.658,
|
| 447 |
+
"eval_samples_per_second": 3.112,
|
| 448 |
+
"eval_steps_per_second": 0.778,
|
| 449 |
+
"eval_wer": 0.32743452795220485,
|
| 450 |
+
"step": 5000
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.5511428108283353,
|
| 454 |
+
"grad_norm": 9.688194274902344,
|
| 455 |
+
"learning_rate": 1.0897777777777778e-05,
|
| 456 |
+
"loss": 0.4115,
|
| 457 |
+
"step": 5100
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5619495326092829,
|
| 461 |
+
"grad_norm": 9.752260208129883,
|
| 462 |
+
"learning_rate": 1.0675555555555558e-05,
|
| 463 |
+
"loss": 0.3854,
|
| 464 |
+
"step": 5200
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.5727562543902307,
|
| 468 |
+
"grad_norm": 10.447392463684082,
|
| 469 |
+
"learning_rate": 1.0453333333333334e-05,
|
| 470 |
+
"loss": 0.4141,
|
| 471 |
+
"step": 5300
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.5835629761711785,
|
| 475 |
+
"grad_norm": 11.185776710510254,
|
| 476 |
+
"learning_rate": 1.0231111111111112e-05,
|
| 477 |
+
"loss": 0.3924,
|
| 478 |
+
"step": 5400
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.5943696979521262,
|
| 482 |
+
"grad_norm": 10.3914794921875,
|
| 483 |
+
"learning_rate": 1.000888888888889e-05,
|
| 484 |
+
"loss": 0.419,
|
| 485 |
+
"step": 5500
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.5943696979521262,
|
| 489 |
+
"eval_loss": 0.3494803309440613,
|
| 490 |
+
"eval_runtime": 6902.9208,
|
| 491 |
+
"eval_samples_per_second": 2.681,
|
| 492 |
+
"eval_steps_per_second": 0.67,
|
| 493 |
+
"eval_wer": 0.314375307908257,
|
| 494 |
+
"step": 5500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.605176419733074,
|
| 498 |
+
"grad_norm": 11.420536041259766,
|
| 499 |
+
"learning_rate": 9.786666666666667e-06,
|
| 500 |
+
"loss": 0.4096,
|
| 501 |
+
"step": 5600
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.6159831415140217,
|
| 505 |
+
"grad_norm": 9.05328369140625,
|
| 506 |
+
"learning_rate": 9.564444444444445e-06,
|
| 507 |
+
"loss": 0.3917,
|
| 508 |
+
"step": 5700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.6267898632949694,
|
| 512 |
+
"grad_norm": 10.281911849975586,
|
| 513 |
+
"learning_rate": 9.342222222222223e-06,
|
| 514 |
+
"loss": 0.3965,
|
| 515 |
+
"step": 5800
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.6375965850759172,
|
| 519 |
+
"grad_norm": 10.587265014648438,
|
| 520 |
+
"learning_rate": 9.12e-06,
|
| 521 |
+
"loss": 0.374,
|
| 522 |
+
"step": 5900
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.648403306856865,
|
| 526 |
+
"grad_norm": 7.721372127532959,
|
| 527 |
+
"learning_rate": 8.897777777777779e-06,
|
| 528 |
+
"loss": 0.3799,
|
| 529 |
+
"step": 6000
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.648403306856865,
|
| 533 |
+
"eval_loss": 0.3397567868232727,
|
| 534 |
+
"eval_runtime": 7002.8513,
|
| 535 |
+
"eval_samples_per_second": 2.642,
|
| 536 |
+
"eval_steps_per_second": 0.661,
|
| 537 |
+
"eval_wer": 0.2921668139413039,
|
| 538 |
+
"step": 6000
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.6592100286378128,
|
| 542 |
+
"grad_norm": 6.785597324371338,
|
| 543 |
+
"learning_rate": 8.675555555555556e-06,
|
| 544 |
+
"loss": 0.3953,
|
| 545 |
+
"step": 6100
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.6700167504187605,
|
| 549 |
+
"grad_norm": 9.53781509399414,
|
| 550 |
+
"learning_rate": 8.453333333333334e-06,
|
| 551 |
+
"loss": 0.3786,
|
| 552 |
+
"step": 6200
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.6808234721997082,
|
| 556 |
+
"grad_norm": 8.857239723205566,
|
| 557 |
+
"learning_rate": 8.231111111111112e-06,
|
| 558 |
+
"loss": 0.3744,
|
| 559 |
+
"step": 6300
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.691630193980656,
|
| 563 |
+
"grad_norm": 9.638261795043945,
|
| 564 |
+
"learning_rate": 8.00888888888889e-06,
|
| 565 |
+
"loss": 0.3809,
|
| 566 |
+
"step": 6400
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.7024369157616037,
|
| 570 |
+
"grad_norm": 8.304004669189453,
|
| 571 |
+
"learning_rate": 7.786666666666666e-06,
|
| 572 |
+
"loss": 0.3802,
|
| 573 |
+
"step": 6500
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 0.7024369157616037,
|
| 577 |
+
"eval_loss": 0.3289755880832672,
|
| 578 |
+
"eval_runtime": 5885.8991,
|
| 579 |
+
"eval_samples_per_second": 3.144,
|
| 580 |
+
"eval_steps_per_second": 0.786,
|
| 581 |
+
"eval_wer": 0.3044049452998538,
|
| 582 |
+
"step": 6500
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.7132436375425515,
|
| 586 |
+
"grad_norm": 9.978581428527832,
|
| 587 |
+
"learning_rate": 7.564444444444446e-06,
|
| 588 |
+
"loss": 0.3537,
|
| 589 |
+
"step": 6600
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 0.7240503593234993,
|
| 593 |
+
"grad_norm": 10.849929809570312,
|
| 594 |
+
"learning_rate": 7.342222222222223e-06,
|
| 595 |
+
"loss": 0.3762,
|
| 596 |
+
"step": 6700
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.7348570811044469,
|
| 600 |
+
"grad_norm": 11.856138229370117,
|
| 601 |
+
"learning_rate": 7.1200000000000004e-06,
|
| 602 |
+
"loss": 0.3477,
|
| 603 |
+
"step": 6800
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.7456638028853947,
|
| 607 |
+
"grad_norm": 10.761491775512695,
|
| 608 |
+
"learning_rate": 6.897777777777779e-06,
|
| 609 |
+
"loss": 0.361,
|
| 610 |
+
"step": 6900
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.7564705246663425,
|
| 614 |
+
"grad_norm": 9.24421501159668,
|
| 615 |
+
"learning_rate": 6.675555555555556e-06,
|
| 616 |
+
"loss": 0.3611,
|
| 617 |
+
"step": 7000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.7564705246663425,
|
| 621 |
+
"eval_loss": 0.3224972188472748,
|
| 622 |
+
"eval_runtime": 5632.6127,
|
| 623 |
+
"eval_samples_per_second": 3.285,
|
| 624 |
+
"eval_steps_per_second": 0.821,
|
| 625 |
+
"eval_wer": 0.2823372093932546,
|
| 626 |
+
"step": 7000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.7672772464472902,
|
| 630 |
+
"grad_norm": 10.52470874786377,
|
| 631 |
+
"learning_rate": 6.453333333333334e-06,
|
| 632 |
+
"loss": 0.3638,
|
| 633 |
+
"step": 7100
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.778083968228238,
|
| 637 |
+
"grad_norm": 9.080463409423828,
|
| 638 |
+
"learning_rate": 6.231111111111111e-06,
|
| 639 |
+
"loss": 0.3532,
|
| 640 |
+
"step": 7200
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.7888906900091858,
|
| 644 |
+
"grad_norm": 8.789374351501465,
|
| 645 |
+
"learning_rate": 6.00888888888889e-06,
|
| 646 |
+
"loss": 0.3592,
|
| 647 |
+
"step": 7300
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.7996974117901334,
|
| 651 |
+
"grad_norm": 8.97732162475586,
|
| 652 |
+
"learning_rate": 5.7866666666666674e-06,
|
| 653 |
+
"loss": 0.3611,
|
| 654 |
+
"step": 7400
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.8105041335710812,
|
| 658 |
+
"grad_norm": 10.455592155456543,
|
| 659 |
+
"learning_rate": 5.5644444444444444e-06,
|
| 660 |
+
"loss": 0.3548,
|
| 661 |
+
"step": 7500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.8105041335710812,
|
| 665 |
+
"eval_loss": 0.31678903102874756,
|
| 666 |
+
"eval_runtime": 3060.9871,
|
| 667 |
+
"eval_samples_per_second": 6.045,
|
| 668 |
+
"eval_steps_per_second": 1.511,
|
| 669 |
+
"eval_wer": 0.27332869353060313,
|
| 670 |
+
"step": 7500
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.821310855352029,
|
| 674 |
+
"grad_norm": 8.56920051574707,
|
| 675 |
+
"learning_rate": 5.342222222222223e-06,
|
| 676 |
+
"loss": 0.3628,
|
| 677 |
+
"step": 7600
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.8321175771329767,
|
| 681 |
+
"grad_norm": 11.37761402130127,
|
| 682 |
+
"learning_rate": 5.12e-06,
|
| 683 |
+
"loss": 0.3353,
|
| 684 |
+
"step": 7700
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.8429242989139245,
|
| 688 |
+
"grad_norm": 9.396086692810059,
|
| 689 |
+
"learning_rate": 4.897777777777778e-06,
|
| 690 |
+
"loss": 0.3704,
|
| 691 |
+
"step": 7800
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.8537310206948722,
|
| 695 |
+
"grad_norm": 10.0977144241333,
|
| 696 |
+
"learning_rate": 4.677777777777778e-06,
|
| 697 |
+
"loss": 0.364,
|
| 698 |
+
"step": 7900
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.8645377424758199,
|
| 702 |
+
"grad_norm": 8.653088569641113,
|
| 703 |
+
"learning_rate": 4.455555555555555e-06,
|
| 704 |
+
"loss": 0.346,
|
| 705 |
+
"step": 8000
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.8645377424758199,
|
| 709 |
+
"eval_loss": 0.3104597330093384,
|
| 710 |
+
"eval_runtime": 3053.8514,
|
| 711 |
+
"eval_samples_per_second": 6.059,
|
| 712 |
+
"eval_steps_per_second": 1.515,
|
| 713 |
+
"eval_wer": 0.26601709428444076,
|
| 714 |
+
"step": 8000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.8753444642567677,
|
| 718 |
+
"grad_norm": 9.058122634887695,
|
| 719 |
+
"learning_rate": 4.233333333333334e-06,
|
| 720 |
+
"loss": 0.3382,
|
| 721 |
+
"step": 8100
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.8861511860377155,
|
| 725 |
+
"grad_norm": 12.135452270507812,
|
| 726 |
+
"learning_rate": 4.011111111111111e-06,
|
| 727 |
+
"loss": 0.3456,
|
| 728 |
+
"step": 8200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.8969579078186632,
|
| 732 |
+
"grad_norm": 6.601293563842773,
|
| 733 |
+
"learning_rate": 3.7888888888888893e-06,
|
| 734 |
+
"loss": 0.3404,
|
| 735 |
+
"step": 8300
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.9077646295996109,
|
| 739 |
+
"grad_norm": 9.51930046081543,
|
| 740 |
+
"learning_rate": 3.566666666666667e-06,
|
| 741 |
+
"loss": 0.3479,
|
| 742 |
+
"step": 8400
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.9185713513805587,
|
| 746 |
+
"grad_norm": 7.031350135803223,
|
| 747 |
+
"learning_rate": 3.3444444444444445e-06,
|
| 748 |
+
"loss": 0.3547,
|
| 749 |
+
"step": 8500
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.9185713513805587,
|
| 753 |
+
"eval_loss": 0.3063461184501648,
|
| 754 |
+
"eval_runtime": 3070.7291,
|
| 755 |
+
"eval_samples_per_second": 6.026,
|
| 756 |
+
"eval_steps_per_second": 1.506,
|
| 757 |
+
"eval_wer": 0.27081068822871623,
|
| 758 |
+
"step": 8500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.9293780731615064,
|
| 762 |
+
"grad_norm": 11.10822868347168,
|
| 763 |
+
"learning_rate": 3.1222222222222228e-06,
|
| 764 |
+
"loss": 0.3454,
|
| 765 |
+
"step": 8600
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.9401847949424542,
|
| 769 |
+
"grad_norm": 9.607211112976074,
|
| 770 |
+
"learning_rate": 2.9e-06,
|
| 771 |
+
"loss": 0.3319,
|
| 772 |
+
"step": 8700
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.950991516723402,
|
| 776 |
+
"grad_norm": 10.614663124084473,
|
| 777 |
+
"learning_rate": 2.677777777777778e-06,
|
| 778 |
+
"loss": 0.3441,
|
| 779 |
+
"step": 8800
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9617982385043498,
|
| 783 |
+
"grad_norm": 8.344138145446777,
|
| 784 |
+
"learning_rate": 2.455555555555556e-06,
|
| 785 |
+
"loss": 0.3466,
|
| 786 |
+
"step": 8900
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.9726049602852974,
|
| 790 |
+
"grad_norm": 11.955930709838867,
|
| 791 |
+
"learning_rate": 2.2333333333333333e-06,
|
| 792 |
+
"loss": 0.3211,
|
| 793 |
+
"step": 9000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.9726049602852974,
|
| 797 |
+
"eval_loss": 0.30189329385757446,
|
| 798 |
+
"eval_runtime": 3095.8164,
|
| 799 |
+
"eval_samples_per_second": 5.977,
|
| 800 |
+
"eval_steps_per_second": 1.494,
|
| 801 |
+
"eval_wer": 0.28268910454413937,
|
| 802 |
+
"step": 9000
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.9834116820662452,
|
| 806 |
+
"grad_norm": 9.438616752624512,
|
| 807 |
+
"learning_rate": 2.011111111111111e-06,
|
| 808 |
+
"loss": 0.343,
|
| 809 |
+
"step": 9100
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.994218403847193,
|
| 813 |
+
"grad_norm": 10.029309272766113,
|
| 814 |
+
"learning_rate": 1.788888888888889e-06,
|
| 815 |
+
"loss": 0.3582,
|
| 816 |
+
"step": 9200
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.0050791592370454,
|
| 820 |
+
"grad_norm": 9.47360610961914,
|
| 821 |
+
"learning_rate": 1.566666666666667e-06,
|
| 822 |
+
"loss": 0.3024,
|
| 823 |
+
"step": 9300
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.0158858810179932,
|
| 827 |
+
"grad_norm": 9.3403959274292,
|
| 828 |
+
"learning_rate": 1.3444444444444446e-06,
|
| 829 |
+
"loss": 0.2811,
|
| 830 |
+
"step": 9400
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.026692602798941,
|
| 834 |
+
"grad_norm": 9.723664283752441,
|
| 835 |
+
"learning_rate": 1.1222222222222222e-06,
|
| 836 |
+
"loss": 0.2718,
|
| 837 |
+
"step": 9500
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.026692602798941,
|
| 841 |
+
"eval_loss": 0.2989746034145355,
|
| 842 |
+
"eval_runtime": 3189.5179,
|
| 843 |
+
"eval_samples_per_second": 5.802,
|
| 844 |
+
"eval_steps_per_second": 1.45,
|
| 845 |
+
"eval_wer": 0.2659936346077151,
|
| 846 |
+
"step": 9500
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.0374993245798887,
|
| 850 |
+
"grad_norm": 7.739469051361084,
|
| 851 |
+
"learning_rate": 9.000000000000001e-07,
|
| 852 |
+
"loss": 0.2765,
|
| 853 |
+
"step": 9600
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 1.0483060463608365,
|
| 857 |
+
"grad_norm": 8.379693984985352,
|
| 858 |
+
"learning_rate": 6.777777777777779e-07,
|
| 859 |
+
"loss": 0.2872,
|
| 860 |
+
"step": 9700
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.0591127681417842,
|
| 864 |
+
"grad_norm": 8.849838256835938,
|
| 865 |
+
"learning_rate": 4.5555555555555563e-07,
|
| 866 |
+
"loss": 0.2782,
|
| 867 |
+
"step": 9800
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 1.069919489922732,
|
| 871 |
+
"grad_norm": 8.006597518920898,
|
| 872 |
+
"learning_rate": 2.3333333333333336e-07,
|
| 873 |
+
"loss": 0.2673,
|
| 874 |
+
"step": 9900
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 1.0807262117036798,
|
| 878 |
+
"grad_norm": 10.859480857849121,
|
| 879 |
+
"learning_rate": 1.1111111111111112e-08,
|
| 880 |
+
"loss": 0.2859,
|
| 881 |
+
"step": 10000
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.0807262117036798,
|
| 885 |
+
"eval_loss": 0.2979792058467865,
|
| 886 |
+
"eval_runtime": 3174.7879,
|
| 887 |
+
"eval_samples_per_second": 5.828,
|
| 888 |
+
"eval_steps_per_second": 1.457,
|
| 889 |
+
"eval_wer": 0.2586507557925852,
|
| 890 |
+
"step": 10000
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 1.0807262117036798,
|
| 894 |
+
"step": 10000,
|
| 895 |
+
"total_flos": 3.93912009474048e+18,
|
| 896 |
+
"train_loss": 0.09938106536865235,
|
| 897 |
+
"train_runtime": 25030.863,
|
| 898 |
+
"train_samples_per_second": 6.392,
|
| 899 |
+
"train_steps_per_second": 0.4
|
| 900 |
+
}
|
| 901 |
+
],
|
| 902 |
+
"logging_steps": 100,
|
| 903 |
+
"max_steps": 10000,
|
| 904 |
+
"num_input_tokens_seen": 0,
|
| 905 |
+
"num_train_epochs": 2,
|
| 906 |
+
"save_steps": 500,
|
| 907 |
+
"stateful_callbacks": {
|
| 908 |
+
"TrainerControl": {
|
| 909 |
+
"args": {
|
| 910 |
+
"should_epoch_stop": false,
|
| 911 |
+
"should_evaluate": false,
|
| 912 |
+
"should_log": false,
|
| 913 |
+
"should_save": true,
|
| 914 |
+
"should_training_stop": true
|
| 915 |
+
},
|
| 916 |
+
"attributes": {}
|
| 917 |
+
}
|
| 918 |
+
},
|
| 919 |
+
"total_flos": 3.93912009474048e+18,
|
| 920 |
+
"train_batch_size": 8,
|
| 921 |
+
"trial_name": null,
|
| 922 |
+
"trial_params": null
|
| 923 |
+
}
|