End of training
Browse files- README.md +12 -9
- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Sep20_15-48-09_ef8d41a2b7b7/events.out.tfevents.1726864201.ef8d41a2b7b7.174.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1487 -0
README.md
CHANGED
@@ -1,39 +1,42 @@
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
base_model: openai/whisper-small
|
4 |
tags:
|
|
|
5 |
- generated_from_trainer
|
6 |
datasets:
|
7 |
-
- common_voice_17_0
|
8 |
metrics:
|
9 |
- wer
|
10 |
model-index:
|
11 |
-
- name:
|
12 |
results:
|
13 |
- task:
|
14 |
name: Automatic Speech Recognition
|
15 |
type: automatic-speech-recognition
|
16 |
dataset:
|
17 |
-
name: common_voice_17_0
|
18 |
-
type: common_voice_17_0
|
19 |
config: pt
|
20 |
split: test
|
21 |
args: pt
|
22 |
metrics:
|
23 |
- name: Wer
|
24 |
type: wer
|
25 |
-
value: 10.
|
26 |
---
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
should probably proofread and complete it, then remove this comment. -->
|
30 |
|
31 |
-
#
|
32 |
|
33 |
-
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
|
34 |
It achieves the following results on the evaluation set:
|
35 |
-
- Loss: 0.
|
36 |
-
- Wer: 10.
|
37 |
|
38 |
## Model description
|
39 |
|
|
|
1 |
---
|
2 |
+
language:
|
3 |
+
- pt
|
4 |
license: apache-2.0
|
5 |
base_model: openai/whisper-small
|
6 |
tags:
|
7 |
+
- whisper-event
|
8 |
- generated_from_trainer
|
9 |
datasets:
|
10 |
+
- mozilla-foundation/common_voice_17_0
|
11 |
metrics:
|
12 |
- wer
|
13 |
model-index:
|
14 |
+
- name: Whisper Small Mixed-Portuguese
|
15 |
results:
|
16 |
- task:
|
17 |
name: Automatic Speech Recognition
|
18 |
type: automatic-speech-recognition
|
19 |
dataset:
|
20 |
+
name: mozilla-foundation/common_voice_17_0 pt
|
21 |
+
type: mozilla-foundation/common_voice_17_0
|
22 |
config: pt
|
23 |
split: test
|
24 |
args: pt
|
25 |
metrics:
|
26 |
- name: Wer
|
27 |
type: wer
|
28 |
+
value: 10.634930784232232
|
29 |
---
|
30 |
|
31 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
32 |
should probably proofread and complete it, then remove this comment. -->
|
33 |
|
34 |
+
# Whisper Small Mixed-Portuguese
|
35 |
|
36 |
+
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the mozilla-foundation/common_voice_17_0 pt dataset.
|
37 |
It achieves the following results on the evaluation set:
|
38 |
+
- Loss: 0.2000
|
39 |
+
- Wer: 10.6349
|
40 |
|
41 |
## Model description
|
42 |
|
all_results.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.062,
|
3 |
+
"eval_loss": 0.2000385820865631,
|
4 |
+
"eval_runtime": 386.067,
|
5 |
+
"eval_samples_per_second": 24.522,
|
6 |
+
"eval_steps_per_second": 3.067,
|
7 |
+
"eval_wer": 10.634930784232232,
|
8 |
+
"total_flos": 9.231442936676352e+19,
|
9 |
+
"train_loss": 0.16698810555934906,
|
10 |
+
"train_runtime": 16450.2071,
|
11 |
+
"train_samples_per_second": 19.453,
|
12 |
+
"train_steps_per_second": 0.304
|
13 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.062,
|
3 |
+
"eval_loss": 0.2000385820865631,
|
4 |
+
"eval_runtime": 386.067,
|
5 |
+
"eval_samples_per_second": 24.522,
|
6 |
+
"eval_steps_per_second": 3.067,
|
7 |
+
"eval_wer": 10.634930784232232
|
8 |
+
}
|
runs/Sep20_15-48-09_ef8d41a2b7b7/events.out.tfevents.1726864201.ef8d41a2b7b7.174.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7e81cdf2741c282fb18771378dd7fd86363d388848b3e8b6e0de886a21a24864
|
3 |
+
size 406
|
train_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 2.062,
|
3 |
+
"total_flos": 9.231442936676352e+19,
|
4 |
+
"train_loss": 0.16698810555934906,
|
5 |
+
"train_runtime": 16450.2071,
|
6 |
+
"train_samples_per_second": 19.453,
|
7 |
+
"train_steps_per_second": 0.304
|
8 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1487 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 10.634930784232232,
|
3 |
+
"best_model_checkpoint": "./whisper-small-mix-pt/checkpoint-4000",
|
4 |
+
"epoch": 2.062,
|
5 |
+
"eval_steps": 1000,
|
6 |
+
"global_step": 5000,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.005,
|
13 |
+
"grad_norm": 8.937539100646973,
|
14 |
+
"learning_rate": 4.6000000000000004e-07,
|
15 |
+
"loss": 0.8175,
|
16 |
+
"step": 25
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.01,
|
20 |
+
"grad_norm": 4.877315044403076,
|
21 |
+
"learning_rate": 9.600000000000001e-07,
|
22 |
+
"loss": 0.7318,
|
23 |
+
"step": 50
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.015,
|
27 |
+
"grad_norm": 4.961461544036865,
|
28 |
+
"learning_rate": 1.46e-06,
|
29 |
+
"loss": 1.0495,
|
30 |
+
"step": 75
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.02,
|
34 |
+
"grad_norm": 4.423418045043945,
|
35 |
+
"learning_rate": 1.9600000000000003e-06,
|
36 |
+
"loss": 0.7303,
|
37 |
+
"step": 100
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.025,
|
41 |
+
"grad_norm": 3.701725482940674,
|
42 |
+
"learning_rate": 2.46e-06,
|
43 |
+
"loss": 0.5486,
|
44 |
+
"step": 125
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.03,
|
48 |
+
"grad_norm": 3.901665449142456,
|
49 |
+
"learning_rate": 2.96e-06,
|
50 |
+
"loss": 0.4327,
|
51 |
+
"step": 150
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.035,
|
55 |
+
"grad_norm": 2.7568914890289307,
|
56 |
+
"learning_rate": 3.46e-06,
|
57 |
+
"loss": 0.3445,
|
58 |
+
"step": 175
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.04,
|
62 |
+
"grad_norm": 3.0615532398223877,
|
63 |
+
"learning_rate": 3.96e-06,
|
64 |
+
"loss": 0.317,
|
65 |
+
"step": 200
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.045,
|
69 |
+
"grad_norm": 2.577845573425293,
|
70 |
+
"learning_rate": 4.4600000000000005e-06,
|
71 |
+
"loss": 0.2995,
|
72 |
+
"step": 225
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.05,
|
76 |
+
"grad_norm": 2.810206890106201,
|
77 |
+
"learning_rate": 4.960000000000001e-06,
|
78 |
+
"loss": 0.2701,
|
79 |
+
"step": 250
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.055,
|
83 |
+
"grad_norm": 2.9507219791412354,
|
84 |
+
"learning_rate": 5.460000000000001e-06,
|
85 |
+
"loss": 0.2275,
|
86 |
+
"step": 275
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.06,
|
90 |
+
"grad_norm": 2.8784899711608887,
|
91 |
+
"learning_rate": 5.9600000000000005e-06,
|
92 |
+
"loss": 0.4164,
|
93 |
+
"step": 300
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.065,
|
97 |
+
"grad_norm": 2.7221858501434326,
|
98 |
+
"learning_rate": 6.460000000000001e-06,
|
99 |
+
"loss": 0.2831,
|
100 |
+
"step": 325
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.07,
|
104 |
+
"grad_norm": 2.760021209716797,
|
105 |
+
"learning_rate": 6.96e-06,
|
106 |
+
"loss": 0.2202,
|
107 |
+
"step": 350
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.075,
|
111 |
+
"grad_norm": 2.8915979862213135,
|
112 |
+
"learning_rate": 7.4600000000000006e-06,
|
113 |
+
"loss": 0.2165,
|
114 |
+
"step": 375
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.08,
|
118 |
+
"grad_norm": 2.7558434009552,
|
119 |
+
"learning_rate": 7.960000000000002e-06,
|
120 |
+
"loss": 0.207,
|
121 |
+
"step": 400
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.085,
|
125 |
+
"grad_norm": 2.5287721157073975,
|
126 |
+
"learning_rate": 8.46e-06,
|
127 |
+
"loss": 0.2529,
|
128 |
+
"step": 425
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.09,
|
132 |
+
"grad_norm": 3.026515483856201,
|
133 |
+
"learning_rate": 8.96e-06,
|
134 |
+
"loss": 0.2134,
|
135 |
+
"step": 450
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.095,
|
139 |
+
"grad_norm": 2.3035366535186768,
|
140 |
+
"learning_rate": 9.460000000000001e-06,
|
141 |
+
"loss": 0.1968,
|
142 |
+
"step": 475
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.1,
|
146 |
+
"grad_norm": 2.298630952835083,
|
147 |
+
"learning_rate": 9.960000000000001e-06,
|
148 |
+
"loss": 0.1881,
|
149 |
+
"step": 500
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.105,
|
153 |
+
"grad_norm": 2.209261417388916,
|
154 |
+
"learning_rate": 9.94888888888889e-06,
|
155 |
+
"loss": 0.162,
|
156 |
+
"step": 525
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.11,
|
160 |
+
"grad_norm": 2.819951295852661,
|
161 |
+
"learning_rate": 9.893333333333334e-06,
|
162 |
+
"loss": 0.1646,
|
163 |
+
"step": 550
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.115,
|
167 |
+
"grad_norm": 2.437563896179199,
|
168 |
+
"learning_rate": 9.837777777777778e-06,
|
169 |
+
"loss": 0.1705,
|
170 |
+
"step": 575
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.12,
|
174 |
+
"grad_norm": 2.5952935218811035,
|
175 |
+
"learning_rate": 9.782222222222222e-06,
|
176 |
+
"loss": 0.1938,
|
177 |
+
"step": 600
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.125,
|
181 |
+
"grad_norm": 2.4692909717559814,
|
182 |
+
"learning_rate": 9.726666666666668e-06,
|
183 |
+
"loss": 0.1869,
|
184 |
+
"step": 625
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 0.13,
|
188 |
+
"grad_norm": 2.4430558681488037,
|
189 |
+
"learning_rate": 9.671111111111112e-06,
|
190 |
+
"loss": 0.1788,
|
191 |
+
"step": 650
|
192 |
+
},
|
193 |
+
{
|
194 |
+
"epoch": 0.135,
|
195 |
+
"grad_norm": 2.2733707427978516,
|
196 |
+
"learning_rate": 9.615555555555558e-06,
|
197 |
+
"loss": 0.1934,
|
198 |
+
"step": 675
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 0.14,
|
202 |
+
"grad_norm": 2.256990909576416,
|
203 |
+
"learning_rate": 9.56e-06,
|
204 |
+
"loss": 0.1622,
|
205 |
+
"step": 700
|
206 |
+
},
|
207 |
+
{
|
208 |
+
"epoch": 0.145,
|
209 |
+
"grad_norm": 2.216128349304199,
|
210 |
+
"learning_rate": 9.504444444444446e-06,
|
211 |
+
"loss": 0.157,
|
212 |
+
"step": 725
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 0.15,
|
216 |
+
"grad_norm": 2.0830190181732178,
|
217 |
+
"learning_rate": 9.44888888888889e-06,
|
218 |
+
"loss": 0.169,
|
219 |
+
"step": 750
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.155,
|
223 |
+
"grad_norm": 2.085129737854004,
|
224 |
+
"learning_rate": 9.393333333333334e-06,
|
225 |
+
"loss": 0.1548,
|
226 |
+
"step": 775
|
227 |
+
},
|
228 |
+
{
|
229 |
+
"epoch": 0.16,
|
230 |
+
"grad_norm": 2.3641719818115234,
|
231 |
+
"learning_rate": 9.33777777777778e-06,
|
232 |
+
"loss": 0.1722,
|
233 |
+
"step": 800
|
234 |
+
},
|
235 |
+
{
|
236 |
+
"epoch": 0.165,
|
237 |
+
"grad_norm": 3.0489625930786133,
|
238 |
+
"learning_rate": 9.282222222222222e-06,
|
239 |
+
"loss": 0.2306,
|
240 |
+
"step": 825
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 0.17,
|
244 |
+
"grad_norm": 2.5903120040893555,
|
245 |
+
"learning_rate": 9.226666666666668e-06,
|
246 |
+
"loss": 0.2255,
|
247 |
+
"step": 850
|
248 |
+
},
|
249 |
+
{
|
250 |
+
"epoch": 0.175,
|
251 |
+
"grad_norm": 2.470798969268799,
|
252 |
+
"learning_rate": 9.171111111111112e-06,
|
253 |
+
"loss": 0.1983,
|
254 |
+
"step": 875
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 0.18,
|
258 |
+
"grad_norm": 1.7610605955123901,
|
259 |
+
"learning_rate": 9.115555555555556e-06,
|
260 |
+
"loss": 0.1991,
|
261 |
+
"step": 900
|
262 |
+
},
|
263 |
+
{
|
264 |
+
"epoch": 0.185,
|
265 |
+
"grad_norm": 2.5426924228668213,
|
266 |
+
"learning_rate": 9.060000000000001e-06,
|
267 |
+
"loss": 0.1584,
|
268 |
+
"step": 925
|
269 |
+
},
|
270 |
+
{
|
271 |
+
"epoch": 0.19,
|
272 |
+
"grad_norm": 2.5950539112091064,
|
273 |
+
"learning_rate": 9.004444444444445e-06,
|
274 |
+
"loss": 0.1707,
|
275 |
+
"step": 950
|
276 |
+
},
|
277 |
+
{
|
278 |
+
"epoch": 0.195,
|
279 |
+
"grad_norm": 2.2430670261383057,
|
280 |
+
"learning_rate": 8.94888888888889e-06,
|
281 |
+
"loss": 0.1682,
|
282 |
+
"step": 975
|
283 |
+
},
|
284 |
+
{
|
285 |
+
"epoch": 0.2,
|
286 |
+
"grad_norm": 2.1364951133728027,
|
287 |
+
"learning_rate": 8.893333333333333e-06,
|
288 |
+
"loss": 0.1575,
|
289 |
+
"step": 1000
|
290 |
+
},
|
291 |
+
{
|
292 |
+
"epoch": 0.2,
|
293 |
+
"eval_loss": 0.21252164244651794,
|
294 |
+
"eval_runtime": 385.7404,
|
295 |
+
"eval_samples_per_second": 24.542,
|
296 |
+
"eval_steps_per_second": 3.069,
|
297 |
+
"eval_wer": 12.185491863550626,
|
298 |
+
"step": 1000
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 0.205,
|
302 |
+
"grad_norm": 2.343320846557617,
|
303 |
+
"learning_rate": 8.83777777777778e-06,
|
304 |
+
"loss": 0.1679,
|
305 |
+
"step": 1025
|
306 |
+
},
|
307 |
+
{
|
308 |
+
"epoch": 0.21,
|
309 |
+
"grad_norm": 2.078854560852051,
|
310 |
+
"learning_rate": 8.782222222222223e-06,
|
311 |
+
"loss": 0.1688,
|
312 |
+
"step": 1050
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"epoch": 0.215,
|
316 |
+
"grad_norm": 2.009870767593384,
|
317 |
+
"learning_rate": 8.726666666666667e-06,
|
318 |
+
"loss": 0.144,
|
319 |
+
"step": 1075
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"epoch": 0.22,
|
323 |
+
"grad_norm": 1.4615126848220825,
|
324 |
+
"learning_rate": 8.671111111111113e-06,
|
325 |
+
"loss": 0.127,
|
326 |
+
"step": 1100
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 0.225,
|
330 |
+
"grad_norm": 2.175366163253784,
|
331 |
+
"learning_rate": 8.615555555555555e-06,
|
332 |
+
"loss": 0.1191,
|
333 |
+
"step": 1125
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 0.23,
|
337 |
+
"grad_norm": 1.3462347984313965,
|
338 |
+
"learning_rate": 8.560000000000001e-06,
|
339 |
+
"loss": 0.1214,
|
340 |
+
"step": 1150
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 0.235,
|
344 |
+
"grad_norm": 1.7791441679000854,
|
345 |
+
"learning_rate": 8.504444444444445e-06,
|
346 |
+
"loss": 0.1213,
|
347 |
+
"step": 1175
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 0.24,
|
351 |
+
"grad_norm": 3.0184803009033203,
|
352 |
+
"learning_rate": 8.448888888888889e-06,
|
353 |
+
"loss": 0.3185,
|
354 |
+
"step": 1200
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 0.245,
|
358 |
+
"grad_norm": 2.1637368202209473,
|
359 |
+
"learning_rate": 8.393333333333335e-06,
|
360 |
+
"loss": 0.2747,
|
361 |
+
"step": 1225
|
362 |
+
},
|
363 |
+
{
|
364 |
+
"epoch": 0.25,
|
365 |
+
"grad_norm": 3.534830093383789,
|
366 |
+
"learning_rate": 8.337777777777777e-06,
|
367 |
+
"loss": 0.1535,
|
368 |
+
"step": 1250
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 0.255,
|
372 |
+
"grad_norm": 3.3045480251312256,
|
373 |
+
"learning_rate": 8.282222222222223e-06,
|
374 |
+
"loss": 0.6384,
|
375 |
+
"step": 1275
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 0.26,
|
379 |
+
"grad_norm": 2.566434383392334,
|
380 |
+
"learning_rate": 8.226666666666667e-06,
|
381 |
+
"loss": 0.4608,
|
382 |
+
"step": 1300
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 0.265,
|
386 |
+
"grad_norm": 3.436089038848877,
|
387 |
+
"learning_rate": 8.171111111111113e-06,
|
388 |
+
"loss": 0.3286,
|
389 |
+
"step": 1325
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 0.27,
|
393 |
+
"grad_norm": 2.8783295154571533,
|
394 |
+
"learning_rate": 8.115555555555557e-06,
|
395 |
+
"loss": 0.3508,
|
396 |
+
"step": 1350
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 0.275,
|
400 |
+
"grad_norm": 3.244107723236084,
|
401 |
+
"learning_rate": 8.06e-06,
|
402 |
+
"loss": 0.3174,
|
403 |
+
"step": 1375
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 0.28,
|
407 |
+
"grad_norm": 2.4489357471466064,
|
408 |
+
"learning_rate": 8.004444444444445e-06,
|
409 |
+
"loss": 0.2287,
|
410 |
+
"step": 1400
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 0.285,
|
414 |
+
"grad_norm": 2.270322561264038,
|
415 |
+
"learning_rate": 7.948888888888889e-06,
|
416 |
+
"loss": 0.1828,
|
417 |
+
"step": 1425
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 0.29,
|
421 |
+
"grad_norm": 2.105621814727783,
|
422 |
+
"learning_rate": 7.893333333333335e-06,
|
423 |
+
"loss": 0.1336,
|
424 |
+
"step": 1450
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 0.295,
|
428 |
+
"grad_norm": 1.8718715906143188,
|
429 |
+
"learning_rate": 7.837777777777779e-06,
|
430 |
+
"loss": 0.1358,
|
431 |
+
"step": 1475
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"epoch": 0.3,
|
435 |
+
"grad_norm": 1.595679759979248,
|
436 |
+
"learning_rate": 7.782222222222223e-06,
|
437 |
+
"loss": 0.1244,
|
438 |
+
"step": 1500
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"epoch": 0.305,
|
442 |
+
"grad_norm": 1.979271411895752,
|
443 |
+
"learning_rate": 7.726666666666667e-06,
|
444 |
+
"loss": 0.1222,
|
445 |
+
"step": 1525
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"epoch": 0.31,
|
449 |
+
"grad_norm": 2.024198532104492,
|
450 |
+
"learning_rate": 7.67111111111111e-06,
|
451 |
+
"loss": 0.1274,
|
452 |
+
"step": 1550
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 0.315,
|
456 |
+
"grad_norm": 1.9602785110473633,
|
457 |
+
"learning_rate": 7.6155555555555564e-06,
|
458 |
+
"loss": 0.1243,
|
459 |
+
"step": 1575
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 0.32,
|
463 |
+
"grad_norm": 2.508364677429199,
|
464 |
+
"learning_rate": 7.5600000000000005e-06,
|
465 |
+
"loss": 0.1896,
|
466 |
+
"step": 1600
|
467 |
+
},
|
468 |
+
{
|
469 |
+
"epoch": 0.325,
|
470 |
+
"grad_norm": 2.0765221118927,
|
471 |
+
"learning_rate": 7.504444444444445e-06,
|
472 |
+
"loss": 0.1707,
|
473 |
+
"step": 1625
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 0.33,
|
477 |
+
"grad_norm": 2.030231475830078,
|
478 |
+
"learning_rate": 7.44888888888889e-06,
|
479 |
+
"loss": 0.1341,
|
480 |
+
"step": 1650
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 0.335,
|
484 |
+
"grad_norm": 2.9829025268554688,
|
485 |
+
"learning_rate": 7.393333333333333e-06,
|
486 |
+
"loss": 0.1305,
|
487 |
+
"step": 1675
|
488 |
+
},
|
489 |
+
{
|
490 |
+
"epoch": 0.34,
|
491 |
+
"grad_norm": 2.1654741764068604,
|
492 |
+
"learning_rate": 7.337777777777778e-06,
|
493 |
+
"loss": 0.122,
|
494 |
+
"step": 1700
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 0.345,
|
498 |
+
"grad_norm": 2.106933832168579,
|
499 |
+
"learning_rate": 7.282222222222222e-06,
|
500 |
+
"loss": 0.1427,
|
501 |
+
"step": 1725
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 0.35,
|
505 |
+
"grad_norm": 3.380559206008911,
|
506 |
+
"learning_rate": 7.226666666666667e-06,
|
507 |
+
"loss": 0.3227,
|
508 |
+
"step": 1750
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"epoch": 0.355,
|
512 |
+
"grad_norm": 2.084108829498291,
|
513 |
+
"learning_rate": 7.171111111111112e-06,
|
514 |
+
"loss": 0.194,
|
515 |
+
"step": 1775
|
516 |
+
},
|
517 |
+
{
|
518 |
+
"epoch": 0.36,
|
519 |
+
"grad_norm": 2.1483328342437744,
|
520 |
+
"learning_rate": 7.115555555555557e-06,
|
521 |
+
"loss": 0.1193,
|
522 |
+
"step": 1800
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"epoch": 0.365,
|
526 |
+
"grad_norm": 1.6093554496765137,
|
527 |
+
"learning_rate": 7.06e-06,
|
528 |
+
"loss": 0.0958,
|
529 |
+
"step": 1825
|
530 |
+
},
|
531 |
+
{
|
532 |
+
"epoch": 0.37,
|
533 |
+
"grad_norm": 1.4673306941986084,
|
534 |
+
"learning_rate": 7.004444444444445e-06,
|
535 |
+
"loss": 0.0896,
|
536 |
+
"step": 1850
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 0.375,
|
540 |
+
"grad_norm": 1.4364008903503418,
|
541 |
+
"learning_rate": 6.948888888888889e-06,
|
542 |
+
"loss": 0.0897,
|
543 |
+
"step": 1875
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 0.38,
|
547 |
+
"grad_norm": 1.4021852016448975,
|
548 |
+
"learning_rate": 6.893333333333334e-06,
|
549 |
+
"loss": 0.0772,
|
550 |
+
"step": 1900
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 0.385,
|
554 |
+
"grad_norm": 1.5660924911499023,
|
555 |
+
"learning_rate": 6.837777777777779e-06,
|
556 |
+
"loss": 0.0773,
|
557 |
+
"step": 1925
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 0.39,
|
561 |
+
"grad_norm": 1.9443961381912231,
|
562 |
+
"learning_rate": 6.782222222222222e-06,
|
563 |
+
"loss": 0.1027,
|
564 |
+
"step": 1950
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 0.395,
|
568 |
+
"grad_norm": 2.002857208251953,
|
569 |
+
"learning_rate": 6.726666666666667e-06,
|
570 |
+
"loss": 0.1945,
|
571 |
+
"step": 1975
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 0.4,
|
575 |
+
"grad_norm": 2.1631336212158203,
|
576 |
+
"learning_rate": 6.671111111111112e-06,
|
577 |
+
"loss": 0.1986,
|
578 |
+
"step": 2000
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 0.4,
|
582 |
+
"eval_loss": 0.20620137453079224,
|
583 |
+
"eval_runtime": 384.1444,
|
584 |
+
"eval_samples_per_second": 24.644,
|
585 |
+
"eval_steps_per_second": 3.082,
|
586 |
+
"eval_wer": 11.570062981553118,
|
587 |
+
"step": 2000
|
588 |
+
},
|
589 |
+
{
|
590 |
+
"epoch": 0.405,
|
591 |
+
"grad_norm": 1.9269295930862427,
|
592 |
+
"learning_rate": 6.615555555555556e-06,
|
593 |
+
"loss": 0.1477,
|
594 |
+
"step": 2025
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"epoch": 0.41,
|
598 |
+
"grad_norm": 1.8260035514831543,
|
599 |
+
"learning_rate": 6.560000000000001e-06,
|
600 |
+
"loss": 0.0935,
|
601 |
+
"step": 2050
|
602 |
+
},
|
603 |
+
{
|
604 |
+
"epoch": 0.415,
|
605 |
+
"grad_norm": 1.3721944093704224,
|
606 |
+
"learning_rate": 6.504444444444446e-06,
|
607 |
+
"loss": 0.0801,
|
608 |
+
"step": 2075
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 0.42,
|
612 |
+
"grad_norm": 1.8245339393615723,
|
613 |
+
"learning_rate": 6.448888888888889e-06,
|
614 |
+
"loss": 0.1047,
|
615 |
+
"step": 2100
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 0.425,
|
619 |
+
"grad_norm": 1.5538787841796875,
|
620 |
+
"learning_rate": 6.393333333333334e-06,
|
621 |
+
"loss": 0.0879,
|
622 |
+
"step": 2125
|
623 |
+
},
|
624 |
+
{
|
625 |
+
"epoch": 0.43,
|
626 |
+
"grad_norm": 1.519700527191162,
|
627 |
+
"learning_rate": 6.3377777777777786e-06,
|
628 |
+
"loss": 0.0813,
|
629 |
+
"step": 2150
|
630 |
+
},
|
631 |
+
{
|
632 |
+
"epoch": 0.435,
|
633 |
+
"grad_norm": 3.4600226879119873,
|
634 |
+
"learning_rate": 6.282222222222223e-06,
|
635 |
+
"loss": 0.1835,
|
636 |
+
"step": 2175
|
637 |
+
},
|
638 |
+
{
|
639 |
+
"epoch": 0.44,
|
640 |
+
"grad_norm": 3.3525190353393555,
|
641 |
+
"learning_rate": 6.2266666666666675e-06,
|
642 |
+
"loss": 0.6174,
|
643 |
+
"step": 2200
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"epoch": 0.445,
|
647 |
+
"grad_norm": 2.3610615730285645,
|
648 |
+
"learning_rate": 6.171111111111112e-06,
|
649 |
+
"loss": 0.6218,
|
650 |
+
"step": 2225
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 0.45,
|
654 |
+
"grad_norm": 1.5530322790145874,
|
655 |
+
"learning_rate": 6.1155555555555555e-06,
|
656 |
+
"loss": 0.2095,
|
657 |
+
"step": 2250
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 0.455,
|
661 |
+
"grad_norm": 1.4927952289581299,
|
662 |
+
"learning_rate": 6.0600000000000004e-06,
|
663 |
+
"loss": 0.0967,
|
664 |
+
"step": 2275
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 0.46,
|
668 |
+
"grad_norm": 1.6570786237716675,
|
669 |
+
"learning_rate": 6.004444444444445e-06,
|
670 |
+
"loss": 0.0895,
|
671 |
+
"step": 2300
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 0.465,
|
675 |
+
"grad_norm": 1.5751352310180664,
|
676 |
+
"learning_rate": 5.948888888888889e-06,
|
677 |
+
"loss": 0.0878,
|
678 |
+
"step": 2325
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 1.001,
|
682 |
+
"grad_norm": 2.1557564735412598,
|
683 |
+
"learning_rate": 5.893333333333334e-06,
|
684 |
+
"loss": 0.095,
|
685 |
+
"step": 2350
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 1.006,
|
689 |
+
"grad_norm": 2.489004611968994,
|
690 |
+
"learning_rate": 5.837777777777777e-06,
|
691 |
+
"loss": 0.1279,
|
692 |
+
"step": 2375
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 1.011,
|
696 |
+
"grad_norm": 3.284184217453003,
|
697 |
+
"learning_rate": 5.782222222222222e-06,
|
698 |
+
"loss": 0.1914,
|
699 |
+
"step": 2400
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 1.016,
|
703 |
+
"grad_norm": 3.20249080657959,
|
704 |
+
"learning_rate": 5.726666666666667e-06,
|
705 |
+
"loss": 0.526,
|
706 |
+
"step": 2425
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 1.021,
|
710 |
+
"grad_norm": 2.9400064945220947,
|
711 |
+
"learning_rate": 5.671111111111112e-06,
|
712 |
+
"loss": 0.2781,
|
713 |
+
"step": 2450
|
714 |
+
},
|
715 |
+
{
|
716 |
+
"epoch": 1.026,
|
717 |
+
"grad_norm": 1.9436590671539307,
|
718 |
+
"learning_rate": 5.615555555555556e-06,
|
719 |
+
"loss": 0.1756,
|
720 |
+
"step": 2475
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 1.031,
|
724 |
+
"grad_norm": 2.086899757385254,
|
725 |
+
"learning_rate": 5.560000000000001e-06,
|
726 |
+
"loss": 0.1469,
|
727 |
+
"step": 2500
|
728 |
+
},
|
729 |
+
{
|
730 |
+
"epoch": 1.036,
|
731 |
+
"grad_norm": 1.6339154243469238,
|
732 |
+
"learning_rate": 5.504444444444444e-06,
|
733 |
+
"loss": 0.1348,
|
734 |
+
"step": 2525
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 1.041,
|
738 |
+
"grad_norm": 2.1945927143096924,
|
739 |
+
"learning_rate": 5.448888888888889e-06,
|
740 |
+
"loss": 0.112,
|
741 |
+
"step": 2550
|
742 |
+
},
|
743 |
+
{
|
744 |
+
"epoch": 1.046,
|
745 |
+
"grad_norm": 1.639339804649353,
|
746 |
+
"learning_rate": 5.393333333333334e-06,
|
747 |
+
"loss": 0.1134,
|
748 |
+
"step": 2575
|
749 |
+
},
|
750 |
+
{
|
751 |
+
"epoch": 1.051,
|
752 |
+
"grad_norm": 1.7829625606536865,
|
753 |
+
"learning_rate": 5.337777777777779e-06,
|
754 |
+
"loss": 0.098,
|
755 |
+
"step": 2600
|
756 |
+
},
|
757 |
+
{
|
758 |
+
"epoch": 1.056,
|
759 |
+
"grad_norm": 1.5553560256958008,
|
760 |
+
"learning_rate": 5.282222222222223e-06,
|
761 |
+
"loss": 0.0706,
|
762 |
+
"step": 2625
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 1.061,
|
766 |
+
"grad_norm": 2.2982029914855957,
|
767 |
+
"learning_rate": 5.226666666666667e-06,
|
768 |
+
"loss": 0.2427,
|
769 |
+
"step": 2650
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"epoch": 1.066,
|
773 |
+
"grad_norm": 1.794986367225647,
|
774 |
+
"learning_rate": 5.171111111111111e-06,
|
775 |
+
"loss": 0.116,
|
776 |
+
"step": 2675
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 1.071,
|
780 |
+
"grad_norm": 1.2313064336776733,
|
781 |
+
"learning_rate": 5.115555555555556e-06,
|
782 |
+
"loss": 0.074,
|
783 |
+
"step": 2700
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 1.076,
|
787 |
+
"grad_norm": 1.4959907531738281,
|
788 |
+
"learning_rate": 5.060000000000001e-06,
|
789 |
+
"loss": 0.0758,
|
790 |
+
"step": 2725
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 1.081,
|
794 |
+
"grad_norm": 1.216587781906128,
|
795 |
+
"learning_rate": 5.004444444444445e-06,
|
796 |
+
"loss": 0.0787,
|
797 |
+
"step": 2750
|
798 |
+
},
|
799 |
+
{
|
800 |
+
"epoch": 1.086,
|
801 |
+
"grad_norm": 1.8736680746078491,
|
802 |
+
"learning_rate": 4.94888888888889e-06,
|
803 |
+
"loss": 0.1223,
|
804 |
+
"step": 2775
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 1.091,
|
808 |
+
"grad_norm": 1.2765681743621826,
|
809 |
+
"learning_rate": 4.893333333333334e-06,
|
810 |
+
"loss": 0.1025,
|
811 |
+
"step": 2800
|
812 |
+
},
|
813 |
+
{
|
814 |
+
"epoch": 1.096,
|
815 |
+
"grad_norm": 1.614762544631958,
|
816 |
+
"learning_rate": 4.837777777777778e-06,
|
817 |
+
"loss": 0.089,
|
818 |
+
"step": 2825
|
819 |
+
},
|
820 |
+
{
|
821 |
+
"epoch": 1.101,
|
822 |
+
"grad_norm": 1.6263494491577148,
|
823 |
+
"learning_rate": 4.7822222222222226e-06,
|
824 |
+
"loss": 0.1046,
|
825 |
+
"step": 2850
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 1.106,
|
829 |
+
"grad_norm": 1.6685870885849,
|
830 |
+
"learning_rate": 4.7266666666666674e-06,
|
831 |
+
"loss": 0.0845,
|
832 |
+
"step": 2875
|
833 |
+
},
|
834 |
+
{
|
835 |
+
"epoch": 1.111,
|
836 |
+
"grad_norm": 1.2902885675430298,
|
837 |
+
"learning_rate": 4.6711111111111115e-06,
|
838 |
+
"loss": 0.0806,
|
839 |
+
"step": 2900
|
840 |
+
},
|
841 |
+
{
|
842 |
+
"epoch": 1.116,
|
843 |
+
"grad_norm": 1.8994227647781372,
|
844 |
+
"learning_rate": 4.6155555555555555e-06,
|
845 |
+
"loss": 0.0767,
|
846 |
+
"step": 2925
|
847 |
+
},
|
848 |
+
{
|
849 |
+
"epoch": 1.121,
|
850 |
+
"grad_norm": 1.3723390102386475,
|
851 |
+
"learning_rate": 4.56e-06,
|
852 |
+
"loss": 0.0716,
|
853 |
+
"step": 2950
|
854 |
+
},
|
855 |
+
{
|
856 |
+
"epoch": 1.126,
|
857 |
+
"grad_norm": 1.6283633708953857,
|
858 |
+
"learning_rate": 4.504444444444444e-06,
|
859 |
+
"loss": 0.0731,
|
860 |
+
"step": 2975
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"epoch": 1.131,
|
864 |
+
"grad_norm": 1.3962900638580322,
|
865 |
+
"learning_rate": 4.448888888888889e-06,
|
866 |
+
"loss": 0.0942,
|
867 |
+
"step": 3000
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 1.131,
|
871 |
+
"eval_loss": 0.19787383079528809,
|
872 |
+
"eval_runtime": 384.7545,
|
873 |
+
"eval_samples_per_second": 24.605,
|
874 |
+
"eval_steps_per_second": 3.077,
|
875 |
+
"eval_wer": 11.015377729467055,
|
876 |
+
"step": 3000
|
877 |
+
},
|
878 |
+
{
|
879 |
+
"epoch": 1.1360000000000001,
|
880 |
+
"grad_norm": 1.321205973625183,
|
881 |
+
"learning_rate": 4.393333333333334e-06,
|
882 |
+
"loss": 0.0731,
|
883 |
+
"step": 3025
|
884 |
+
},
|
885 |
+
{
|
886 |
+
"epoch": 1.141,
|
887 |
+
"grad_norm": 1.2057939767837524,
|
888 |
+
"learning_rate": 4.337777777777778e-06,
|
889 |
+
"loss": 0.0646,
|
890 |
+
"step": 3050
|
891 |
+
},
|
892 |
+
{
|
893 |
+
"epoch": 1.146,
|
894 |
+
"grad_norm": 1.0545603036880493,
|
895 |
+
"learning_rate": 4.282222222222222e-06,
|
896 |
+
"loss": 0.0628,
|
897 |
+
"step": 3075
|
898 |
+
},
|
899 |
+
{
|
900 |
+
"epoch": 1.151,
|
901 |
+
"grad_norm": 1.363511562347412,
|
902 |
+
"learning_rate": 4.226666666666667e-06,
|
903 |
+
"loss": 0.0614,
|
904 |
+
"step": 3100
|
905 |
+
},
|
906 |
+
{
|
907 |
+
"epoch": 1.156,
|
908 |
+
"grad_norm": 1.1930261850357056,
|
909 |
+
"learning_rate": 4.171111111111111e-06,
|
910 |
+
"loss": 0.0634,
|
911 |
+
"step": 3125
|
912 |
+
},
|
913 |
+
{
|
914 |
+
"epoch": 1.161,
|
915 |
+
"grad_norm": 3.0981078147888184,
|
916 |
+
"learning_rate": 4.115555555555556e-06,
|
917 |
+
"loss": 0.2114,
|
918 |
+
"step": 3150
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 1.166,
|
922 |
+
"grad_norm": 1.6411216259002686,
|
923 |
+
"learning_rate": 4.060000000000001e-06,
|
924 |
+
"loss": 0.1761,
|
925 |
+
"step": 3175
|
926 |
+
},
|
927 |
+
{
|
928 |
+
"epoch": 1.171,
|
929 |
+
"grad_norm": 1.1803615093231201,
|
930 |
+
"learning_rate": 4.004444444444445e-06,
|
931 |
+
"loss": 0.0729,
|
932 |
+
"step": 3200
|
933 |
+
},
|
934 |
+
{
|
935 |
+
"epoch": 1.176,
|
936 |
+
"grad_norm": 3.052395820617676,
|
937 |
+
"learning_rate": 3.948888888888889e-06,
|
938 |
+
"loss": 0.3801,
|
939 |
+
"step": 3225
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 1.181,
|
943 |
+
"grad_norm": 2.7270925045013428,
|
944 |
+
"learning_rate": 3.893333333333333e-06,
|
945 |
+
"loss": 0.3596,
|
946 |
+
"step": 3250
|
947 |
+
},
|
948 |
+
{
|
949 |
+
"epoch": 1.186,
|
950 |
+
"grad_norm": 1.7751537561416626,
|
951 |
+
"learning_rate": 3.837777777777778e-06,
|
952 |
+
"loss": 0.2001,
|
953 |
+
"step": 3275
|
954 |
+
},
|
955 |
+
{
|
956 |
+
"epoch": 1.191,
|
957 |
+
"grad_norm": 1.7908082008361816,
|
958 |
+
"learning_rate": 3.782222222222223e-06,
|
959 |
+
"loss": 0.1933,
|
960 |
+
"step": 3300
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 1.196,
|
964 |
+
"grad_norm": 1.8746870756149292,
|
965 |
+
"learning_rate": 3.726666666666667e-06,
|
966 |
+
"loss": 0.2034,
|
967 |
+
"step": 3325
|
968 |
+
},
|
969 |
+
{
|
970 |
+
"epoch": 1.201,
|
971 |
+
"grad_norm": 1.8763881921768188,
|
972 |
+
"learning_rate": 3.6711111111111113e-06,
|
973 |
+
"loss": 0.1608,
|
974 |
+
"step": 3350
|
975 |
+
},
|
976 |
+
{
|
977 |
+
"epoch": 1.206,
|
978 |
+
"grad_norm": 1.5297502279281616,
|
979 |
+
"learning_rate": 3.615555555555556e-06,
|
980 |
+
"loss": 0.1099,
|
981 |
+
"step": 3375
|
982 |
+
},
|
983 |
+
{
|
984 |
+
"epoch": 1.211,
|
985 |
+
"grad_norm": 1.4193453788757324,
|
986 |
+
"learning_rate": 3.5600000000000002e-06,
|
987 |
+
"loss": 0.079,
|
988 |
+
"step": 3400
|
989 |
+
},
|
990 |
+
{
|
991 |
+
"epoch": 1.216,
|
992 |
+
"grad_norm": 1.5194439888000488,
|
993 |
+
"learning_rate": 3.5044444444444447e-06,
|
994 |
+
"loss": 0.0652,
|
995 |
+
"step": 3425
|
996 |
+
},
|
997 |
+
{
|
998 |
+
"epoch": 1.221,
|
999 |
+
"grad_norm": 1.3446123600006104,
|
1000 |
+
"learning_rate": 3.4488888888888896e-06,
|
1001 |
+
"loss": 0.0623,
|
1002 |
+
"step": 3450
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 1.226,
|
1006 |
+
"grad_norm": 1.293591856956482,
|
1007 |
+
"learning_rate": 3.3933333333333336e-06,
|
1008 |
+
"loss": 0.063,
|
1009 |
+
"step": 3475
|
1010 |
+
},
|
1011 |
+
{
|
1012 |
+
"epoch": 1.231,
|
1013 |
+
"grad_norm": 1.3113939762115479,
|
1014 |
+
"learning_rate": 3.337777777777778e-06,
|
1015 |
+
"loss": 0.0691,
|
1016 |
+
"step": 3500
|
1017 |
+
},
|
1018 |
+
{
|
1019 |
+
"epoch": 1.236,
|
1020 |
+
"grad_norm": 1.8869807720184326,
|
1021 |
+
"learning_rate": 3.282222222222223e-06,
|
1022 |
+
"loss": 0.0697,
|
1023 |
+
"step": 3525
|
1024 |
+
},
|
1025 |
+
{
|
1026 |
+
"epoch": 1.241,
|
1027 |
+
"grad_norm": 1.311045527458191,
|
1028 |
+
"learning_rate": 3.226666666666667e-06,
|
1029 |
+
"loss": 0.1023,
|
1030 |
+
"step": 3550
|
1031 |
+
},
|
1032 |
+
{
|
1033 |
+
"epoch": 1.246,
|
1034 |
+
"grad_norm": 1.4343574047088623,
|
1035 |
+
"learning_rate": 3.1711111111111114e-06,
|
1036 |
+
"loss": 0.0833,
|
1037 |
+
"step": 3575
|
1038 |
+
},
|
1039 |
+
{
|
1040 |
+
"epoch": 1.251,
|
1041 |
+
"grad_norm": 1.6197088956832886,
|
1042 |
+
"learning_rate": 3.1155555555555555e-06,
|
1043 |
+
"loss": 0.0812,
|
1044 |
+
"step": 3600
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 1.256,
|
1048 |
+
"grad_norm": 1.4370137453079224,
|
1049 |
+
"learning_rate": 3.0600000000000003e-06,
|
1050 |
+
"loss": 0.0666,
|
1051 |
+
"step": 3625
|
1052 |
+
},
|
1053 |
+
{
|
1054 |
+
"epoch": 1.2610000000000001,
|
1055 |
+
"grad_norm": 1.3375777006149292,
|
1056 |
+
"learning_rate": 3.004444444444445e-06,
|
1057 |
+
"loss": 0.058,
|
1058 |
+
"step": 3650
|
1059 |
+
},
|
1060 |
+
{
|
1061 |
+
"epoch": 1.266,
|
1062 |
+
"grad_norm": 1.6165860891342163,
|
1063 |
+
"learning_rate": 2.948888888888889e-06,
|
1064 |
+
"loss": 0.0736,
|
1065 |
+
"step": 3675
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 1.271,
|
1069 |
+
"grad_norm": 2.2871458530426025,
|
1070 |
+
"learning_rate": 2.8933333333333337e-06,
|
1071 |
+
"loss": 0.0943,
|
1072 |
+
"step": 3700
|
1073 |
+
},
|
1074 |
+
{
|
1075 |
+
"epoch": 1.276,
|
1076 |
+
"grad_norm": 1.4543273448944092,
|
1077 |
+
"learning_rate": 2.837777777777778e-06,
|
1078 |
+
"loss": 0.0904,
|
1079 |
+
"step": 3725
|
1080 |
+
},
|
1081 |
+
{
|
1082 |
+
"epoch": 1.2810000000000001,
|
1083 |
+
"grad_norm": 1.64260995388031,
|
1084 |
+
"learning_rate": 2.7822222222222222e-06,
|
1085 |
+
"loss": 0.0736,
|
1086 |
+
"step": 3750
|
1087 |
+
},
|
1088 |
+
{
|
1089 |
+
"epoch": 1.286,
|
1090 |
+
"grad_norm": 1.1954628229141235,
|
1091 |
+
"learning_rate": 2.726666666666667e-06,
|
1092 |
+
"loss": 0.0634,
|
1093 |
+
"step": 3775
|
1094 |
+
},
|
1095 |
+
{
|
1096 |
+
"epoch": 1.291,
|
1097 |
+
"grad_norm": 1.245037317276001,
|
1098 |
+
"learning_rate": 2.6711111111111116e-06,
|
1099 |
+
"loss": 0.0759,
|
1100 |
+
"step": 3800
|
1101 |
+
},
|
1102 |
+
{
|
1103 |
+
"epoch": 1.296,
|
1104 |
+
"grad_norm": 1.1573712825775146,
|
1105 |
+
"learning_rate": 2.6155555555555556e-06,
|
1106 |
+
"loss": 0.0695,
|
1107 |
+
"step": 3825
|
1108 |
+
},
|
1109 |
+
{
|
1110 |
+
"epoch": 1.301,
|
1111 |
+
"grad_norm": 1.0521385669708252,
|
1112 |
+
"learning_rate": 2.56e-06,
|
1113 |
+
"loss": 0.058,
|
1114 |
+
"step": 3850
|
1115 |
+
},
|
1116 |
+
{
|
1117 |
+
"epoch": 1.306,
|
1118 |
+
"grad_norm": 1.6519237756729126,
|
1119 |
+
"learning_rate": 2.504444444444445e-06,
|
1120 |
+
"loss": 0.0706,
|
1121 |
+
"step": 3875
|
1122 |
+
},
|
1123 |
+
{
|
1124 |
+
"epoch": 1.311,
|
1125 |
+
"grad_norm": 1.3316805362701416,
|
1126 |
+
"learning_rate": 2.448888888888889e-06,
|
1127 |
+
"loss": 0.0681,
|
1128 |
+
"step": 3900
|
1129 |
+
},
|
1130 |
+
{
|
1131 |
+
"epoch": 1.316,
|
1132 |
+
"grad_norm": 1.8995634317398071,
|
1133 |
+
"learning_rate": 2.3933333333333334e-06,
|
1134 |
+
"loss": 0.1458,
|
1135 |
+
"step": 3925
|
1136 |
+
},
|
1137 |
+
{
|
1138 |
+
"epoch": 1.321,
|
1139 |
+
"grad_norm": 1.9890100955963135,
|
1140 |
+
"learning_rate": 2.337777777777778e-06,
|
1141 |
+
"loss": 0.1334,
|
1142 |
+
"step": 3950
|
1143 |
+
},
|
1144 |
+
{
|
1145 |
+
"epoch": 1.326,
|
1146 |
+
"grad_norm": 1.2592449188232422,
|
1147 |
+
"learning_rate": 2.2822222222222223e-06,
|
1148 |
+
"loss": 0.1247,
|
1149 |
+
"step": 3975
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 1.331,
|
1153 |
+
"grad_norm": 1.3058019876480103,
|
1154 |
+
"learning_rate": 2.226666666666667e-06,
|
1155 |
+
"loss": 0.0577,
|
1156 |
+
"step": 4000
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 1.331,
|
1160 |
+
"eval_loss": 0.2000385820865631,
|
1161 |
+
"eval_runtime": 386.3157,
|
1162 |
+
"eval_samples_per_second": 24.506,
|
1163 |
+
"eval_steps_per_second": 3.065,
|
1164 |
+
"eval_wer": 10.634930784232232,
|
1165 |
+
"step": 4000
|
1166 |
+
},
|
1167 |
+
{
|
1168 |
+
"epoch": 1.336,
|
1169 |
+
"grad_norm": 1.0941780805587769,
|
1170 |
+
"learning_rate": 2.1711111111111113e-06,
|
1171 |
+
"loss": 0.0498,
|
1172 |
+
"step": 4025
|
1173 |
+
},
|
1174 |
+
{
|
1175 |
+
"epoch": 1.341,
|
1176 |
+
"grad_norm": 1.407287836074829,
|
1177 |
+
"learning_rate": 2.1155555555555557e-06,
|
1178 |
+
"loss": 0.0479,
|
1179 |
+
"step": 4050
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 1.346,
|
1183 |
+
"grad_norm": 0.8003798723220825,
|
1184 |
+
"learning_rate": 2.06e-06,
|
1185 |
+
"loss": 0.0518,
|
1186 |
+
"step": 4075
|
1187 |
+
},
|
1188 |
+
{
|
1189 |
+
"epoch": 1.351,
|
1190 |
+
"grad_norm": 1.1387590169906616,
|
1191 |
+
"learning_rate": 2.0044444444444446e-06,
|
1192 |
+
"loss": 0.0453,
|
1193 |
+
"step": 4100
|
1194 |
+
},
|
1195 |
+
{
|
1196 |
+
"epoch": 1.3559999999999999,
|
1197 |
+
"grad_norm": 2.531583786010742,
|
1198 |
+
"learning_rate": 1.948888888888889e-06,
|
1199 |
+
"loss": 0.1058,
|
1200 |
+
"step": 4125
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 1.361,
|
1204 |
+
"grad_norm": 2.965607166290283,
|
1205 |
+
"learning_rate": 1.8933333333333333e-06,
|
1206 |
+
"loss": 0.4739,
|
1207 |
+
"step": 4150
|
1208 |
+
},
|
1209 |
+
{
|
1210 |
+
"epoch": 1.366,
|
1211 |
+
"grad_norm": 3.0888776779174805,
|
1212 |
+
"learning_rate": 1.837777777777778e-06,
|
1213 |
+
"loss": 0.5685,
|
1214 |
+
"step": 4175
|
1215 |
+
},
|
1216 |
+
{
|
1217 |
+
"epoch": 1.371,
|
1218 |
+
"grad_norm": 1.201741099357605,
|
1219 |
+
"learning_rate": 1.7822222222222225e-06,
|
1220 |
+
"loss": 0.203,
|
1221 |
+
"step": 4200
|
1222 |
+
},
|
1223 |
+
{
|
1224 |
+
"epoch": 1.376,
|
1225 |
+
"grad_norm": 1.1447727680206299,
|
1226 |
+
"learning_rate": 1.7266666666666667e-06,
|
1227 |
+
"loss": 0.0638,
|
1228 |
+
"step": 4225
|
1229 |
+
},
|
1230 |
+
{
|
1231 |
+
"epoch": 1.381,
|
1232 |
+
"grad_norm": 0.9738747477531433,
|
1233 |
+
"learning_rate": 1.6711111111111112e-06,
|
1234 |
+
"loss": 0.0554,
|
1235 |
+
"step": 4250
|
1236 |
+
},
|
1237 |
+
{
|
1238 |
+
"epoch": 1.3860000000000001,
|
1239 |
+
"grad_norm": 0.9229967594146729,
|
1240 |
+
"learning_rate": 1.6155555555555559e-06,
|
1241 |
+
"loss": 0.0513,
|
1242 |
+
"step": 4275
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 1.391,
|
1246 |
+
"grad_norm": 1.1114579439163208,
|
1247 |
+
"learning_rate": 1.56e-06,
|
1248 |
+
"loss": 0.0515,
|
1249 |
+
"step": 4300
|
1250 |
+
},
|
1251 |
+
{
|
1252 |
+
"epoch": 1.396,
|
1253 |
+
"grad_norm": 2.101980447769165,
|
1254 |
+
"learning_rate": 1.5044444444444446e-06,
|
1255 |
+
"loss": 0.078,
|
1256 |
+
"step": 4325
|
1257 |
+
},
|
1258 |
+
{
|
1259 |
+
"epoch": 1.401,
|
1260 |
+
"grad_norm": 1.728270173072815,
|
1261 |
+
"learning_rate": 1.4488888888888892e-06,
|
1262 |
+
"loss": 0.1397,
|
1263 |
+
"step": 4350
|
1264 |
+
},
|
1265 |
+
{
|
1266 |
+
"epoch": 1.4060000000000001,
|
1267 |
+
"grad_norm": 1.6103638410568237,
|
1268 |
+
"learning_rate": 1.3933333333333335e-06,
|
1269 |
+
"loss": 0.0851,
|
1270 |
+
"step": 4375
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 1.411,
|
1274 |
+
"grad_norm": 1.334058403968811,
|
1275 |
+
"learning_rate": 1.337777777777778e-06,
|
1276 |
+
"loss": 0.0715,
|
1277 |
+
"step": 4400
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 1.416,
|
1281 |
+
"grad_norm": 1.138947606086731,
|
1282 |
+
"learning_rate": 1.2822222222222222e-06,
|
1283 |
+
"loss": 0.0652,
|
1284 |
+
"step": 4425
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 1.421,
|
1288 |
+
"grad_norm": 1.4695849418640137,
|
1289 |
+
"learning_rate": 1.2266666666666666e-06,
|
1290 |
+
"loss": 0.0723,
|
1291 |
+
"step": 4450
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 1.426,
|
1295 |
+
"grad_norm": 4.378330707550049,
|
1296 |
+
"learning_rate": 1.171111111111111e-06,
|
1297 |
+
"loss": 0.1358,
|
1298 |
+
"step": 4475
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 1.431,
|
1302 |
+
"grad_norm": 1.7877916097640991,
|
1303 |
+
"learning_rate": 1.1155555555555558e-06,
|
1304 |
+
"loss": 0.2452,
|
1305 |
+
"step": 4500
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 1.436,
|
1309 |
+
"grad_norm": 1.5879491567611694,
|
1310 |
+
"learning_rate": 1.06e-06,
|
1311 |
+
"loss": 0.1187,
|
1312 |
+
"step": 4525
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 1.441,
|
1316 |
+
"grad_norm": 1.056880235671997,
|
1317 |
+
"learning_rate": 1.0044444444444445e-06,
|
1318 |
+
"loss": 0.0584,
|
1319 |
+
"step": 4550
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 1.446,
|
1323 |
+
"grad_norm": 0.8786937594413757,
|
1324 |
+
"learning_rate": 9.488888888888889e-07,
|
1325 |
+
"loss": 0.0487,
|
1326 |
+
"step": 4575
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 1.451,
|
1330 |
+
"grad_norm": 0.9723278880119324,
|
1331 |
+
"learning_rate": 8.933333333333334e-07,
|
1332 |
+
"loss": 0.0413,
|
1333 |
+
"step": 4600
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 1.456,
|
1337 |
+
"grad_norm": 1.1325929164886475,
|
1338 |
+
"learning_rate": 8.37777777777778e-07,
|
1339 |
+
"loss": 0.0444,
|
1340 |
+
"step": 4625
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 1.461,
|
1344 |
+
"grad_norm": 1.0554845333099365,
|
1345 |
+
"learning_rate": 7.822222222222223e-07,
|
1346 |
+
"loss": 0.0408,
|
1347 |
+
"step": 4650
|
1348 |
+
},
|
1349 |
+
{
|
1350 |
+
"epoch": 1.466,
|
1351 |
+
"grad_norm": 1.6001170873641968,
|
1352 |
+
"learning_rate": 7.266666666666668e-07,
|
1353 |
+
"loss": 0.0486,
|
1354 |
+
"step": 4675
|
1355 |
+
},
|
1356 |
+
{
|
1357 |
+
"epoch": 2.002,
|
1358 |
+
"grad_norm": 1.434520959854126,
|
1359 |
+
"learning_rate": 6.711111111111111e-07,
|
1360 |
+
"loss": 0.0605,
|
1361 |
+
"step": 4700
|
1362 |
+
},
|
1363 |
+
{
|
1364 |
+
"epoch": 2.007,
|
1365 |
+
"grad_norm": 2.3628861904144287,
|
1366 |
+
"learning_rate": 6.155555555555556e-07,
|
1367 |
+
"loss": 0.0848,
|
1368 |
+
"step": 4725
|
1369 |
+
},
|
1370 |
+
{
|
1371 |
+
"epoch": 2.012,
|
1372 |
+
"grad_norm": 1.8358659744262695,
|
1373 |
+
"learning_rate": 5.6e-07,
|
1374 |
+
"loss": 0.1237,
|
1375 |
+
"step": 4750
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 2.017,
|
1379 |
+
"grad_norm": 1.7591222524642944,
|
1380 |
+
"learning_rate": 5.044444444444445e-07,
|
1381 |
+
"loss": 0.0846,
|
1382 |
+
"step": 4775
|
1383 |
+
},
|
1384 |
+
{
|
1385 |
+
"epoch": 2.022,
|
1386 |
+
"grad_norm": 1.3934890031814575,
|
1387 |
+
"learning_rate": 4.488888888888889e-07,
|
1388 |
+
"loss": 0.0855,
|
1389 |
+
"step": 4800
|
1390 |
+
},
|
1391 |
+
{
|
1392 |
+
"epoch": 2.027,
|
1393 |
+
"grad_norm": 1.0459132194519043,
|
1394 |
+
"learning_rate": 3.9333333333333336e-07,
|
1395 |
+
"loss": 0.0702,
|
1396 |
+
"step": 4825
|
1397 |
+
},
|
1398 |
+
{
|
1399 |
+
"epoch": 2.032,
|
1400 |
+
"grad_norm": 1.2503104209899902,
|
1401 |
+
"learning_rate": 3.3777777777777777e-07,
|
1402 |
+
"loss": 0.0637,
|
1403 |
+
"step": 4850
|
1404 |
+
},
|
1405 |
+
{
|
1406 |
+
"epoch": 2.037,
|
1407 |
+
"grad_norm": 1.1123186349868774,
|
1408 |
+
"learning_rate": 2.822222222222222e-07,
|
1409 |
+
"loss": 0.0555,
|
1410 |
+
"step": 4875
|
1411 |
+
},
|
1412 |
+
{
|
1413 |
+
"epoch": 2.042,
|
1414 |
+
"grad_norm": 1.0896210670471191,
|
1415 |
+
"learning_rate": 2.266666666666667e-07,
|
1416 |
+
"loss": 0.0538,
|
1417 |
+
"step": 4900
|
1418 |
+
},
|
1419 |
+
{
|
1420 |
+
"epoch": 2.047,
|
1421 |
+
"grad_norm": 1.138146162033081,
|
1422 |
+
"learning_rate": 1.7111111111111114e-07,
|
1423 |
+
"loss": 0.0486,
|
1424 |
+
"step": 4925
|
1425 |
+
},
|
1426 |
+
{
|
1427 |
+
"epoch": 2.052,
|
1428 |
+
"grad_norm": 1.1482701301574707,
|
1429 |
+
"learning_rate": 1.1555555555555556e-07,
|
1430 |
+
"loss": 0.0668,
|
1431 |
+
"step": 4950
|
1432 |
+
},
|
1433 |
+
{
|
1434 |
+
"epoch": 2.057,
|
1435 |
+
"grad_norm": 0.9316118955612183,
|
1436 |
+
"learning_rate": 6.000000000000001e-08,
|
1437 |
+
"loss": 0.0581,
|
1438 |
+
"step": 4975
|
1439 |
+
},
|
1440 |
+
{
|
1441 |
+
"epoch": 2.062,
|
1442 |
+
"grad_norm": 1.259885549545288,
|
1443 |
+
"learning_rate": 4.444444444444445e-09,
|
1444 |
+
"loss": 0.0516,
|
1445 |
+
"step": 5000
|
1446 |
+
},
|
1447 |
+
{
|
1448 |
+
"epoch": 2.062,
|
1449 |
+
"eval_loss": 0.20070870220661163,
|
1450 |
+
"eval_runtime": 383.0104,
|
1451 |
+
"eval_samples_per_second": 24.717,
|
1452 |
+
"eval_steps_per_second": 3.091,
|
1453 |
+
"eval_wer": 10.670098148917804,
|
1454 |
+
"step": 5000
|
1455 |
+
},
|
1456 |
+
{
|
1457 |
+
"epoch": 2.062,
|
1458 |
+
"step": 5000,
|
1459 |
+
"total_flos": 9.231442936676352e+19,
|
1460 |
+
"train_loss": 0.16698810555934906,
|
1461 |
+
"train_runtime": 16450.2071,
|
1462 |
+
"train_samples_per_second": 19.453,
|
1463 |
+
"train_steps_per_second": 0.304
|
1464 |
+
}
|
1465 |
+
],
|
1466 |
+
"logging_steps": 25,
|
1467 |
+
"max_steps": 5000,
|
1468 |
+
"num_input_tokens_seen": 0,
|
1469 |
+
"num_train_epochs": 9223372036854775807,
|
1470 |
+
"save_steps": 1000,
|
1471 |
+
"stateful_callbacks": {
|
1472 |
+
"TrainerControl": {
|
1473 |
+
"args": {
|
1474 |
+
"should_epoch_stop": false,
|
1475 |
+
"should_evaluate": false,
|
1476 |
+
"should_log": false,
|
1477 |
+
"should_save": true,
|
1478 |
+
"should_training_stop": true
|
1479 |
+
},
|
1480 |
+
"attributes": {}
|
1481 |
+
}
|
1482 |
+
},
|
1483 |
+
"total_flos": 9.231442936676352e+19,
|
1484 |
+
"train_batch_size": 64,
|
1485 |
+
"trial_name": null,
|
1486 |
+
"trial_params": null
|
1487 |
+
}
|