update model card README.md
Browse files
README.md
CHANGED
@@ -1,42 +1,38 @@
|
|
1 |
---
|
2 |
-
language:
|
3 |
-
- lv
|
4 |
license: apache-2.0
|
5 |
tags:
|
6 |
-
- whisper-event
|
7 |
-
- hf-asr-leaderboard
|
8 |
- generated_from_trainer
|
9 |
datasets:
|
10 |
-
-
|
11 |
metrics:
|
12 |
- wer
|
13 |
model-index:
|
14 |
-
- name:
|
15 |
results:
|
16 |
- task:
|
17 |
name: Automatic Speech Recognition
|
18 |
type: automatic-speech-recognition
|
19 |
dataset:
|
20 |
-
name:
|
21 |
-
type:
|
22 |
config: lv
|
23 |
split: test
|
24 |
args: lv
|
25 |
metrics:
|
26 |
- name: Wer
|
27 |
type: wer
|
28 |
-
value:
|
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 |
-
#
|
35 |
|
36 |
-
This model is a fine-tuned version of [
|
37 |
It achieves the following results on the evaluation set:
|
38 |
-
- Loss: 0.
|
39 |
-
- Wer:
|
40 |
|
41 |
## Model description
|
42 |
|
@@ -55,27 +51,26 @@ More information needed
|
|
55 |
### Training hyperparameters
|
56 |
|
57 |
The following hyperparameters were used during training:
|
58 |
-
- learning_rate:
|
59 |
-
- train_batch_size:
|
60 |
- eval_batch_size: 32
|
61 |
- seed: 42
|
62 |
- distributed_type: multi-GPU
|
|
|
|
|
63 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
64 |
- lr_scheduler_type: cosine
|
65 |
-
- lr_scheduler_warmup_steps:
|
66 |
-
- training_steps:
|
67 |
|
68 |
### Training results
|
69 |
|
70 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
71 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
72 |
-
| 0.
|
73 |
-
| 0.
|
74 |
-
| 0.
|
75 |
-
| 0.
|
76 |
-
| 0.2034 | 16.03 | 1000 | 0.3179 | 27.4763 |
|
77 |
-
| 0.1478 | 19.04 | 1200 | 0.3193 | 27.5237 |
|
78 |
-
| 0.2169 | 23.01 | 1400 | 0.3198 | 27.5047 |
|
79 |
|
80 |
|
81 |
### Framework versions
|
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
tags:
|
|
|
|
|
4 |
- generated_from_trainer
|
5 |
datasets:
|
6 |
+
- common_voice_11_0
|
7 |
metrics:
|
8 |
- wer
|
9 |
model-index:
|
10 |
+
- name: p4b/whisper-large-v2-lv
|
11 |
results:
|
12 |
- task:
|
13 |
name: Automatic Speech Recognition
|
14 |
type: automatic-speech-recognition
|
15 |
dataset:
|
16 |
+
name: common_voice_11_0
|
17 |
+
type: common_voice_11_0
|
18 |
config: lv
|
19 |
split: test
|
20 |
args: lv
|
21 |
metrics:
|
22 |
- name: Wer
|
23 |
type: wer
|
24 |
+
value: 19.97153700189753
|
25 |
---
|
26 |
|
27 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
28 |
should probably proofread and complete it, then remove this comment. -->
|
29 |
|
30 |
+
# p4b/whisper-large-v2-lv
|
31 |
|
32 |
+
This model is a fine-tuned version of [p4b/whisper-large-v2-lv](https://huggingface.co/p4b/whisper-large-v2-lv) on the common_voice_11_0 dataset.
|
33 |
It achieves the following results on the evaluation set:
|
34 |
+
- Loss: 0.2593
|
35 |
+
- Wer: 19.9715
|
36 |
|
37 |
## Model description
|
38 |
|
|
|
51 |
### Training hyperparameters
|
52 |
|
53 |
The following hyperparameters were used during training:
|
54 |
+
- learning_rate: 1e-07
|
55 |
+
- train_batch_size: 32
|
56 |
- eval_batch_size: 32
|
57 |
- seed: 42
|
58 |
- distributed_type: multi-GPU
|
59 |
+
- gradient_accumulation_steps: 2
|
60 |
+
- total_train_batch_size: 64
|
61 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
62 |
- lr_scheduler_type: cosine
|
63 |
+
- lr_scheduler_warmup_steps: 100
|
64 |
+
- training_steps: 900
|
65 |
|
66 |
### Training results
|
67 |
|
68 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
69 |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
70 |
+
| 0.7919 | 3.03 | 200 | 0.2793 | 22.5806 |
|
71 |
+
| 0.4409 | 6.05 | 400 | 0.2651 | 20.6072 |
|
72 |
+
| 0.4393 | 10.01 | 600 | 0.2600 | 20.0664 |
|
73 |
+
| 0.4975 | 13.04 | 800 | 0.2593 | 19.9715 |
|
|
|
|
|
|
|
74 |
|
75 |
|
76 |
### Framework versions
|