Commit
•
f363fc1
1
Parent(s):
c90221a
End of training
Browse files
README.md
CHANGED
@@ -10,7 +10,7 @@ datasets:
|
|
10 |
metrics:
|
11 |
- wer
|
12 |
model-index:
|
13 |
-
- name: BANG -
|
14 |
results:
|
15 |
- task:
|
16 |
name: Automatic Speech Recognition
|
@@ -18,24 +18,24 @@ model-index:
|
|
18 |
dataset:
|
19 |
name: Radio-Modified Common Voice 11.0
|
20 |
type: mozilla-foundation/common_voice_11_0
|
21 |
-
config:
|
22 |
split: test
|
23 |
args: 'config: en, split: test'
|
24 |
metrics:
|
25 |
- name: Wer
|
26 |
type: wer
|
27 |
-
value:
|
28 |
---
|
29 |
|
30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
31 |
should probably proofread and complete it, then remove this comment. -->
|
32 |
|
33 |
-
# BANG -
|
34 |
|
35 |
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Radio-Modified Common Voice 11.0 dataset.
|
36 |
It achieves the following results on the evaluation set:
|
37 |
-
- Loss: 0.
|
38 |
-
- Wer:
|
39 |
|
40 |
## Model description
|
41 |
|
@@ -66,12 +66,12 @@ The following hyperparameters were used during training:
|
|
66 |
|
67 |
### Training results
|
68 |
|
69 |
-
| Training Loss | Epoch
|
70 |
-
|
71 |
-
| 0.
|
72 |
-
| 0.
|
73 |
-
| 0.
|
74 |
-
| 0.
|
75 |
|
76 |
|
77 |
### Framework versions
|
|
|
10 |
metrics:
|
11 |
- wer
|
12 |
model-index:
|
13 |
+
- name: BANG - v2 (EN)
|
14 |
results:
|
15 |
- task:
|
16 |
name: Automatic Speech Recognition
|
|
|
18 |
dataset:
|
19 |
name: Radio-Modified Common Voice 11.0
|
20 |
type: mozilla-foundation/common_voice_11_0
|
21 |
+
config: en
|
22 |
split: test
|
23 |
args: 'config: en, split: test'
|
24 |
metrics:
|
25 |
- name: Wer
|
26 |
type: wer
|
27 |
+
value: 20.561047043748857
|
28 |
---
|
29 |
|
30 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
31 |
should probably proofread and complete it, then remove this comment. -->
|
32 |
|
33 |
+
# BANG - v2 (EN)
|
34 |
|
35 |
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co/openai/whisper-large-v3) on the Radio-Modified Common Voice 11.0 dataset.
|
36 |
It achieves the following results on the evaluation set:
|
37 |
+
- Loss: 0.2650
|
38 |
+
- Wer: 20.5610
|
39 |
|
40 |
## Model description
|
41 |
|
|
|
66 |
|
67 |
### Training results
|
68 |
|
69 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
70 |
+
|:-------------:|:------:|:----:|:---------------:|:-------:|
|
71 |
+
| 0.7709 | 0.25 | 1000 | 0.6383 | 35.6607 |
|
72 |
+
| 0.4424 | 1.2443 | 2000 | 0.4248 | 26.8037 |
|
73 |
+
| 0.2823 | 2.2385 | 3000 | 0.3117 | 22.4425 |
|
74 |
+
| 0.2429 | 3.2328 | 4000 | 0.2650 | 20.5610 |
|
75 |
|
76 |
|
77 |
### Framework versions
|