End of training
Browse files- README.md +16 -4
- all_results.json +15 -0
- eval_results.json +9 -0
- runs/Jan04_14-43-51_srvrocgpu011.uct.ac.za/events.out.tfevents.1736012714.srvrocgpu011.uct.ac.za +3 -0
- train_results.json +9 -0
- trainer_state.json +1221 -0
README.md
CHANGED
|
@@ -4,11 +4,23 @@ license: apache-2.0
|
|
| 4 |
base_model: openai/whisper-medium
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
|
|
|
|
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
model-index:
|
| 10 |
- name: whisper-medium-swagen-combined-20hrs-model
|
| 11 |
-
results:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
@@ -16,10 +28,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# whisper-medium-swagen-combined-20hrs-model
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Wer: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 4 |
base_model: openai/whisper-medium
|
| 5 |
tags:
|
| 6 |
- generated_from_trainer
|
| 7 |
+
datasets:
|
| 8 |
+
- swagen
|
| 9 |
metrics:
|
| 10 |
- wer
|
| 11 |
model-index:
|
| 12 |
- name: whisper-medium-swagen-combined-20hrs-model
|
| 13 |
+
results:
|
| 14 |
+
- task:
|
| 15 |
+
name: Automatic Speech Recognition
|
| 16 |
+
type: automatic-speech-recognition
|
| 17 |
+
dataset:
|
| 18 |
+
name: swagen
|
| 19 |
+
type: swagen
|
| 20 |
+
metrics:
|
| 21 |
+
- name: Wer
|
| 22 |
+
type: wer
|
| 23 |
+
value: 0.22899756493506493
|
| 24 |
---
|
| 25 |
|
| 26 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
|
|
|
| 28 |
|
| 29 |
# whisper-medium-swagen-combined-20hrs-model
|
| 30 |
|
| 31 |
+
This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the swagen dataset.
|
| 32 |
It achieves the following results on the evaluation set:
|
| 33 |
+
- Loss: 0.3817
|
| 34 |
+
- Wer: 0.2290
|
| 35 |
|
| 36 |
## Model description
|
| 37 |
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.2291021671826625,
|
| 3 |
+
"eval_loss": 0.3816666007041931,
|
| 4 |
+
"eval_runtime": 537.2536,
|
| 5 |
+
"eval_samples": 1087,
|
| 6 |
+
"eval_samples_per_second": 2.023,
|
| 7 |
+
"eval_steps_per_second": 1.013,
|
| 8 |
+
"eval_wer": 0.22899756493506493,
|
| 9 |
+
"total_flos": 2.939138012676096e+19,
|
| 10 |
+
"train_loss": 1.2231303668022155,
|
| 11 |
+
"train_runtime": 17395.8198,
|
| 12 |
+
"train_samples": 12919,
|
| 13 |
+
"train_samples_per_second": 22.279,
|
| 14 |
+
"train_steps_per_second": 2.785
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.2291021671826625,
|
| 3 |
+
"eval_loss": 0.3816666007041931,
|
| 4 |
+
"eval_runtime": 537.2536,
|
| 5 |
+
"eval_samples": 1087,
|
| 6 |
+
"eval_samples_per_second": 2.023,
|
| 7 |
+
"eval_steps_per_second": 1.013,
|
| 8 |
+
"eval_wer": 0.22899756493506493
|
| 9 |
+
}
|
runs/Jan04_14-43-51_srvrocgpu011.uct.ac.za/events.out.tfevents.1736012714.srvrocgpu011.uct.ac.za
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7ed9e80852a6fb5f09a5ab4083790a41f71306aff21dabba9a7f018cce1643d
|
| 3 |
+
size 40
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.2291021671826625,
|
| 3 |
+
"total_flos": 2.939138012676096e+19,
|
| 4 |
+
"train_loss": 1.2231303668022155,
|
| 5 |
+
"train_runtime": 17395.8198,
|
| 6 |
+
"train_samples": 12919,
|
| 7 |
+
"train_samples_per_second": 22.279,
|
| 8 |
+
"train_steps_per_second": 2.785
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1221 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.3816666007041931,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/speech/results/whisper-medium-swagen-combined-20hrs-model/checkpoint-3000",
|
| 4 |
+
"epoch": 2.2291021671826625,
|
| 5 |
+
"eval_steps": 200,
|
| 6 |
+
"global_step": 3600,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.015479876160990712,
|
| 13 |
+
"grad_norm": 138.2083740234375,
|
| 14 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 15 |
+
"loss": 10.5632,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.030959752321981424,
|
| 20 |
+
"grad_norm": 90.87299346923828,
|
| 21 |
+
"learning_rate": 9.000000000000001e-07,
|
| 22 |
+
"loss": 7.9292,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04643962848297214,
|
| 27 |
+
"grad_norm": 79.41549682617188,
|
| 28 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 29 |
+
"loss": 5.8787,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06191950464396285,
|
| 34 |
+
"grad_norm": 68.67957305908203,
|
| 35 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 36 |
+
"loss": 4.1248,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07739938080495357,
|
| 41 |
+
"grad_norm": 65.86703491210938,
|
| 42 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 43 |
+
"loss": 3.5039,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09287925696594428,
|
| 48 |
+
"grad_norm": 62.98320388793945,
|
| 49 |
+
"learning_rate": 2.9e-06,
|
| 50 |
+
"loss": 3.201,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.10835913312693499,
|
| 55 |
+
"grad_norm": 62.89395523071289,
|
| 56 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 57 |
+
"loss": 3.181,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.1238390092879257,
|
| 62 |
+
"grad_norm": 62.861602783203125,
|
| 63 |
+
"learning_rate": 3.900000000000001e-06,
|
| 64 |
+
"loss": 2.7478,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1238390092879257,
|
| 69 |
+
"eval_loss": 0.797576904296875,
|
| 70 |
+
"eval_runtime": 563.0008,
|
| 71 |
+
"eval_samples_per_second": 1.931,
|
| 72 |
+
"eval_steps_per_second": 0.966,
|
| 73 |
+
"eval_wer": 0.4823457792207792,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.1393188854489164,
|
| 78 |
+
"grad_norm": 63.16032409667969,
|
| 79 |
+
"learning_rate": 4.4e-06,
|
| 80 |
+
"loss": 2.6638,
|
| 81 |
+
"step": 225
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.15479876160990713,
|
| 85 |
+
"grad_norm": 62.850440979003906,
|
| 86 |
+
"learning_rate": 4.9000000000000005e-06,
|
| 87 |
+
"loss": 2.1734,
|
| 88 |
+
"step": 250
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.17027863777089783,
|
| 92 |
+
"grad_norm": 52.81850051879883,
|
| 93 |
+
"learning_rate": 5.400000000000001e-06,
|
| 94 |
+
"loss": 2.3938,
|
| 95 |
+
"step": 275
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.18575851393188855,
|
| 99 |
+
"grad_norm": 44.25548553466797,
|
| 100 |
+
"learning_rate": 5.9e-06,
|
| 101 |
+
"loss": 2.2972,
|
| 102 |
+
"step": 300
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.20123839009287925,
|
| 106 |
+
"grad_norm": 68.06084442138672,
|
| 107 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 108 |
+
"loss": 2.02,
|
| 109 |
+
"step": 325
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.21671826625386997,
|
| 113 |
+
"grad_norm": 68.29243469238281,
|
| 114 |
+
"learning_rate": 6.9e-06,
|
| 115 |
+
"loss": 2.0561,
|
| 116 |
+
"step": 350
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.23219814241486067,
|
| 120 |
+
"grad_norm": 65.92505645751953,
|
| 121 |
+
"learning_rate": 7.4e-06,
|
| 122 |
+
"loss": 2.1382,
|
| 123 |
+
"step": 375
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.2476780185758514,
|
| 127 |
+
"grad_norm": 50.10117721557617,
|
| 128 |
+
"learning_rate": 7.9e-06,
|
| 129 |
+
"loss": 1.6879,
|
| 130 |
+
"step": 400
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.2476780185758514,
|
| 134 |
+
"eval_loss": 0.6310324668884277,
|
| 135 |
+
"eval_runtime": 559.0623,
|
| 136 |
+
"eval_samples_per_second": 1.944,
|
| 137 |
+
"eval_steps_per_second": 0.973,
|
| 138 |
+
"eval_wer": 0.38950892857142855,
|
| 139 |
+
"step": 400
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.2631578947368421,
|
| 143 |
+
"grad_norm": 56.55495071411133,
|
| 144 |
+
"learning_rate": 8.400000000000001e-06,
|
| 145 |
+
"loss": 1.9753,
|
| 146 |
+
"step": 425
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.2786377708978328,
|
| 150 |
+
"grad_norm": 65.00907135009766,
|
| 151 |
+
"learning_rate": 8.900000000000001e-06,
|
| 152 |
+
"loss": 1.8005,
|
| 153 |
+
"step": 450
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.29411764705882354,
|
| 157 |
+
"grad_norm": 34.83129119873047,
|
| 158 |
+
"learning_rate": 9.4e-06,
|
| 159 |
+
"loss": 1.7464,
|
| 160 |
+
"step": 475
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.30959752321981426,
|
| 164 |
+
"grad_norm": 69.51799011230469,
|
| 165 |
+
"learning_rate": 9.9e-06,
|
| 166 |
+
"loss": 1.991,
|
| 167 |
+
"step": 500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 0.32507739938080493,
|
| 171 |
+
"grad_norm": 59.508968353271484,
|
| 172 |
+
"learning_rate": 9.995828988529718e-06,
|
| 173 |
+
"loss": 1.9345,
|
| 174 |
+
"step": 525
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 0.34055727554179566,
|
| 178 |
+
"grad_norm": 38.56643295288086,
|
| 179 |
+
"learning_rate": 9.990615224191867e-06,
|
| 180 |
+
"loss": 1.7101,
|
| 181 |
+
"step": 550
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.3560371517027864,
|
| 185 |
+
"grad_norm": 61.955345153808594,
|
| 186 |
+
"learning_rate": 9.985401459854015e-06,
|
| 187 |
+
"loss": 1.8107,
|
| 188 |
+
"step": 575
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.3715170278637771,
|
| 192 |
+
"grad_norm": 44.76349639892578,
|
| 193 |
+
"learning_rate": 9.980187695516164e-06,
|
| 194 |
+
"loss": 1.7394,
|
| 195 |
+
"step": 600
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.3715170278637771,
|
| 199 |
+
"eval_loss": 0.5530644655227661,
|
| 200 |
+
"eval_runtime": 582.5012,
|
| 201 |
+
"eval_samples_per_second": 1.866,
|
| 202 |
+
"eval_steps_per_second": 0.934,
|
| 203 |
+
"eval_wer": 0.3317775974025974,
|
| 204 |
+
"step": 600
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.38699690402476783,
|
| 208 |
+
"grad_norm": 53.88811111450195,
|
| 209 |
+
"learning_rate": 9.974973931178311e-06,
|
| 210 |
+
"loss": 1.4966,
|
| 211 |
+
"step": 625
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.4024767801857585,
|
| 215 |
+
"grad_norm": 36.39348602294922,
|
| 216 |
+
"learning_rate": 9.96976016684046e-06,
|
| 217 |
+
"loss": 1.6524,
|
| 218 |
+
"step": 650
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.4179566563467492,
|
| 222 |
+
"grad_norm": 50.56556701660156,
|
| 223 |
+
"learning_rate": 9.964546402502608e-06,
|
| 224 |
+
"loss": 1.6449,
|
| 225 |
+
"step": 675
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.43343653250773995,
|
| 229 |
+
"grad_norm": 49.1718635559082,
|
| 230 |
+
"learning_rate": 9.959332638164755e-06,
|
| 231 |
+
"loss": 1.6947,
|
| 232 |
+
"step": 700
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.44891640866873067,
|
| 236 |
+
"grad_norm": 44.10185623168945,
|
| 237 |
+
"learning_rate": 9.954118873826903e-06,
|
| 238 |
+
"loss": 1.6989,
|
| 239 |
+
"step": 725
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.46439628482972134,
|
| 243 |
+
"grad_norm": 47.85836410522461,
|
| 244 |
+
"learning_rate": 9.948905109489052e-06,
|
| 245 |
+
"loss": 1.5363,
|
| 246 |
+
"step": 750
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.47987616099071206,
|
| 250 |
+
"grad_norm": 50.4083251953125,
|
| 251 |
+
"learning_rate": 9.9436913451512e-06,
|
| 252 |
+
"loss": 1.5924,
|
| 253 |
+
"step": 775
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.4953560371517028,
|
| 257 |
+
"grad_norm": 68.3183364868164,
|
| 258 |
+
"learning_rate": 9.938477580813348e-06,
|
| 259 |
+
"loss": 1.4248,
|
| 260 |
+
"step": 800
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.4953560371517028,
|
| 264 |
+
"eval_loss": 0.5066032409667969,
|
| 265 |
+
"eval_runtime": 568.3038,
|
| 266 |
+
"eval_samples_per_second": 1.913,
|
| 267 |
+
"eval_steps_per_second": 0.957,
|
| 268 |
+
"eval_wer": 0.2980925324675325,
|
| 269 |
+
"step": 800
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5108359133126935,
|
| 273 |
+
"grad_norm": 39.81132888793945,
|
| 274 |
+
"learning_rate": 9.933263816475496e-06,
|
| 275 |
+
"loss": 1.5851,
|
| 276 |
+
"step": 825
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.5263157894736842,
|
| 280 |
+
"grad_norm": 35.51476287841797,
|
| 281 |
+
"learning_rate": 9.928050052137645e-06,
|
| 282 |
+
"loss": 1.5689,
|
| 283 |
+
"step": 850
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.541795665634675,
|
| 287 |
+
"grad_norm": 43.64626693725586,
|
| 288 |
+
"learning_rate": 9.922836287799792e-06,
|
| 289 |
+
"loss": 1.2399,
|
| 290 |
+
"step": 875
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5572755417956656,
|
| 294 |
+
"grad_norm": 46.96986770629883,
|
| 295 |
+
"learning_rate": 9.91762252346194e-06,
|
| 296 |
+
"loss": 1.4891,
|
| 297 |
+
"step": 900
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5727554179566563,
|
| 301 |
+
"grad_norm": 48.96132278442383,
|
| 302 |
+
"learning_rate": 9.912408759124087e-06,
|
| 303 |
+
"loss": 1.5456,
|
| 304 |
+
"step": 925
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5882352941176471,
|
| 308 |
+
"grad_norm": 46.15045166015625,
|
| 309 |
+
"learning_rate": 9.907194994786236e-06,
|
| 310 |
+
"loss": 1.4521,
|
| 311 |
+
"step": 950
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6037151702786377,
|
| 315 |
+
"grad_norm": 41.92116928100586,
|
| 316 |
+
"learning_rate": 9.901981230448384e-06,
|
| 317 |
+
"loss": 1.4187,
|
| 318 |
+
"step": 975
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6191950464396285,
|
| 322 |
+
"grad_norm": 34.402976989746094,
|
| 323 |
+
"learning_rate": 9.896767466110533e-06,
|
| 324 |
+
"loss": 1.2897,
|
| 325 |
+
"step": 1000
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.6191950464396285,
|
| 329 |
+
"eval_loss": 0.4649743437767029,
|
| 330 |
+
"eval_runtime": 564.2182,
|
| 331 |
+
"eval_samples_per_second": 1.927,
|
| 332 |
+
"eval_steps_per_second": 0.964,
|
| 333 |
+
"eval_wer": 0.2688717532467532,
|
| 334 |
+
"step": 1000
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.6346749226006192,
|
| 338 |
+
"grad_norm": 31.557924270629883,
|
| 339 |
+
"learning_rate": 9.89155370177268e-06,
|
| 340 |
+
"loss": 1.4588,
|
| 341 |
+
"step": 1025
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 0.6501547987616099,
|
| 345 |
+
"grad_norm": 46.03443908691406,
|
| 346 |
+
"learning_rate": 9.88633993743483e-06,
|
| 347 |
+
"loss": 1.3974,
|
| 348 |
+
"step": 1050
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.6656346749226006,
|
| 352 |
+
"grad_norm": 56.12520980834961,
|
| 353 |
+
"learning_rate": 9.881126173096977e-06,
|
| 354 |
+
"loss": 1.4078,
|
| 355 |
+
"step": 1075
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.6811145510835913,
|
| 359 |
+
"grad_norm": 49.19106674194336,
|
| 360 |
+
"learning_rate": 9.875912408759124e-06,
|
| 361 |
+
"loss": 1.2417,
|
| 362 |
+
"step": 1100
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.6965944272445821,
|
| 366 |
+
"grad_norm": 51.47815704345703,
|
| 367 |
+
"learning_rate": 9.870698644421273e-06,
|
| 368 |
+
"loss": 1.4088,
|
| 369 |
+
"step": 1125
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.7120743034055728,
|
| 373 |
+
"grad_norm": 36.05540084838867,
|
| 374 |
+
"learning_rate": 9.865484880083421e-06,
|
| 375 |
+
"loss": 1.4249,
|
| 376 |
+
"step": 1150
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.7275541795665634,
|
| 380 |
+
"grad_norm": 50.757347106933594,
|
| 381 |
+
"learning_rate": 9.860271115745568e-06,
|
| 382 |
+
"loss": 1.3757,
|
| 383 |
+
"step": 1175
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.7430340557275542,
|
| 387 |
+
"grad_norm": 43.31879425048828,
|
| 388 |
+
"learning_rate": 9.855057351407717e-06,
|
| 389 |
+
"loss": 1.357,
|
| 390 |
+
"step": 1200
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.7430340557275542,
|
| 394 |
+
"eval_loss": 0.4457937777042389,
|
| 395 |
+
"eval_runtime": 572.8759,
|
| 396 |
+
"eval_samples_per_second": 1.897,
|
| 397 |
+
"eval_steps_per_second": 0.95,
|
| 398 |
+
"eval_wer": 0.27384334415584416,
|
| 399 |
+
"step": 1200
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.7585139318885449,
|
| 403 |
+
"grad_norm": 52.25995635986328,
|
| 404 |
+
"learning_rate": 9.849843587069865e-06,
|
| 405 |
+
"loss": 1.0826,
|
| 406 |
+
"step": 1225
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.7739938080495357,
|
| 410 |
+
"grad_norm": 43.19371032714844,
|
| 411 |
+
"learning_rate": 9.844629822732014e-06,
|
| 412 |
+
"loss": 1.4315,
|
| 413 |
+
"step": 1250
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.7894736842105263,
|
| 417 |
+
"grad_norm": 28.991748809814453,
|
| 418 |
+
"learning_rate": 9.839416058394161e-06,
|
| 419 |
+
"loss": 1.2291,
|
| 420 |
+
"step": 1275
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.804953560371517,
|
| 424 |
+
"grad_norm": 38.30876922607422,
|
| 425 |
+
"learning_rate": 9.83420229405631e-06,
|
| 426 |
+
"loss": 1.3047,
|
| 427 |
+
"step": 1300
|
| 428 |
+
},
|
| 429 |
+
{
|
| 430 |
+
"epoch": 0.8204334365325078,
|
| 431 |
+
"grad_norm": 44.77728271484375,
|
| 432 |
+
"learning_rate": 9.828988529718458e-06,
|
| 433 |
+
"loss": 1.4421,
|
| 434 |
+
"step": 1325
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.8359133126934984,
|
| 438 |
+
"grad_norm": 64.83729553222656,
|
| 439 |
+
"learning_rate": 9.823774765380605e-06,
|
| 440 |
+
"loss": 1.3138,
|
| 441 |
+
"step": 1350
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.8513931888544891,
|
| 445 |
+
"grad_norm": 36.1328010559082,
|
| 446 |
+
"learning_rate": 9.818561001042753e-06,
|
| 447 |
+
"loss": 1.2132,
|
| 448 |
+
"step": 1375
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.8668730650154799,
|
| 452 |
+
"grad_norm": 40.22114562988281,
|
| 453 |
+
"learning_rate": 9.813347236704902e-06,
|
| 454 |
+
"loss": 1.3443,
|
| 455 |
+
"step": 1400
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.8668730650154799,
|
| 459 |
+
"eval_loss": 0.43931862711906433,
|
| 460 |
+
"eval_runtime": 580.0591,
|
| 461 |
+
"eval_samples_per_second": 1.874,
|
| 462 |
+
"eval_steps_per_second": 0.938,
|
| 463 |
+
"eval_wer": 0.270900974025974,
|
| 464 |
+
"step": 1400
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.8823529411764706,
|
| 468 |
+
"grad_norm": 27.64898681640625,
|
| 469 |
+
"learning_rate": 9.80813347236705e-06,
|
| 470 |
+
"loss": 1.1846,
|
| 471 |
+
"step": 1425
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.8978328173374613,
|
| 475 |
+
"grad_norm": 19.260305404663086,
|
| 476 |
+
"learning_rate": 9.802919708029199e-06,
|
| 477 |
+
"loss": 1.1317,
|
| 478 |
+
"step": 1450
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.913312693498452,
|
| 482 |
+
"grad_norm": 30.223257064819336,
|
| 483 |
+
"learning_rate": 9.797705943691346e-06,
|
| 484 |
+
"loss": 1.3269,
|
| 485 |
+
"step": 1475
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.9287925696594427,
|
| 489 |
+
"grad_norm": 35.81090545654297,
|
| 490 |
+
"learning_rate": 9.792492179353495e-06,
|
| 491 |
+
"loss": 1.0545,
|
| 492 |
+
"step": 1500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.9442724458204335,
|
| 496 |
+
"grad_norm": 32.93012237548828,
|
| 497 |
+
"learning_rate": 9.787278415015643e-06,
|
| 498 |
+
"loss": 1.1677,
|
| 499 |
+
"step": 1525
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.9597523219814241,
|
| 503 |
+
"grad_norm": 46.6026496887207,
|
| 504 |
+
"learning_rate": 9.78206465067779e-06,
|
| 505 |
+
"loss": 1.2468,
|
| 506 |
+
"step": 1550
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.9752321981424149,
|
| 510 |
+
"grad_norm": 29.324565887451172,
|
| 511 |
+
"learning_rate": 9.776850886339937e-06,
|
| 512 |
+
"loss": 1.081,
|
| 513 |
+
"step": 1575
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.9907120743034056,
|
| 517 |
+
"grad_norm": 41.78209686279297,
|
| 518 |
+
"learning_rate": 9.771637122002087e-06,
|
| 519 |
+
"loss": 1.2284,
|
| 520 |
+
"step": 1600
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.9907120743034056,
|
| 524 |
+
"eval_loss": 0.41507846117019653,
|
| 525 |
+
"eval_runtime": 580.4169,
|
| 526 |
+
"eval_samples_per_second": 1.873,
|
| 527 |
+
"eval_steps_per_second": 0.937,
|
| 528 |
+
"eval_wer": 0.27445211038961037,
|
| 529 |
+
"step": 1600
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.0061919504643964,
|
| 533 |
+
"grad_norm": 25.455608367919922,
|
| 534 |
+
"learning_rate": 9.766423357664234e-06,
|
| 535 |
+
"loss": 0.9931,
|
| 536 |
+
"step": 1625
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.021671826625387,
|
| 540 |
+
"grad_norm": 22.430744171142578,
|
| 541 |
+
"learning_rate": 9.761209593326383e-06,
|
| 542 |
+
"loss": 0.6197,
|
| 543 |
+
"step": 1650
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.0371517027863777,
|
| 547 |
+
"grad_norm": 21.164098739624023,
|
| 548 |
+
"learning_rate": 9.75599582898853e-06,
|
| 549 |
+
"loss": 0.6202,
|
| 550 |
+
"step": 1675
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.0526315789473684,
|
| 554 |
+
"grad_norm": 23.99432373046875,
|
| 555 |
+
"learning_rate": 9.75078206465068e-06,
|
| 556 |
+
"loss": 0.7075,
|
| 557 |
+
"step": 1700
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.068111455108359,
|
| 561 |
+
"grad_norm": 32.372291564941406,
|
| 562 |
+
"learning_rate": 9.745568300312827e-06,
|
| 563 |
+
"loss": 0.6919,
|
| 564 |
+
"step": 1725
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.08359133126935,
|
| 568 |
+
"grad_norm": 29.520769119262695,
|
| 569 |
+
"learning_rate": 9.740354535974975e-06,
|
| 570 |
+
"loss": 0.5938,
|
| 571 |
+
"step": 1750
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.0990712074303406,
|
| 575 |
+
"grad_norm": 34.81272888183594,
|
| 576 |
+
"learning_rate": 9.735140771637122e-06,
|
| 577 |
+
"loss": 0.7783,
|
| 578 |
+
"step": 1775
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.1145510835913313,
|
| 582 |
+
"grad_norm": 26.11049461364746,
|
| 583 |
+
"learning_rate": 9.729927007299271e-06,
|
| 584 |
+
"loss": 0.6748,
|
| 585 |
+
"step": 1800
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.1145510835913313,
|
| 589 |
+
"eval_loss": 0.41134530305862427,
|
| 590 |
+
"eval_runtime": 569.9787,
|
| 591 |
+
"eval_samples_per_second": 1.907,
|
| 592 |
+
"eval_steps_per_second": 0.954,
|
| 593 |
+
"eval_wer": 0.2520292207792208,
|
| 594 |
+
"step": 1800
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 1.130030959752322,
|
| 598 |
+
"grad_norm": 31.540014266967773,
|
| 599 |
+
"learning_rate": 9.724713242961419e-06,
|
| 600 |
+
"loss": 0.5986,
|
| 601 |
+
"step": 1825
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.1455108359133126,
|
| 605 |
+
"grad_norm": 29.100893020629883,
|
| 606 |
+
"learning_rate": 9.719499478623568e-06,
|
| 607 |
+
"loss": 0.6972,
|
| 608 |
+
"step": 1850
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.1609907120743035,
|
| 612 |
+
"grad_norm": 24.08401870727539,
|
| 613 |
+
"learning_rate": 9.714285714285715e-06,
|
| 614 |
+
"loss": 0.7194,
|
| 615 |
+
"step": 1875
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 1.1764705882352942,
|
| 619 |
+
"grad_norm": 17.05337142944336,
|
| 620 |
+
"learning_rate": 9.709071949947864e-06,
|
| 621 |
+
"loss": 0.7229,
|
| 622 |
+
"step": 1900
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 1.1919504643962848,
|
| 626 |
+
"grad_norm": 34.69645309448242,
|
| 627 |
+
"learning_rate": 9.703858185610012e-06,
|
| 628 |
+
"loss": 0.7934,
|
| 629 |
+
"step": 1925
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 1.2074303405572755,
|
| 633 |
+
"grad_norm": 23.91748046875,
|
| 634 |
+
"learning_rate": 9.69864442127216e-06,
|
| 635 |
+
"loss": 0.6275,
|
| 636 |
+
"step": 1950
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 1.2229102167182662,
|
| 640 |
+
"grad_norm": 29.427534103393555,
|
| 641 |
+
"learning_rate": 9.693430656934307e-06,
|
| 642 |
+
"loss": 0.6677,
|
| 643 |
+
"step": 1975
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 1.238390092879257,
|
| 647 |
+
"grad_norm": 25.043569564819336,
|
| 648 |
+
"learning_rate": 9.688216892596456e-06,
|
| 649 |
+
"loss": 0.7242,
|
| 650 |
+
"step": 2000
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.238390092879257,
|
| 654 |
+
"eval_loss": 0.4232771694660187,
|
| 655 |
+
"eval_runtime": 576.1378,
|
| 656 |
+
"eval_samples_per_second": 1.887,
|
| 657 |
+
"eval_steps_per_second": 0.944,
|
| 658 |
+
"eval_wer": 0.26704545454545453,
|
| 659 |
+
"step": 2000
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 1.2538699690402477,
|
| 663 |
+
"grad_norm": 37.21471405029297,
|
| 664 |
+
"learning_rate": 9.683003128258603e-06,
|
| 665 |
+
"loss": 0.6917,
|
| 666 |
+
"step": 2025
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 1.2693498452012384,
|
| 670 |
+
"grad_norm": 19.723859786987305,
|
| 671 |
+
"learning_rate": 9.677789363920752e-06,
|
| 672 |
+
"loss": 0.6583,
|
| 673 |
+
"step": 2050
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 1.284829721362229,
|
| 677 |
+
"grad_norm": 25.643213272094727,
|
| 678 |
+
"learning_rate": 9.6725755995829e-06,
|
| 679 |
+
"loss": 0.6923,
|
| 680 |
+
"step": 2075
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 1.3003095975232197,
|
| 684 |
+
"grad_norm": 27.70577621459961,
|
| 685 |
+
"learning_rate": 9.667361835245049e-06,
|
| 686 |
+
"loss": 0.5808,
|
| 687 |
+
"step": 2100
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 1.3157894736842106,
|
| 691 |
+
"grad_norm": 23.764432907104492,
|
| 692 |
+
"learning_rate": 9.662148070907196e-06,
|
| 693 |
+
"loss": 0.5764,
|
| 694 |
+
"step": 2125
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 1.3312693498452013,
|
| 698 |
+
"grad_norm": 32.69343185424805,
|
| 699 |
+
"learning_rate": 9.656934306569344e-06,
|
| 700 |
+
"loss": 0.6372,
|
| 701 |
+
"step": 2150
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 1.346749226006192,
|
| 705 |
+
"grad_norm": 20.59185791015625,
|
| 706 |
+
"learning_rate": 9.651720542231491e-06,
|
| 707 |
+
"loss": 0.5799,
|
| 708 |
+
"step": 2175
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 1.3622291021671826,
|
| 712 |
+
"grad_norm": 20.03351593017578,
|
| 713 |
+
"learning_rate": 9.64650677789364e-06,
|
| 714 |
+
"loss": 0.6472,
|
| 715 |
+
"step": 2200
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 1.3622291021671826,
|
| 719 |
+
"eval_loss": 0.42063385248184204,
|
| 720 |
+
"eval_runtime": 574.5155,
|
| 721 |
+
"eval_samples_per_second": 1.892,
|
| 722 |
+
"eval_steps_per_second": 0.947,
|
| 723 |
+
"eval_wer": 0.25923295454545453,
|
| 724 |
+
"step": 2200
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.3777089783281733,
|
| 728 |
+
"grad_norm": 23.296613693237305,
|
| 729 |
+
"learning_rate": 9.641293013555788e-06,
|
| 730 |
+
"loss": 0.6845,
|
| 731 |
+
"step": 2225
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.3931888544891642,
|
| 735 |
+
"grad_norm": 30.750635147094727,
|
| 736 |
+
"learning_rate": 9.636079249217935e-06,
|
| 737 |
+
"loss": 0.6608,
|
| 738 |
+
"step": 2250
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.4086687306501549,
|
| 742 |
+
"grad_norm": 25.292583465576172,
|
| 743 |
+
"learning_rate": 9.630865484880084e-06,
|
| 744 |
+
"loss": 0.7074,
|
| 745 |
+
"step": 2275
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.4241486068111455,
|
| 749 |
+
"grad_norm": 20.848609924316406,
|
| 750 |
+
"learning_rate": 9.625651720542232e-06,
|
| 751 |
+
"loss": 0.5744,
|
| 752 |
+
"step": 2300
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.4396284829721362,
|
| 756 |
+
"grad_norm": 24.94603729248047,
|
| 757 |
+
"learning_rate": 9.620437956204381e-06,
|
| 758 |
+
"loss": 0.5716,
|
| 759 |
+
"step": 2325
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.4551083591331269,
|
| 763 |
+
"grad_norm": 35.606422424316406,
|
| 764 |
+
"learning_rate": 9.615224191866528e-06,
|
| 765 |
+
"loss": 0.6942,
|
| 766 |
+
"step": 2350
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.4705882352941178,
|
| 770 |
+
"grad_norm": 33.53971481323242,
|
| 771 |
+
"learning_rate": 9.610010427528676e-06,
|
| 772 |
+
"loss": 0.6933,
|
| 773 |
+
"step": 2375
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.4860681114551084,
|
| 777 |
+
"grad_norm": 26.1285457611084,
|
| 778 |
+
"learning_rate": 9.604796663190825e-06,
|
| 779 |
+
"loss": 0.6917,
|
| 780 |
+
"step": 2400
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.4860681114551084,
|
| 784 |
+
"eval_loss": 0.39901289343833923,
|
| 785 |
+
"eval_runtime": 584.9316,
|
| 786 |
+
"eval_samples_per_second": 1.858,
|
| 787 |
+
"eval_steps_per_second": 0.93,
|
| 788 |
+
"eval_wer": 0.27901785714285715,
|
| 789 |
+
"step": 2400
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1.501547987616099,
|
| 793 |
+
"grad_norm": 11.766839027404785,
|
| 794 |
+
"learning_rate": 9.599582898852972e-06,
|
| 795 |
+
"loss": 0.5896,
|
| 796 |
+
"step": 2425
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 1.5170278637770898,
|
| 800 |
+
"grad_norm": 30.828351974487305,
|
| 801 |
+
"learning_rate": 9.59436913451512e-06,
|
| 802 |
+
"loss": 0.7007,
|
| 803 |
+
"step": 2450
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 1.5325077399380804,
|
| 807 |
+
"grad_norm": 21.049291610717773,
|
| 808 |
+
"learning_rate": 9.589155370177269e-06,
|
| 809 |
+
"loss": 0.855,
|
| 810 |
+
"step": 2475
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 1.5479876160990713,
|
| 814 |
+
"grad_norm": 17.50288963317871,
|
| 815 |
+
"learning_rate": 9.583941605839416e-06,
|
| 816 |
+
"loss": 0.5517,
|
| 817 |
+
"step": 2500
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 1.5634674922600618,
|
| 821 |
+
"grad_norm": 40.06394577026367,
|
| 822 |
+
"learning_rate": 9.578727841501565e-06,
|
| 823 |
+
"loss": 0.6987,
|
| 824 |
+
"step": 2525
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 1.5789473684210527,
|
| 828 |
+
"grad_norm": 26.996692657470703,
|
| 829 |
+
"learning_rate": 9.573514077163713e-06,
|
| 830 |
+
"loss": 0.6327,
|
| 831 |
+
"step": 2550
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 1.5944272445820433,
|
| 835 |
+
"grad_norm": 15.66840934753418,
|
| 836 |
+
"learning_rate": 9.56830031282586e-06,
|
| 837 |
+
"loss": 0.6396,
|
| 838 |
+
"step": 2575
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.609907120743034,
|
| 842 |
+
"grad_norm": 21.837093353271484,
|
| 843 |
+
"learning_rate": 9.56308654848801e-06,
|
| 844 |
+
"loss": 0.7172,
|
| 845 |
+
"step": 2600
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1.609907120743034,
|
| 849 |
+
"eval_loss": 0.3971739113330841,
|
| 850 |
+
"eval_runtime": 571.0488,
|
| 851 |
+
"eval_samples_per_second": 1.904,
|
| 852 |
+
"eval_steps_per_second": 0.953,
|
| 853 |
+
"eval_wer": 0.3005275974025974,
|
| 854 |
+
"step": 2600
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.6253869969040249,
|
| 858 |
+
"grad_norm": 22.674137115478516,
|
| 859 |
+
"learning_rate": 9.557872784150157e-06,
|
| 860 |
+
"loss": 0.5473,
|
| 861 |
+
"step": 2625
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 1.6408668730650153,
|
| 865 |
+
"grad_norm": 30.52071762084961,
|
| 866 |
+
"learning_rate": 9.552659019812304e-06,
|
| 867 |
+
"loss": 0.594,
|
| 868 |
+
"step": 2650
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 1.6563467492260062,
|
| 872 |
+
"grad_norm": 4.544656276702881,
|
| 873 |
+
"learning_rate": 9.547445255474453e-06,
|
| 874 |
+
"loss": 0.6123,
|
| 875 |
+
"step": 2675
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 1.671826625386997,
|
| 879 |
+
"grad_norm": 31.003204345703125,
|
| 880 |
+
"learning_rate": 9.542231491136601e-06,
|
| 881 |
+
"loss": 0.6435,
|
| 882 |
+
"step": 2700
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 1.6873065015479876,
|
| 886 |
+
"grad_norm": 24.042932510375977,
|
| 887 |
+
"learning_rate": 9.53701772679875e-06,
|
| 888 |
+
"loss": 0.6526,
|
| 889 |
+
"step": 2725
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 1.7027863777089784,
|
| 893 |
+
"grad_norm": 25.72020149230957,
|
| 894 |
+
"learning_rate": 9.531803962460897e-06,
|
| 895 |
+
"loss": 0.6013,
|
| 896 |
+
"step": 2750
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.718266253869969,
|
| 900 |
+
"grad_norm": 32.10405349731445,
|
| 901 |
+
"learning_rate": 9.526590198123047e-06,
|
| 902 |
+
"loss": 0.7614,
|
| 903 |
+
"step": 2775
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 1.7337461300309598,
|
| 907 |
+
"grad_norm": 28.47552490234375,
|
| 908 |
+
"learning_rate": 9.521376433785194e-06,
|
| 909 |
+
"loss": 0.6105,
|
| 910 |
+
"step": 2800
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 1.7337461300309598,
|
| 914 |
+
"eval_loss": 0.39262855052948,
|
| 915 |
+
"eval_runtime": 546.1891,
|
| 916 |
+
"eval_samples_per_second": 1.99,
|
| 917 |
+
"eval_steps_per_second": 0.996,
|
| 918 |
+
"eval_wer": 0.23133116883116883,
|
| 919 |
+
"step": 2800
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.7492260061919505,
|
| 923 |
+
"grad_norm": 29.18010139465332,
|
| 924 |
+
"learning_rate": 9.516162669447341e-06,
|
| 925 |
+
"loss": 0.7792,
|
| 926 |
+
"step": 2825
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.7647058823529411,
|
| 930 |
+
"grad_norm": 23.199079513549805,
|
| 931 |
+
"learning_rate": 9.510948905109489e-06,
|
| 932 |
+
"loss": 0.5944,
|
| 933 |
+
"step": 2850
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.780185758513932,
|
| 937 |
+
"grad_norm": 19.647172927856445,
|
| 938 |
+
"learning_rate": 9.505735140771638e-06,
|
| 939 |
+
"loss": 0.7278,
|
| 940 |
+
"step": 2875
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.7956656346749225,
|
| 944 |
+
"grad_norm": 26.793537139892578,
|
| 945 |
+
"learning_rate": 9.500521376433785e-06,
|
| 946 |
+
"loss": 0.6646,
|
| 947 |
+
"step": 2900
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.8111455108359134,
|
| 951 |
+
"grad_norm": 30.784910202026367,
|
| 952 |
+
"learning_rate": 9.495307612095935e-06,
|
| 953 |
+
"loss": 0.7125,
|
| 954 |
+
"step": 2925
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.826625386996904,
|
| 958 |
+
"grad_norm": 27.740474700927734,
|
| 959 |
+
"learning_rate": 9.490093847758082e-06,
|
| 960 |
+
"loss": 0.6095,
|
| 961 |
+
"step": 2950
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.8421052631578947,
|
| 965 |
+
"grad_norm": 58.08521270751953,
|
| 966 |
+
"learning_rate": 9.484880083420231e-06,
|
| 967 |
+
"loss": 0.657,
|
| 968 |
+
"step": 2975
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.8575851393188856,
|
| 972 |
+
"grad_norm": 34.06402587890625,
|
| 973 |
+
"learning_rate": 9.479666319082379e-06,
|
| 974 |
+
"loss": 0.7442,
|
| 975 |
+
"step": 3000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.8575851393188856,
|
| 979 |
+
"eval_loss": 0.3816666007041931,
|
| 980 |
+
"eval_runtime": 538.3038,
|
| 981 |
+
"eval_samples_per_second": 2.019,
|
| 982 |
+
"eval_steps_per_second": 1.011,
|
| 983 |
+
"eval_wer": 0.22899756493506493,
|
| 984 |
+
"step": 3000
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.873065015479876,
|
| 988 |
+
"grad_norm": 20.675968170166016,
|
| 989 |
+
"learning_rate": 9.474452554744526e-06,
|
| 990 |
+
"loss": 0.7677,
|
| 991 |
+
"step": 3025
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.888544891640867,
|
| 995 |
+
"grad_norm": 30.081329345703125,
|
| 996 |
+
"learning_rate": 9.469238790406673e-06,
|
| 997 |
+
"loss": 0.592,
|
| 998 |
+
"step": 3050
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 1.9040247678018576,
|
| 1002 |
+
"grad_norm": 26.64177703857422,
|
| 1003 |
+
"learning_rate": 9.464025026068823e-06,
|
| 1004 |
+
"loss": 0.6696,
|
| 1005 |
+
"step": 3075
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 1.9195046439628483,
|
| 1009 |
+
"grad_norm": 28.337804794311523,
|
| 1010 |
+
"learning_rate": 9.45881126173097e-06,
|
| 1011 |
+
"loss": 0.6002,
|
| 1012 |
+
"step": 3100
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.9349845201238391,
|
| 1016 |
+
"grad_norm": 15.685124397277832,
|
| 1017 |
+
"learning_rate": 9.453597497393119e-06,
|
| 1018 |
+
"loss": 0.6714,
|
| 1019 |
+
"step": 3125
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.9504643962848296,
|
| 1023 |
+
"grad_norm": 28.92962646484375,
|
| 1024 |
+
"learning_rate": 9.448383733055267e-06,
|
| 1025 |
+
"loss": 0.6535,
|
| 1026 |
+
"step": 3150
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 1.9659442724458205,
|
| 1030 |
+
"grad_norm": 29.42474365234375,
|
| 1031 |
+
"learning_rate": 9.443169968717416e-06,
|
| 1032 |
+
"loss": 0.5845,
|
| 1033 |
+
"step": 3175
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.9814241486068112,
|
| 1037 |
+
"grad_norm": 17.18292999267578,
|
| 1038 |
+
"learning_rate": 9.437956204379563e-06,
|
| 1039 |
+
"loss": 0.7074,
|
| 1040 |
+
"step": 3200
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.9814241486068112,
|
| 1044 |
+
"eval_loss": 0.3848791718482971,
|
| 1045 |
+
"eval_runtime": 535.7133,
|
| 1046 |
+
"eval_samples_per_second": 2.029,
|
| 1047 |
+
"eval_steps_per_second": 1.015,
|
| 1048 |
+
"eval_wer": 0.2262581168831169,
|
| 1049 |
+
"step": 3200
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 1.9969040247678018,
|
| 1053 |
+
"grad_norm": 30.033981323242188,
|
| 1054 |
+
"learning_rate": 9.43274244004171e-06,
|
| 1055 |
+
"loss": 0.7561,
|
| 1056 |
+
"step": 3225
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.0123839009287927,
|
| 1060 |
+
"grad_norm": 11.292386054992676,
|
| 1061 |
+
"learning_rate": 9.427528675703858e-06,
|
| 1062 |
+
"loss": 0.324,
|
| 1063 |
+
"step": 3250
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.027863777089783,
|
| 1067 |
+
"grad_norm": 14.040888786315918,
|
| 1068 |
+
"learning_rate": 9.422314911366007e-06,
|
| 1069 |
+
"loss": 0.2739,
|
| 1070 |
+
"step": 3275
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.043343653250774,
|
| 1074 |
+
"grad_norm": 10.198466300964355,
|
| 1075 |
+
"learning_rate": 9.417101147028155e-06,
|
| 1076 |
+
"loss": 0.2792,
|
| 1077 |
+
"step": 3300
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.0588235294117645,
|
| 1081 |
+
"grad_norm": 11.23904800415039,
|
| 1082 |
+
"learning_rate": 9.411887382690304e-06,
|
| 1083 |
+
"loss": 0.2426,
|
| 1084 |
+
"step": 3325
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.0743034055727554,
|
| 1088 |
+
"grad_norm": 15.241331100463867,
|
| 1089 |
+
"learning_rate": 9.406673618352451e-06,
|
| 1090 |
+
"loss": 0.2859,
|
| 1091 |
+
"step": 3350
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.0897832817337463,
|
| 1095 |
+
"grad_norm": 12.682251930236816,
|
| 1096 |
+
"learning_rate": 9.4014598540146e-06,
|
| 1097 |
+
"loss": 0.2581,
|
| 1098 |
+
"step": 3375
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.1052631578947367,
|
| 1102 |
+
"grad_norm": 8.495072364807129,
|
| 1103 |
+
"learning_rate": 9.396246089676748e-06,
|
| 1104 |
+
"loss": 0.2659,
|
| 1105 |
+
"step": 3400
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.1052631578947367,
|
| 1109 |
+
"eval_loss": 0.400187611579895,
|
| 1110 |
+
"eval_runtime": 545.9468,
|
| 1111 |
+
"eval_samples_per_second": 1.991,
|
| 1112 |
+
"eval_steps_per_second": 0.996,
|
| 1113 |
+
"eval_wer": 0.24857954545454544,
|
| 1114 |
+
"step": 3400
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 2.1207430340557276,
|
| 1118 |
+
"grad_norm": 10.508811950683594,
|
| 1119 |
+
"learning_rate": 9.391032325338895e-06,
|
| 1120 |
+
"loss": 0.2665,
|
| 1121 |
+
"step": 3425
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 2.136222910216718,
|
| 1125 |
+
"grad_norm": 27.000606536865234,
|
| 1126 |
+
"learning_rate": 9.385818561001043e-06,
|
| 1127 |
+
"loss": 0.2768,
|
| 1128 |
+
"step": 3450
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 2.151702786377709,
|
| 1132 |
+
"grad_norm": 25.08758544921875,
|
| 1133 |
+
"learning_rate": 9.380604796663192e-06,
|
| 1134 |
+
"loss": 0.2693,
|
| 1135 |
+
"step": 3475
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 2.1671826625387,
|
| 1139 |
+
"grad_norm": 16.024877548217773,
|
| 1140 |
+
"learning_rate": 9.375391032325339e-06,
|
| 1141 |
+
"loss": 0.2919,
|
| 1142 |
+
"step": 3500
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 2.1826625386996903,
|
| 1146 |
+
"grad_norm": 11.162302017211914,
|
| 1147 |
+
"learning_rate": 9.370177267987488e-06,
|
| 1148 |
+
"loss": 0.2464,
|
| 1149 |
+
"step": 3525
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 2.198142414860681,
|
| 1153 |
+
"grad_norm": 14.112415313720703,
|
| 1154 |
+
"learning_rate": 9.364963503649636e-06,
|
| 1155 |
+
"loss": 0.3192,
|
| 1156 |
+
"step": 3550
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 2.2136222910216716,
|
| 1160 |
+
"grad_norm": 14.262118339538574,
|
| 1161 |
+
"learning_rate": 9.359749739311785e-06,
|
| 1162 |
+
"loss": 0.2935,
|
| 1163 |
+
"step": 3575
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 2.2291021671826625,
|
| 1167 |
+
"grad_norm": 14.042037963867188,
|
| 1168 |
+
"learning_rate": 9.354535974973932e-06,
|
| 1169 |
+
"loss": 0.3132,
|
| 1170 |
+
"step": 3600
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 2.2291021671826625,
|
| 1174 |
+
"eval_loss": 0.39583250880241394,
|
| 1175 |
+
"eval_runtime": 534.4305,
|
| 1176 |
+
"eval_samples_per_second": 2.034,
|
| 1177 |
+
"eval_steps_per_second": 1.018,
|
| 1178 |
+
"eval_wer": 0.22280844155844157,
|
| 1179 |
+
"step": 3600
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.2291021671826625,
|
| 1183 |
+
"step": 3600,
|
| 1184 |
+
"total_flos": 2.939138012676096e+19,
|
| 1185 |
+
"train_loss": 1.2231303668022155,
|
| 1186 |
+
"train_runtime": 17395.8198,
|
| 1187 |
+
"train_samples_per_second": 22.279,
|
| 1188 |
+
"train_steps_per_second": 2.785
|
| 1189 |
+
}
|
| 1190 |
+
],
|
| 1191 |
+
"logging_steps": 25,
|
| 1192 |
+
"max_steps": 48450,
|
| 1193 |
+
"num_input_tokens_seen": 0,
|
| 1194 |
+
"num_train_epochs": 30,
|
| 1195 |
+
"save_steps": 200,
|
| 1196 |
+
"stateful_callbacks": {
|
| 1197 |
+
"EarlyStoppingCallback": {
|
| 1198 |
+
"args": {
|
| 1199 |
+
"early_stopping_patience": 3,
|
| 1200 |
+
"early_stopping_threshold": 0.0
|
| 1201 |
+
},
|
| 1202 |
+
"attributes": {
|
| 1203 |
+
"early_stopping_patience_counter": 3
|
| 1204 |
+
}
|
| 1205 |
+
},
|
| 1206 |
+
"TrainerControl": {
|
| 1207 |
+
"args": {
|
| 1208 |
+
"should_epoch_stop": false,
|
| 1209 |
+
"should_evaluate": false,
|
| 1210 |
+
"should_log": false,
|
| 1211 |
+
"should_save": true,
|
| 1212 |
+
"should_training_stop": true
|
| 1213 |
+
},
|
| 1214 |
+
"attributes": {}
|
| 1215 |
+
}
|
| 1216 |
+
},
|
| 1217 |
+
"total_flos": 2.939138012676096e+19,
|
| 1218 |
+
"train_batch_size": 2,
|
| 1219 |
+
"trial_name": null,
|
| 1220 |
+
"trial_params": null
|
| 1221 |
+
}
|