End of training
Browse files- README.md +12 -9
- all_results.json +12 -0
- eval_results.json +8 -0
- runs/Feb07_19-14-38_ip-172-31-68-44/events.out.tfevents.1707404326.ip-172-31-68-44.1925.1 +3 -0
- train_results.json +7 -0
- trainer_state.json +1320 -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_16_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_16_0
|
18 |
-
type: common_voice_16_0
|
19 |
config: ar
|
20 |
split: test
|
21 |
args: ar
|
22 |
metrics:
|
23 |
- name: Wer
|
24 |
type: wer
|
25 |
-
value:
|
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_16_0 dataset.
|
34 |
It achieves the following results on the evaluation set:
|
35 |
-
- Loss: 0.
|
36 |
-
- Wer:
|
37 |
|
38 |
## Model description
|
39 |
|
|
|
1 |
---
|
2 |
+
language:
|
3 |
+
- ar
|
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_16_0
|
11 |
metrics:
|
12 |
- wer
|
13 |
model-index:
|
14 |
+
- name: Whisper Small Arabic
|
15 |
results:
|
16 |
- task:
|
17 |
name: Automatic Speech Recognition
|
18 |
type: automatic-speech-recognition
|
19 |
dataset:
|
20 |
+
name: mozilla-foundation/common_voice_16_0 ar
|
21 |
+
type: mozilla-foundation/common_voice_16_0
|
22 |
config: ar
|
23 |
split: test
|
24 |
args: ar
|
25 |
metrics:
|
26 |
- name: Wer
|
27 |
type: wer
|
28 |
+
value: 58.90729282066525
|
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 Arabic
|
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_16_0 ar dataset.
|
37 |
It achieves the following results on the evaluation set:
|
38 |
+
- Loss: 0.4005
|
39 |
+
- Wer: 58.9073
|
40 |
|
41 |
## Model description
|
42 |
|
all_results.json
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 15.34,
|
3 |
+
"eval_loss": 0.40053287148475647,
|
4 |
+
"eval_runtime": 2688.0099,
|
5 |
+
"eval_samples_per_second": 3.894,
|
6 |
+
"eval_steps_per_second": 0.122,
|
7 |
+
"eval_wer": 58.90729282066525,
|
8 |
+
"train_loss": 0.23562390689849855,
|
9 |
+
"train_runtime": 67707.284,
|
10 |
+
"train_samples_per_second": 4.726,
|
11 |
+
"train_steps_per_second": 0.074
|
12 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 15.34,
|
3 |
+
"eval_loss": 0.40053287148475647,
|
4 |
+
"eval_runtime": 2688.0099,
|
5 |
+
"eval_samples_per_second": 3.894,
|
6 |
+
"eval_steps_per_second": 0.122,
|
7 |
+
"eval_wer": 58.90729282066525
|
8 |
+
}
|
runs/Feb07_19-14-38_ip-172-31-68-44/events.out.tfevents.1707404326.ip-172-31-68-44.1925.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b496f23590b2bc1d73f2282c058ff4f20889900d34c13e669fab88aa433ef27d
|
3 |
+
size 406
|
train_results.json
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 15.34,
|
3 |
+
"train_loss": 0.23562390689849855,
|
4 |
+
"train_runtime": 67707.284,
|
5 |
+
"train_samples_per_second": 4.726,
|
6 |
+
"train_steps_per_second": 0.074
|
7 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1320 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 58.90729282066525,
|
3 |
+
"best_model_checkpoint": "./checkpoint-3500",
|
4 |
+
"epoch": 15.337423312883436,
|
5 |
+
"eval_steps": 500,
|
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": 0.08,
|
13 |
+
"learning_rate": 4.6e-07,
|
14 |
+
"loss": 1.0735,
|
15 |
+
"step": 25
|
16 |
+
},
|
17 |
+
{
|
18 |
+
"epoch": 0.15,
|
19 |
+
"learning_rate": 9.6e-07,
|
20 |
+
"loss": 0.9014,
|
21 |
+
"step": 50
|
22 |
+
},
|
23 |
+
{
|
24 |
+
"epoch": 0.23,
|
25 |
+
"learning_rate": 9.953535353535352e-07,
|
26 |
+
"loss": 0.651,
|
27 |
+
"step": 75
|
28 |
+
},
|
29 |
+
{
|
30 |
+
"epoch": 0.31,
|
31 |
+
"learning_rate": 9.903030303030303e-07,
|
32 |
+
"loss": 0.4902,
|
33 |
+
"step": 100
|
34 |
+
},
|
35 |
+
{
|
36 |
+
"epoch": 0.38,
|
37 |
+
"learning_rate": 9.852525252525253e-07,
|
38 |
+
"loss": 0.4484,
|
39 |
+
"step": 125
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.46,
|
43 |
+
"learning_rate": 9.802020202020202e-07,
|
44 |
+
"loss": 0.4164,
|
45 |
+
"step": 150
|
46 |
+
},
|
47 |
+
{
|
48 |
+
"epoch": 0.54,
|
49 |
+
"learning_rate": 9.751515151515151e-07,
|
50 |
+
"loss": 0.4303,
|
51 |
+
"step": 175
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.61,
|
55 |
+
"learning_rate": 9.7010101010101e-07,
|
56 |
+
"loss": 0.4073,
|
57 |
+
"step": 200
|
58 |
+
},
|
59 |
+
{
|
60 |
+
"epoch": 0.69,
|
61 |
+
"learning_rate": 9.65050505050505e-07,
|
62 |
+
"loss": 0.3977,
|
63 |
+
"step": 225
|
64 |
+
},
|
65 |
+
{
|
66 |
+
"epoch": 0.77,
|
67 |
+
"learning_rate": 9.6e-07,
|
68 |
+
"loss": 0.3906,
|
69 |
+
"step": 250
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.84,
|
73 |
+
"learning_rate": 9.549494949494948e-07,
|
74 |
+
"loss": 0.3687,
|
75 |
+
"step": 275
|
76 |
+
},
|
77 |
+
{
|
78 |
+
"epoch": 0.92,
|
79 |
+
"learning_rate": 9.498989898989899e-07,
|
80 |
+
"loss": 0.3817,
|
81 |
+
"step": 300
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 1.0,
|
85 |
+
"learning_rate": 9.448484848484848e-07,
|
86 |
+
"loss": 0.3733,
|
87 |
+
"step": 325
|
88 |
+
},
|
89 |
+
{
|
90 |
+
"epoch": 1.07,
|
91 |
+
"learning_rate": 9.397979797979797e-07,
|
92 |
+
"loss": 0.3608,
|
93 |
+
"step": 350
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 1.15,
|
97 |
+
"learning_rate": 9.347474747474747e-07,
|
98 |
+
"loss": 0.3519,
|
99 |
+
"step": 375
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 1.23,
|
103 |
+
"learning_rate": 9.296969696969696e-07,
|
104 |
+
"loss": 0.3286,
|
105 |
+
"step": 400
|
106 |
+
},
|
107 |
+
{
|
108 |
+
"epoch": 1.3,
|
109 |
+
"learning_rate": 9.246464646464645e-07,
|
110 |
+
"loss": 0.3468,
|
111 |
+
"step": 425
|
112 |
+
},
|
113 |
+
{
|
114 |
+
"epoch": 1.38,
|
115 |
+
"learning_rate": 9.195959595959596e-07,
|
116 |
+
"loss": 0.3386,
|
117 |
+
"step": 450
|
118 |
+
},
|
119 |
+
{
|
120 |
+
"epoch": 1.46,
|
121 |
+
"learning_rate": 9.145454545454545e-07,
|
122 |
+
"loss": 0.3356,
|
123 |
+
"step": 475
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 1.53,
|
127 |
+
"learning_rate": 9.094949494949494e-07,
|
128 |
+
"loss": 0.3404,
|
129 |
+
"step": 500
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.53,
|
133 |
+
"eval_loss": 0.46061962842941284,
|
134 |
+
"eval_runtime": 2718.8588,
|
135 |
+
"eval_samples_per_second": 3.849,
|
136 |
+
"eval_steps_per_second": 0.121,
|
137 |
+
"eval_wer": 66.62163257585992,
|
138 |
+
"step": 500
|
139 |
+
},
|
140 |
+
{
|
141 |
+
"epoch": 1.61,
|
142 |
+
"learning_rate": 9.044444444444445e-07,
|
143 |
+
"loss": 0.3266,
|
144 |
+
"step": 525
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 1.69,
|
148 |
+
"learning_rate": 8.993939393939394e-07,
|
149 |
+
"loss": 0.3369,
|
150 |
+
"step": 550
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 1.76,
|
154 |
+
"learning_rate": 8.943434343434342e-07,
|
155 |
+
"loss": 0.3154,
|
156 |
+
"step": 575
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 1.84,
|
160 |
+
"learning_rate": 8.892929292929292e-07,
|
161 |
+
"loss": 0.3285,
|
162 |
+
"step": 600
|
163 |
+
},
|
164 |
+
{
|
165 |
+
"epoch": 1.92,
|
166 |
+
"learning_rate": 8.842424242424242e-07,
|
167 |
+
"loss": 0.3154,
|
168 |
+
"step": 625
|
169 |
+
},
|
170 |
+
{
|
171 |
+
"epoch": 1.99,
|
172 |
+
"learning_rate": 8.791919191919191e-07,
|
173 |
+
"loss": 0.3243,
|
174 |
+
"step": 650
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 2.07,
|
178 |
+
"learning_rate": 8.741414141414141e-07,
|
179 |
+
"loss": 0.3045,
|
180 |
+
"step": 675
|
181 |
+
},
|
182 |
+
{
|
183 |
+
"epoch": 2.15,
|
184 |
+
"learning_rate": 8.690909090909091e-07,
|
185 |
+
"loss": 0.2976,
|
186 |
+
"step": 700
|
187 |
+
},
|
188 |
+
{
|
189 |
+
"epoch": 2.22,
|
190 |
+
"learning_rate": 8.64040404040404e-07,
|
191 |
+
"loss": 0.2947,
|
192 |
+
"step": 725
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 2.3,
|
196 |
+
"learning_rate": 8.58989898989899e-07,
|
197 |
+
"loss": 0.2951,
|
198 |
+
"step": 750
|
199 |
+
},
|
200 |
+
{
|
201 |
+
"epoch": 2.38,
|
202 |
+
"learning_rate": 8.539393939393939e-07,
|
203 |
+
"loss": 0.2865,
|
204 |
+
"step": 775
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 2.45,
|
208 |
+
"learning_rate": 8.488888888888888e-07,
|
209 |
+
"loss": 0.3027,
|
210 |
+
"step": 800
|
211 |
+
},
|
212 |
+
{
|
213 |
+
"epoch": 2.53,
|
214 |
+
"learning_rate": 8.438383838383838e-07,
|
215 |
+
"loss": 0.2963,
|
216 |
+
"step": 825
|
217 |
+
},
|
218 |
+
{
|
219 |
+
"epoch": 2.61,
|
220 |
+
"learning_rate": 8.387878787878787e-07,
|
221 |
+
"loss": 0.2903,
|
222 |
+
"step": 850
|
223 |
+
},
|
224 |
+
{
|
225 |
+
"epoch": 2.68,
|
226 |
+
"learning_rate": 8.337373737373737e-07,
|
227 |
+
"loss": 0.2887,
|
228 |
+
"step": 875
|
229 |
+
},
|
230 |
+
{
|
231 |
+
"epoch": 2.76,
|
232 |
+
"learning_rate": 8.286868686868687e-07,
|
233 |
+
"loss": 0.2903,
|
234 |
+
"step": 900
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 2.84,
|
238 |
+
"learning_rate": 8.236363636363636e-07,
|
239 |
+
"loss": 0.2836,
|
240 |
+
"step": 925
|
241 |
+
},
|
242 |
+
{
|
243 |
+
"epoch": 2.91,
|
244 |
+
"learning_rate": 8.185858585858586e-07,
|
245 |
+
"loss": 0.2832,
|
246 |
+
"step": 950
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 2.99,
|
250 |
+
"learning_rate": 8.135353535353536e-07,
|
251 |
+
"loss": 0.2984,
|
252 |
+
"step": 975
|
253 |
+
},
|
254 |
+
{
|
255 |
+
"epoch": 3.07,
|
256 |
+
"learning_rate": 8.084848484848484e-07,
|
257 |
+
"loss": 0.2707,
|
258 |
+
"step": 1000
|
259 |
+
},
|
260 |
+
{
|
261 |
+
"epoch": 3.07,
|
262 |
+
"eval_loss": 0.42951369285583496,
|
263 |
+
"eval_runtime": 2739.5863,
|
264 |
+
"eval_samples_per_second": 3.82,
|
265 |
+
"eval_steps_per_second": 0.12,
|
266 |
+
"eval_wer": 66.84995541625011,
|
267 |
+
"step": 1000
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 3.14,
|
271 |
+
"learning_rate": 8.034343434343433e-07,
|
272 |
+
"loss": 0.2778,
|
273 |
+
"step": 1025
|
274 |
+
},
|
275 |
+
{
|
276 |
+
"epoch": 3.22,
|
277 |
+
"learning_rate": 7.983838383838384e-07,
|
278 |
+
"loss": 0.2742,
|
279 |
+
"step": 1050
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 3.3,
|
283 |
+
"learning_rate": 7.933333333333333e-07,
|
284 |
+
"loss": 0.2626,
|
285 |
+
"step": 1075
|
286 |
+
},
|
287 |
+
{
|
288 |
+
"epoch": 3.37,
|
289 |
+
"learning_rate": 7.882828282828282e-07,
|
290 |
+
"loss": 0.2658,
|
291 |
+
"step": 1100
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 3.45,
|
295 |
+
"learning_rate": 7.832323232323233e-07,
|
296 |
+
"loss": 0.2654,
|
297 |
+
"step": 1125
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 3.53,
|
301 |
+
"learning_rate": 7.781818181818182e-07,
|
302 |
+
"loss": 0.2634,
|
303 |
+
"step": 1150
|
304 |
+
},
|
305 |
+
{
|
306 |
+
"epoch": 3.6,
|
307 |
+
"learning_rate": 7.731313131313131e-07,
|
308 |
+
"loss": 0.2609,
|
309 |
+
"step": 1175
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 3.68,
|
313 |
+
"learning_rate": 7.68080808080808e-07,
|
314 |
+
"loss": 0.262,
|
315 |
+
"step": 1200
|
316 |
+
},
|
317 |
+
{
|
318 |
+
"epoch": 3.76,
|
319 |
+
"learning_rate": 7.63030303030303e-07,
|
320 |
+
"loss": 0.2701,
|
321 |
+
"step": 1225
|
322 |
+
},
|
323 |
+
{
|
324 |
+
"epoch": 3.83,
|
325 |
+
"learning_rate": 7.579797979797979e-07,
|
326 |
+
"loss": 0.2646,
|
327 |
+
"step": 1250
|
328 |
+
},
|
329 |
+
{
|
330 |
+
"epoch": 3.91,
|
331 |
+
"learning_rate": 7.529292929292928e-07,
|
332 |
+
"loss": 0.2692,
|
333 |
+
"step": 1275
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 3.99,
|
337 |
+
"learning_rate": 7.478787878787879e-07,
|
338 |
+
"loss": 0.2588,
|
339 |
+
"step": 1300
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 4.06,
|
343 |
+
"learning_rate": 7.428282828282828e-07,
|
344 |
+
"loss": 0.2549,
|
345 |
+
"step": 1325
|
346 |
+
},
|
347 |
+
{
|
348 |
+
"epoch": 4.14,
|
349 |
+
"learning_rate": 7.377777777777777e-07,
|
350 |
+
"loss": 0.2376,
|
351 |
+
"step": 1350
|
352 |
+
},
|
353 |
+
{
|
354 |
+
"epoch": 4.22,
|
355 |
+
"learning_rate": 7.327272727272728e-07,
|
356 |
+
"loss": 0.2542,
|
357 |
+
"step": 1375
|
358 |
+
},
|
359 |
+
{
|
360 |
+
"epoch": 4.29,
|
361 |
+
"learning_rate": 7.276767676767677e-07,
|
362 |
+
"loss": 0.2634,
|
363 |
+
"step": 1400
|
364 |
+
},
|
365 |
+
{
|
366 |
+
"epoch": 4.37,
|
367 |
+
"learning_rate": 7.226262626262625e-07,
|
368 |
+
"loss": 0.2453,
|
369 |
+
"step": 1425
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 4.45,
|
373 |
+
"learning_rate": 7.175757575757575e-07,
|
374 |
+
"loss": 0.2393,
|
375 |
+
"step": 1450
|
376 |
+
},
|
377 |
+
{
|
378 |
+
"epoch": 4.52,
|
379 |
+
"learning_rate": 7.125252525252525e-07,
|
380 |
+
"loss": 0.2466,
|
381 |
+
"step": 1475
|
382 |
+
},
|
383 |
+
{
|
384 |
+
"epoch": 4.6,
|
385 |
+
"learning_rate": 7.074747474747474e-07,
|
386 |
+
"loss": 0.2427,
|
387 |
+
"step": 1500
|
388 |
+
},
|
389 |
+
{
|
390 |
+
"epoch": 4.6,
|
391 |
+
"eval_loss": 0.4123845398426056,
|
392 |
+
"eval_runtime": 2708.1735,
|
393 |
+
"eval_samples_per_second": 3.865,
|
394 |
+
"eval_steps_per_second": 0.121,
|
395 |
+
"eval_wer": 61.166202815531356,
|
396 |
+
"step": 1500
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 4.68,
|
400 |
+
"learning_rate": 7.024242424242424e-07,
|
401 |
+
"loss": 0.247,
|
402 |
+
"step": 1525
|
403 |
+
},
|
404 |
+
{
|
405 |
+
"epoch": 4.75,
|
406 |
+
"learning_rate": 6.973737373737374e-07,
|
407 |
+
"loss": 0.2531,
|
408 |
+
"step": 1550
|
409 |
+
},
|
410 |
+
{
|
411 |
+
"epoch": 4.83,
|
412 |
+
"learning_rate": 6.923232323232323e-07,
|
413 |
+
"loss": 0.2496,
|
414 |
+
"step": 1575
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 4.91,
|
418 |
+
"learning_rate": 6.872727272727273e-07,
|
419 |
+
"loss": 0.2366,
|
420 |
+
"step": 1600
|
421 |
+
},
|
422 |
+
{
|
423 |
+
"epoch": 4.98,
|
424 |
+
"learning_rate": 6.822222222222221e-07,
|
425 |
+
"loss": 0.2287,
|
426 |
+
"step": 1625
|
427 |
+
},
|
428 |
+
{
|
429 |
+
"epoch": 5.06,
|
430 |
+
"learning_rate": 6.771717171717171e-07,
|
431 |
+
"loss": 0.2352,
|
432 |
+
"step": 1650
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"epoch": 5.14,
|
436 |
+
"learning_rate": 6.721212121212121e-07,
|
437 |
+
"loss": 0.2242,
|
438 |
+
"step": 1675
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"epoch": 5.21,
|
442 |
+
"learning_rate": 6.67070707070707e-07,
|
443 |
+
"loss": 0.2288,
|
444 |
+
"step": 1700
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 5.29,
|
448 |
+
"learning_rate": 6.62020202020202e-07,
|
449 |
+
"loss": 0.2374,
|
450 |
+
"step": 1725
|
451 |
+
},
|
452 |
+
{
|
453 |
+
"epoch": 5.37,
|
454 |
+
"learning_rate": 6.56969696969697e-07,
|
455 |
+
"loss": 0.2312,
|
456 |
+
"step": 1750
|
457 |
+
},
|
458 |
+
{
|
459 |
+
"epoch": 5.44,
|
460 |
+
"learning_rate": 6.519191919191919e-07,
|
461 |
+
"loss": 0.2279,
|
462 |
+
"step": 1775
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 5.52,
|
466 |
+
"learning_rate": 6.468686868686868e-07,
|
467 |
+
"loss": 0.2263,
|
468 |
+
"step": 1800
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"epoch": 5.6,
|
472 |
+
"learning_rate": 6.418181818181818e-07,
|
473 |
+
"loss": 0.2305,
|
474 |
+
"step": 1825
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 5.67,
|
478 |
+
"learning_rate": 6.367676767676767e-07,
|
479 |
+
"loss": 0.2251,
|
480 |
+
"step": 1850
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 5.75,
|
484 |
+
"learning_rate": 6.317171717171716e-07,
|
485 |
+
"loss": 0.2285,
|
486 |
+
"step": 1875
|
487 |
+
},
|
488 |
+
{
|
489 |
+
"epoch": 5.83,
|
490 |
+
"learning_rate": 6.266666666666667e-07,
|
491 |
+
"loss": 0.2308,
|
492 |
+
"step": 1900
|
493 |
+
},
|
494 |
+
{
|
495 |
+
"epoch": 5.9,
|
496 |
+
"learning_rate": 6.216161616161616e-07,
|
497 |
+
"loss": 0.2343,
|
498 |
+
"step": 1925
|
499 |
+
},
|
500 |
+
{
|
501 |
+
"epoch": 5.98,
|
502 |
+
"learning_rate": 6.165656565656565e-07,
|
503 |
+
"loss": 0.2335,
|
504 |
+
"step": 1950
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 6.06,
|
508 |
+
"learning_rate": 6.115151515151516e-07,
|
509 |
+
"loss": 0.2246,
|
510 |
+
"step": 1975
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"epoch": 6.13,
|
514 |
+
"learning_rate": 6.064646464646465e-07,
|
515 |
+
"loss": 0.2131,
|
516 |
+
"step": 2000
|
517 |
+
},
|
518 |
+
{
|
519 |
+
"epoch": 6.13,
|
520 |
+
"eval_loss": 0.4056357443332672,
|
521 |
+
"eval_runtime": 2709.0159,
|
522 |
+
"eval_samples_per_second": 3.863,
|
523 |
+
"eval_steps_per_second": 0.121,
|
524 |
+
"eval_wer": 62.303763949309634,
|
525 |
+
"step": 2000
|
526 |
+
},
|
527 |
+
{
|
528 |
+
"epoch": 6.21,
|
529 |
+
"learning_rate": 6.014141414141413e-07,
|
530 |
+
"loss": 0.2113,
|
531 |
+
"step": 2025
|
532 |
+
},
|
533 |
+
{
|
534 |
+
"epoch": 6.29,
|
535 |
+
"learning_rate": 5.963636363636363e-07,
|
536 |
+
"loss": 0.2207,
|
537 |
+
"step": 2050
|
538 |
+
},
|
539 |
+
{
|
540 |
+
"epoch": 6.37,
|
541 |
+
"learning_rate": 5.913131313131313e-07,
|
542 |
+
"loss": 0.2171,
|
543 |
+
"step": 2075
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 6.44,
|
547 |
+
"learning_rate": 5.862626262626262e-07,
|
548 |
+
"loss": 0.2112,
|
549 |
+
"step": 2100
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 6.52,
|
553 |
+
"learning_rate": 5.812121212121212e-07,
|
554 |
+
"loss": 0.2098,
|
555 |
+
"step": 2125
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 6.6,
|
559 |
+
"learning_rate": 5.761616161616162e-07,
|
560 |
+
"loss": 0.2094,
|
561 |
+
"step": 2150
|
562 |
+
},
|
563 |
+
{
|
564 |
+
"epoch": 6.67,
|
565 |
+
"learning_rate": 5.711111111111111e-07,
|
566 |
+
"loss": 0.2149,
|
567 |
+
"step": 2175
|
568 |
+
},
|
569 |
+
{
|
570 |
+
"epoch": 6.75,
|
571 |
+
"learning_rate": 5.660606060606061e-07,
|
572 |
+
"loss": 0.2162,
|
573 |
+
"step": 2200
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 6.83,
|
577 |
+
"learning_rate": 5.610101010101009e-07,
|
578 |
+
"loss": 0.2216,
|
579 |
+
"step": 2225
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 6.9,
|
583 |
+
"learning_rate": 5.559595959595959e-07,
|
584 |
+
"loss": 0.2169,
|
585 |
+
"step": 2250
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 6.98,
|
589 |
+
"learning_rate": 5.509090909090909e-07,
|
590 |
+
"loss": 0.2209,
|
591 |
+
"step": 2275
|
592 |
+
},
|
593 |
+
{
|
594 |
+
"epoch": 7.06,
|
595 |
+
"learning_rate": 5.458585858585858e-07,
|
596 |
+
"loss": 0.2108,
|
597 |
+
"step": 2300
|
598 |
+
},
|
599 |
+
{
|
600 |
+
"epoch": 7.13,
|
601 |
+
"learning_rate": 5.408080808080808e-07,
|
602 |
+
"loss": 0.198,
|
603 |
+
"step": 2325
|
604 |
+
},
|
605 |
+
{
|
606 |
+
"epoch": 7.21,
|
607 |
+
"learning_rate": 5.357575757575758e-07,
|
608 |
+
"loss": 0.2075,
|
609 |
+
"step": 2350
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 7.29,
|
613 |
+
"learning_rate": 5.307070707070707e-07,
|
614 |
+
"loss": 0.2067,
|
615 |
+
"step": 2375
|
616 |
+
},
|
617 |
+
{
|
618 |
+
"epoch": 7.36,
|
619 |
+
"learning_rate": 5.256565656565657e-07,
|
620 |
+
"loss": 0.2193,
|
621 |
+
"step": 2400
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 7.44,
|
625 |
+
"learning_rate": 5.206060606060607e-07,
|
626 |
+
"loss": 0.1996,
|
627 |
+
"step": 2425
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 7.52,
|
631 |
+
"learning_rate": 5.155555555555555e-07,
|
632 |
+
"loss": 0.2118,
|
633 |
+
"step": 2450
|
634 |
+
},
|
635 |
+
{
|
636 |
+
"epoch": 7.59,
|
637 |
+
"learning_rate": 5.105050505050504e-07,
|
638 |
+
"loss": 0.1938,
|
639 |
+
"step": 2475
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 7.67,
|
643 |
+
"learning_rate": 5.054545454545454e-07,
|
644 |
+
"loss": 0.2085,
|
645 |
+
"step": 2500
|
646 |
+
},
|
647 |
+
{
|
648 |
+
"epoch": 7.67,
|
649 |
+
"eval_loss": 0.401197612285614,
|
650 |
+
"eval_runtime": 2721.3063,
|
651 |
+
"eval_samples_per_second": 3.846,
|
652 |
+
"eval_steps_per_second": 0.121,
|
653 |
+
"eval_wer": 62.27539247210137,
|
654 |
+
"step": 2500
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 7.75,
|
658 |
+
"learning_rate": 5.004040404040404e-07,
|
659 |
+
"loss": 0.1983,
|
660 |
+
"step": 2525
|
661 |
+
},
|
662 |
+
{
|
663 |
+
"epoch": 7.82,
|
664 |
+
"learning_rate": 4.953535353535353e-07,
|
665 |
+
"loss": 0.2098,
|
666 |
+
"step": 2550
|
667 |
+
},
|
668 |
+
{
|
669 |
+
"epoch": 7.9,
|
670 |
+
"learning_rate": 4.903030303030302e-07,
|
671 |
+
"loss": 0.2036,
|
672 |
+
"step": 2575
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"epoch": 7.98,
|
676 |
+
"learning_rate": 4.852525252525253e-07,
|
677 |
+
"loss": 0.2024,
|
678 |
+
"step": 2600
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 8.05,
|
682 |
+
"learning_rate": 4.802020202020202e-07,
|
683 |
+
"loss": 0.2074,
|
684 |
+
"step": 2625
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 8.13,
|
688 |
+
"learning_rate": 4.7515151515151514e-07,
|
689 |
+
"loss": 0.1989,
|
690 |
+
"step": 2650
|
691 |
+
},
|
692 |
+
{
|
693 |
+
"epoch": 8.21,
|
694 |
+
"learning_rate": 4.701010101010101e-07,
|
695 |
+
"loss": 0.1904,
|
696 |
+
"step": 2675
|
697 |
+
},
|
698 |
+
{
|
699 |
+
"epoch": 8.28,
|
700 |
+
"learning_rate": 4.65050505050505e-07,
|
701 |
+
"loss": 0.1921,
|
702 |
+
"step": 2700
|
703 |
+
},
|
704 |
+
{
|
705 |
+
"epoch": 8.36,
|
706 |
+
"learning_rate": 4.6e-07,
|
707 |
+
"loss": 0.1993,
|
708 |
+
"step": 2725
|
709 |
+
},
|
710 |
+
{
|
711 |
+
"epoch": 8.44,
|
712 |
+
"learning_rate": 4.549494949494949e-07,
|
713 |
+
"loss": 0.1966,
|
714 |
+
"step": 2750
|
715 |
+
},
|
716 |
+
{
|
717 |
+
"epoch": 8.51,
|
718 |
+
"learning_rate": 4.498989898989899e-07,
|
719 |
+
"loss": 0.1912,
|
720 |
+
"step": 2775
|
721 |
+
},
|
722 |
+
{
|
723 |
+
"epoch": 8.59,
|
724 |
+
"learning_rate": 4.448484848484848e-07,
|
725 |
+
"loss": 0.2001,
|
726 |
+
"step": 2800
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"epoch": 8.67,
|
730 |
+
"learning_rate": 4.3979797979797977e-07,
|
731 |
+
"loss": 0.2025,
|
732 |
+
"step": 2825
|
733 |
+
},
|
734 |
+
{
|
735 |
+
"epoch": 8.74,
|
736 |
+
"learning_rate": 4.3474747474747475e-07,
|
737 |
+
"loss": 0.192,
|
738 |
+
"step": 2850
|
739 |
+
},
|
740 |
+
{
|
741 |
+
"epoch": 8.82,
|
742 |
+
"learning_rate": 4.296969696969697e-07,
|
743 |
+
"loss": 0.1979,
|
744 |
+
"step": 2875
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 8.9,
|
748 |
+
"learning_rate": 4.246464646464646e-07,
|
749 |
+
"loss": 0.1928,
|
750 |
+
"step": 2900
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"epoch": 8.97,
|
754 |
+
"learning_rate": 4.1959595959595955e-07,
|
755 |
+
"loss": 0.1845,
|
756 |
+
"step": 2925
|
757 |
+
},
|
758 |
+
{
|
759 |
+
"epoch": 9.05,
|
760 |
+
"learning_rate": 4.1454545454545453e-07,
|
761 |
+
"loss": 0.1846,
|
762 |
+
"step": 2950
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 9.13,
|
766 |
+
"learning_rate": 4.0949494949494946e-07,
|
767 |
+
"loss": 0.1926,
|
768 |
+
"step": 2975
|
769 |
+
},
|
770 |
+
{
|
771 |
+
"epoch": 9.2,
|
772 |
+
"learning_rate": 4.044444444444444e-07,
|
773 |
+
"loss": 0.1904,
|
774 |
+
"step": 3000
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 9.2,
|
778 |
+
"eval_loss": 0.3976161777973175,
|
779 |
+
"eval_runtime": 2684.5933,
|
780 |
+
"eval_samples_per_second": 3.899,
|
781 |
+
"eval_steps_per_second": 0.122,
|
782 |
+
"eval_wer": 59.734118727877004,
|
783 |
+
"step": 3000
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 9.28,
|
787 |
+
"learning_rate": 3.993939393939394e-07,
|
788 |
+
"loss": 0.1972,
|
789 |
+
"step": 3025
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 9.36,
|
793 |
+
"learning_rate": 3.943434343434343e-07,
|
794 |
+
"loss": 0.1881,
|
795 |
+
"step": 3050
|
796 |
+
},
|
797 |
+
{
|
798 |
+
"epoch": 9.43,
|
799 |
+
"learning_rate": 3.892929292929293e-07,
|
800 |
+
"loss": 0.1935,
|
801 |
+
"step": 3075
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 9.51,
|
805 |
+
"learning_rate": 3.8424242424242423e-07,
|
806 |
+
"loss": 0.1911,
|
807 |
+
"step": 3100
|
808 |
+
},
|
809 |
+
{
|
810 |
+
"epoch": 9.59,
|
811 |
+
"learning_rate": 3.7919191919191916e-07,
|
812 |
+
"loss": 0.1852,
|
813 |
+
"step": 3125
|
814 |
+
},
|
815 |
+
{
|
816 |
+
"epoch": 9.66,
|
817 |
+
"learning_rate": 3.7414141414141414e-07,
|
818 |
+
"loss": 0.1939,
|
819 |
+
"step": 3150
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 9.74,
|
823 |
+
"learning_rate": 3.690909090909091e-07,
|
824 |
+
"loss": 0.1834,
|
825 |
+
"step": 3175
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 9.82,
|
829 |
+
"learning_rate": 3.6404040404040406e-07,
|
830 |
+
"loss": 0.1794,
|
831 |
+
"step": 3200
|
832 |
+
},
|
833 |
+
{
|
834 |
+
"epoch": 9.89,
|
835 |
+
"learning_rate": 3.5898989898989894e-07,
|
836 |
+
"loss": 0.1878,
|
837 |
+
"step": 3225
|
838 |
+
},
|
839 |
+
{
|
840 |
+
"epoch": 9.97,
|
841 |
+
"learning_rate": 3.539393939393939e-07,
|
842 |
+
"loss": 0.1884,
|
843 |
+
"step": 3250
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 10.05,
|
847 |
+
"learning_rate": 3.488888888888889e-07,
|
848 |
+
"loss": 0.1844,
|
849 |
+
"step": 3275
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 10.12,
|
853 |
+
"learning_rate": 3.4383838383838384e-07,
|
854 |
+
"loss": 0.1744,
|
855 |
+
"step": 3300
|
856 |
+
},
|
857 |
+
{
|
858 |
+
"epoch": 10.2,
|
859 |
+
"learning_rate": 3.3878787878787877e-07,
|
860 |
+
"loss": 0.1848,
|
861 |
+
"step": 3325
|
862 |
+
},
|
863 |
+
{
|
864 |
+
"epoch": 10.28,
|
865 |
+
"learning_rate": 3.337373737373737e-07,
|
866 |
+
"loss": 0.1792,
|
867 |
+
"step": 3350
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 10.35,
|
871 |
+
"learning_rate": 3.286868686868687e-07,
|
872 |
+
"loss": 0.1757,
|
873 |
+
"step": 3375
|
874 |
+
},
|
875 |
+
{
|
876 |
+
"epoch": 10.43,
|
877 |
+
"learning_rate": 3.236363636363636e-07,
|
878 |
+
"loss": 0.1815,
|
879 |
+
"step": 3400
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 10.51,
|
883 |
+
"learning_rate": 3.1858585858585855e-07,
|
884 |
+
"loss": 0.1758,
|
885 |
+
"step": 3425
|
886 |
+
},
|
887 |
+
{
|
888 |
+
"epoch": 10.58,
|
889 |
+
"learning_rate": 3.1353535353535353e-07,
|
890 |
+
"loss": 0.1895,
|
891 |
+
"step": 3450
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 10.66,
|
895 |
+
"learning_rate": 3.0848484848484847e-07,
|
896 |
+
"loss": 0.177,
|
897 |
+
"step": 3475
|
898 |
+
},
|
899 |
+
{
|
900 |
+
"epoch": 10.74,
|
901 |
+
"learning_rate": 3.0343434343434345e-07,
|
902 |
+
"loss": 0.1836,
|
903 |
+
"step": 3500
|
904 |
+
},
|
905 |
+
{
|
906 |
+
"epoch": 10.74,
|
907 |
+
"eval_loss": 0.40053287148475647,
|
908 |
+
"eval_runtime": 2681.3744,
|
909 |
+
"eval_samples_per_second": 3.903,
|
910 |
+
"eval_steps_per_second": 0.122,
|
911 |
+
"eval_wer": 58.90729282066525,
|
912 |
+
"step": 3500
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 10.81,
|
916 |
+
"learning_rate": 2.9838383838383833e-07,
|
917 |
+
"loss": 0.1861,
|
918 |
+
"step": 3525
|
919 |
+
},
|
920 |
+
{
|
921 |
+
"epoch": 10.89,
|
922 |
+
"learning_rate": 2.933333333333333e-07,
|
923 |
+
"loss": 0.1817,
|
924 |
+
"step": 3550
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 10.97,
|
928 |
+
"learning_rate": 2.882828282828283e-07,
|
929 |
+
"loss": 0.1854,
|
930 |
+
"step": 3575
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"epoch": 11.04,
|
934 |
+
"learning_rate": 2.8323232323232323e-07,
|
935 |
+
"loss": 0.1796,
|
936 |
+
"step": 3600
|
937 |
+
},
|
938 |
+
{
|
939 |
+
"epoch": 11.12,
|
940 |
+
"learning_rate": 2.7818181818181816e-07,
|
941 |
+
"loss": 0.1719,
|
942 |
+
"step": 3625
|
943 |
+
},
|
944 |
+
{
|
945 |
+
"epoch": 11.2,
|
946 |
+
"learning_rate": 2.731313131313131e-07,
|
947 |
+
"loss": 0.1718,
|
948 |
+
"step": 3650
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 11.27,
|
952 |
+
"learning_rate": 2.680808080808081e-07,
|
953 |
+
"loss": 0.1774,
|
954 |
+
"step": 3675
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 11.35,
|
958 |
+
"learning_rate": 2.63030303030303e-07,
|
959 |
+
"loss": 0.1795,
|
960 |
+
"step": 3700
|
961 |
+
},
|
962 |
+
{
|
963 |
+
"epoch": 11.43,
|
964 |
+
"learning_rate": 2.5797979797979794e-07,
|
965 |
+
"loss": 0.1784,
|
966 |
+
"step": 3725
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 11.5,
|
970 |
+
"learning_rate": 2.529292929292929e-07,
|
971 |
+
"loss": 0.1689,
|
972 |
+
"step": 3750
|
973 |
+
},
|
974 |
+
{
|
975 |
+
"epoch": 11.58,
|
976 |
+
"learning_rate": 2.4787878787878786e-07,
|
977 |
+
"loss": 0.1792,
|
978 |
+
"step": 3775
|
979 |
+
},
|
980 |
+
{
|
981 |
+
"epoch": 11.66,
|
982 |
+
"learning_rate": 2.4282828282828284e-07,
|
983 |
+
"loss": 0.1749,
|
984 |
+
"step": 3800
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 11.73,
|
988 |
+
"learning_rate": 2.3777777777777777e-07,
|
989 |
+
"loss": 0.1787,
|
990 |
+
"step": 3825
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 11.81,
|
994 |
+
"learning_rate": 2.3272727272727273e-07,
|
995 |
+
"loss": 0.1816,
|
996 |
+
"step": 3850
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 11.89,
|
1000 |
+
"learning_rate": 2.2767676767676766e-07,
|
1001 |
+
"loss": 0.171,
|
1002 |
+
"step": 3875
|
1003 |
+
},
|
1004 |
+
{
|
1005 |
+
"epoch": 11.96,
|
1006 |
+
"learning_rate": 2.2262626262626262e-07,
|
1007 |
+
"loss": 0.1794,
|
1008 |
+
"step": 3900
|
1009 |
+
},
|
1010 |
+
{
|
1011 |
+
"epoch": 12.04,
|
1012 |
+
"learning_rate": 2.1757575757575758e-07,
|
1013 |
+
"loss": 0.1651,
|
1014 |
+
"step": 3925
|
1015 |
+
},
|
1016 |
+
{
|
1017 |
+
"epoch": 12.12,
|
1018 |
+
"learning_rate": 2.1252525252525254e-07,
|
1019 |
+
"loss": 0.1689,
|
1020 |
+
"step": 3950
|
1021 |
+
},
|
1022 |
+
{
|
1023 |
+
"epoch": 12.19,
|
1024 |
+
"learning_rate": 2.0747474747474747e-07,
|
1025 |
+
"loss": 0.1747,
|
1026 |
+
"step": 3975
|
1027 |
+
},
|
1028 |
+
{
|
1029 |
+
"epoch": 12.27,
|
1030 |
+
"learning_rate": 2.0242424242424242e-07,
|
1031 |
+
"loss": 0.1653,
|
1032 |
+
"step": 4000
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 12.27,
|
1036 |
+
"eval_loss": 0.3988991677761078,
|
1037 |
+
"eval_runtime": 2674.7246,
|
1038 |
+
"eval_samples_per_second": 3.913,
|
1039 |
+
"eval_steps_per_second": 0.123,
|
1040 |
+
"eval_wer": 59.77735145505147,
|
1041 |
+
"step": 4000
|
1042 |
+
},
|
1043 |
+
{
|
1044 |
+
"epoch": 12.35,
|
1045 |
+
"learning_rate": 1.9737373737373736e-07,
|
1046 |
+
"loss": 0.1748,
|
1047 |
+
"step": 4025
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 12.42,
|
1051 |
+
"learning_rate": 1.9232323232323231e-07,
|
1052 |
+
"loss": 0.1722,
|
1053 |
+
"step": 4050
|
1054 |
+
},
|
1055 |
+
{
|
1056 |
+
"epoch": 12.5,
|
1057 |
+
"learning_rate": 1.8727272727272727e-07,
|
1058 |
+
"loss": 0.1678,
|
1059 |
+
"step": 4075
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 12.58,
|
1063 |
+
"learning_rate": 1.8222222222222223e-07,
|
1064 |
+
"loss": 0.1784,
|
1065 |
+
"step": 4100
|
1066 |
+
},
|
1067 |
+
{
|
1068 |
+
"epoch": 12.65,
|
1069 |
+
"learning_rate": 1.7717171717171716e-07,
|
1070 |
+
"loss": 0.1795,
|
1071 |
+
"step": 4125
|
1072 |
+
},
|
1073 |
+
{
|
1074 |
+
"epoch": 12.73,
|
1075 |
+
"learning_rate": 1.7212121212121212e-07,
|
1076 |
+
"loss": 0.1752,
|
1077 |
+
"step": 4150
|
1078 |
+
},
|
1079 |
+
{
|
1080 |
+
"epoch": 12.81,
|
1081 |
+
"learning_rate": 1.6707070707070705e-07,
|
1082 |
+
"loss": 0.1725,
|
1083 |
+
"step": 4175
|
1084 |
+
},
|
1085 |
+
{
|
1086 |
+
"epoch": 12.88,
|
1087 |
+
"learning_rate": 1.62020202020202e-07,
|
1088 |
+
"loss": 0.1792,
|
1089 |
+
"step": 4200
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 12.96,
|
1093 |
+
"learning_rate": 1.5696969696969697e-07,
|
1094 |
+
"loss": 0.169,
|
1095 |
+
"step": 4225
|
1096 |
+
},
|
1097 |
+
{
|
1098 |
+
"epoch": 13.04,
|
1099 |
+
"learning_rate": 1.5191919191919193e-07,
|
1100 |
+
"loss": 0.1643,
|
1101 |
+
"step": 4250
|
1102 |
+
},
|
1103 |
+
{
|
1104 |
+
"epoch": 13.11,
|
1105 |
+
"learning_rate": 1.4686868686868686e-07,
|
1106 |
+
"loss": 0.1653,
|
1107 |
+
"step": 4275
|
1108 |
+
},
|
1109 |
+
{
|
1110 |
+
"epoch": 13.19,
|
1111 |
+
"learning_rate": 1.4181818181818182e-07,
|
1112 |
+
"loss": 0.1743,
|
1113 |
+
"step": 4300
|
1114 |
+
},
|
1115 |
+
{
|
1116 |
+
"epoch": 13.27,
|
1117 |
+
"learning_rate": 1.3676767676767675e-07,
|
1118 |
+
"loss": 0.1677,
|
1119 |
+
"step": 4325
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 13.34,
|
1123 |
+
"learning_rate": 1.317171717171717e-07,
|
1124 |
+
"loss": 0.1724,
|
1125 |
+
"step": 4350
|
1126 |
+
},
|
1127 |
+
{
|
1128 |
+
"epoch": 13.42,
|
1129 |
+
"learning_rate": 1.2666666666666666e-07,
|
1130 |
+
"loss": 0.17,
|
1131 |
+
"step": 4375
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 13.5,
|
1135 |
+
"learning_rate": 1.2161616161616162e-07,
|
1136 |
+
"loss": 0.1665,
|
1137 |
+
"step": 4400
|
1138 |
+
},
|
1139 |
+
{
|
1140 |
+
"epoch": 13.57,
|
1141 |
+
"learning_rate": 1.1656565656565657e-07,
|
1142 |
+
"loss": 0.1733,
|
1143 |
+
"step": 4425
|
1144 |
+
},
|
1145 |
+
{
|
1146 |
+
"epoch": 13.65,
|
1147 |
+
"learning_rate": 1.1151515151515151e-07,
|
1148 |
+
"loss": 0.1655,
|
1149 |
+
"step": 4450
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 13.73,
|
1153 |
+
"learning_rate": 1.0646464646464647e-07,
|
1154 |
+
"loss": 0.1642,
|
1155 |
+
"step": 4475
|
1156 |
+
},
|
1157 |
+
{
|
1158 |
+
"epoch": 13.8,
|
1159 |
+
"learning_rate": 1.0141414141414141e-07,
|
1160 |
+
"loss": 0.1693,
|
1161 |
+
"step": 4500
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 13.8,
|
1165 |
+
"eval_loss": 0.3983347713947296,
|
1166 |
+
"eval_runtime": 2686.1318,
|
1167 |
+
"eval_samples_per_second": 3.896,
|
1168 |
+
"eval_steps_per_second": 0.122,
|
1169 |
+
"eval_wer": 59.94622929557675,
|
1170 |
+
"step": 4500
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"epoch": 13.88,
|
1174 |
+
"learning_rate": 9.636363636363636e-08,
|
1175 |
+
"loss": 0.1635,
|
1176 |
+
"step": 4525
|
1177 |
+
},
|
1178 |
+
{
|
1179 |
+
"epoch": 13.96,
|
1180 |
+
"learning_rate": 9.131313131313132e-08,
|
1181 |
+
"loss": 0.1769,
|
1182 |
+
"step": 4550
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"epoch": 14.03,
|
1186 |
+
"learning_rate": 8.626262626262626e-08,
|
1187 |
+
"loss": 0.1747,
|
1188 |
+
"step": 4575
|
1189 |
+
},
|
1190 |
+
{
|
1191 |
+
"epoch": 14.11,
|
1192 |
+
"learning_rate": 8.12121212121212e-08,
|
1193 |
+
"loss": 0.1684,
|
1194 |
+
"step": 4600
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 14.19,
|
1198 |
+
"learning_rate": 7.616161616161616e-08,
|
1199 |
+
"loss": 0.1664,
|
1200 |
+
"step": 4625
|
1201 |
+
},
|
1202 |
+
{
|
1203 |
+
"epoch": 14.26,
|
1204 |
+
"learning_rate": 7.111111111111111e-08,
|
1205 |
+
"loss": 0.1736,
|
1206 |
+
"step": 4650
|
1207 |
+
},
|
1208 |
+
{
|
1209 |
+
"epoch": 14.34,
|
1210 |
+
"learning_rate": 6.606060606060605e-08,
|
1211 |
+
"loss": 0.1609,
|
1212 |
+
"step": 4675
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"epoch": 14.42,
|
1216 |
+
"learning_rate": 6.101010101010101e-08,
|
1217 |
+
"loss": 0.1662,
|
1218 |
+
"step": 4700
|
1219 |
+
},
|
1220 |
+
{
|
1221 |
+
"epoch": 14.49,
|
1222 |
+
"learning_rate": 5.5959595959595956e-08,
|
1223 |
+
"loss": 0.1746,
|
1224 |
+
"step": 4725
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 14.57,
|
1228 |
+
"learning_rate": 5.090909090909091e-08,
|
1229 |
+
"loss": 0.1677,
|
1230 |
+
"step": 4750
|
1231 |
+
},
|
1232 |
+
{
|
1233 |
+
"epoch": 14.65,
|
1234 |
+
"learning_rate": 4.585858585858585e-08,
|
1235 |
+
"loss": 0.1674,
|
1236 |
+
"step": 4775
|
1237 |
+
},
|
1238 |
+
{
|
1239 |
+
"epoch": 14.72,
|
1240 |
+
"learning_rate": 4.0808080808080803e-08,
|
1241 |
+
"loss": 0.168,
|
1242 |
+
"step": 4800
|
1243 |
+
},
|
1244 |
+
{
|
1245 |
+
"epoch": 14.8,
|
1246 |
+
"learning_rate": 3.5757575757575755e-08,
|
1247 |
+
"loss": 0.1691,
|
1248 |
+
"step": 4825
|
1249 |
+
},
|
1250 |
+
{
|
1251 |
+
"epoch": 14.88,
|
1252 |
+
"learning_rate": 3.0707070707070706e-08,
|
1253 |
+
"loss": 0.1641,
|
1254 |
+
"step": 4850
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"epoch": 14.95,
|
1258 |
+
"learning_rate": 2.5656565656565654e-08,
|
1259 |
+
"loss": 0.1678,
|
1260 |
+
"step": 4875
|
1261 |
+
},
|
1262 |
+
{
|
1263 |
+
"epoch": 15.03,
|
1264 |
+
"learning_rate": 2.0606060606060606e-08,
|
1265 |
+
"loss": 0.1672,
|
1266 |
+
"step": 4900
|
1267 |
+
},
|
1268 |
+
{
|
1269 |
+
"epoch": 15.11,
|
1270 |
+
"learning_rate": 1.5555555555555554e-08,
|
1271 |
+
"loss": 0.1656,
|
1272 |
+
"step": 4925
|
1273 |
+
},
|
1274 |
+
{
|
1275 |
+
"epoch": 15.18,
|
1276 |
+
"learning_rate": 1.0505050505050505e-08,
|
1277 |
+
"loss": 0.1637,
|
1278 |
+
"step": 4950
|
1279 |
+
},
|
1280 |
+
{
|
1281 |
+
"epoch": 15.26,
|
1282 |
+
"learning_rate": 5.454545454545455e-09,
|
1283 |
+
"loss": 0.1614,
|
1284 |
+
"step": 4975
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 15.34,
|
1288 |
+
"learning_rate": 4.0404040404040403e-10,
|
1289 |
+
"loss": 0.1616,
|
1290 |
+
"step": 5000
|
1291 |
+
},
|
1292 |
+
{
|
1293 |
+
"epoch": 15.34,
|
1294 |
+
"eval_loss": 0.3983960449695587,
|
1295 |
+
"eval_runtime": 2696.9238,
|
1296 |
+
"eval_samples_per_second": 3.881,
|
1297 |
+
"eval_steps_per_second": 0.122,
|
1298 |
+
"eval_wer": 59.830041341295356,
|
1299 |
+
"step": 5000
|
1300 |
+
},
|
1301 |
+
{
|
1302 |
+
"epoch": 15.34,
|
1303 |
+
"step": 5000,
|
1304 |
+
"total_flos": 9.22520949202944e+19,
|
1305 |
+
"train_loss": 0.23562390689849855,
|
1306 |
+
"train_runtime": 67707.284,
|
1307 |
+
"train_samples_per_second": 4.726,
|
1308 |
+
"train_steps_per_second": 0.074
|
1309 |
+
}
|
1310 |
+
],
|
1311 |
+
"logging_steps": 25,
|
1312 |
+
"max_steps": 5000,
|
1313 |
+
"num_input_tokens_seen": 0,
|
1314 |
+
"num_train_epochs": 16,
|
1315 |
+
"save_steps": 500,
|
1316 |
+
"total_flos": 9.22520949202944e+19,
|
1317 |
+
"train_batch_size": 32,
|
1318 |
+
"trial_name": null,
|
1319 |
+
"trial_params": null
|
1320 |
+
}
|