Seyfelislem
commited on
Commit
•
5a13aea
1
Parent(s):
3d63bf4
Update README.md
Browse files
README.md
CHANGED
@@ -6,7 +6,7 @@ datasets:
|
|
6 |
metrics:
|
7 |
- wer
|
8 |
model-index:
|
9 |
-
- name:
|
10 |
results:
|
11 |
- task:
|
12 |
name: Automatic Speech Recognition
|
@@ -20,18 +20,16 @@ model-index:
|
|
20 |
metrics:
|
21 |
- name: Wer
|
22 |
type: wer
|
23 |
-
value: 6.
|
24 |
---
|
25 |
|
26 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
27 |
should probably proofread and complete it, then remove this comment. -->
|
28 |
|
29 |
-
#
|
|
|
|
|
30 |
|
31 |
-
This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the afrispeech-200 dataset.
|
32 |
-
It achieves the following results on the evaluation set:
|
33 |
-
- Loss: 0.1213
|
34 |
-
- Wer: 6.1529
|
35 |
|
36 |
## Model description
|
37 |
|
@@ -64,19 +62,7 @@ The following hyperparameters were used during training:
|
|
64 |
|
65 |
### Training results
|
66 |
|
67 |
-
|
68 |
-
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
69 |
-
| 0.1578 | 0.1 | 200 | 0.1283 | 5.8313 |
|
70 |
-
| 0.1253 | 0.2 | 400 | 0.1475 | 6.8282 |
|
71 |
-
| 0.236 | 0.3 | 600 | 0.1505 | 7.2033 |
|
72 |
-
| 0.1628 | 0.4 | 800 | 0.1398 | 8.2217 |
|
73 |
-
| 0.1213 | 1.05 | 1000 | 0.1316 | 6.1743 |
|
74 |
-
| 0.0569 | 1.15 | 1200 | 0.1417 | 6.7960 |
|
75 |
-
| 0.0598 | 1.25 | 1400 | 0.1445 | 7.0104 |
|
76 |
-
| 0.1262 | 1.35 | 1600 | 0.1269 | 6.4101 |
|
77 |
-
| 0.0462 | 2.0 | 1800 | 0.1186 | 5.9706 |
|
78 |
-
| 0.0487 | 2.1 | 2000 | 0.1213 | 6.1529 |
|
79 |
-
|
80 |
|
81 |
### Framework versions
|
82 |
|
|
|
6 |
metrics:
|
7 |
- wer
|
8 |
model-index:
|
9 |
+
- name: afripspeech_data_aug
|
10 |
results:
|
11 |
- task:
|
12 |
name: Automatic Speech Recognition
|
|
|
20 |
metrics:
|
21 |
- name: Wer
|
22 |
type: wer
|
23 |
+
value: 6.008
|
24 |
---
|
25 |
|
26 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
27 |
should probably proofread and complete it, then remove this comment. -->
|
28 |
|
29 |
+
# afripspeech_data_aug
|
30 |
+
|
31 |
+
This model is a fine-tuned version of [afrispeech_large_A100](https://huggingface.co/Seyfelislem/afrispeech_large_A100) on the afrispeech-200 dataset.
|
32 |
|
|
|
|
|
|
|
|
|
33 |
|
34 |
## Model description
|
35 |
|
|
|
62 |
|
63 |
### Training results
|
64 |
|
65 |
+
https://huggingface.co/Seyfelislem/afrispeech_large_A100/tensorboard
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
66 |
|
67 |
### Framework versions
|
68 |
|