Update README.md
Browse files
README.md
CHANGED
@@ -17,11 +17,12 @@ should probably proofread and complete it, then remove this comment. -->
|
|
17 |
# Whisper Small GA-EN Speech Translation
|
18 |
|
19 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on an unknown dataset.
|
20 |
-
|
21 |
-
- Loss: 1.
|
22 |
-
- Bleu:
|
23 |
-
- Chrf:
|
24 |
-
- Wer:
|
|
|
25 |
|
26 |
## Model description
|
27 |
|
@@ -64,11 +65,11 @@ The following hyperparameters were used during training:
|
|
64 |
| 0.9166 | 0.86 | 800 | 26.87 | 43.16 | 1.4104 | 71.3192 |
|
65 |
| 0.848 | 0.97 | 900 | 25.95 | 42.61 | 1.3556 | 75.6866 |
|
66 |
| 0.3712 | 1.08 | 1000 | 22.4 | 41.02 | 1.3936 | 87.2580 |
|
67 |
-
| 0.4415
|
68 |
-
| 0.4166
|
69 |
-
| 0.387
|
70 |
-
| 0.3714
|
71 |
-
| 0.3695
|
72 |
|
73 |
|
74 |
### Framework versions
|
|
|
17 |
# Whisper Small GA-EN Speech Translation
|
18 |
|
19 |
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on an unknown dataset.
|
20 |
+
The best model is at checkpoint 1400, epoch 1.51, and it achieves the following results on the evaluation set:
|
21 |
+
- Loss: 1.3989
|
22 |
+
- Bleu: 28.53
|
23 |
+
- Chrf: 44.93
|
24 |
+
- Wer: 68.1675
|
25 |
+
|
26 |
|
27 |
## Model description
|
28 |
|
|
|
65 |
| 0.9166 | 0.86 | 800 | 26.87 | 43.16 | 1.4104 | 71.3192 |
|
66 |
| 0.848 | 0.97 | 900 | 25.95 | 42.61 | 1.3556 | 75.6866 |
|
67 |
| 0.3712 | 1.08 | 1000 | 22.4 | 41.02 | 1.3936 | 87.2580 |
|
68 |
+
| 0.4415 | 1.19 | 1100 | 28.13 | 43.0 | 1.4157 | 68.0324 |
|
69 |
+
| 0.4166 | 1.29 | 1200 | 27.75 | 44.39 | 1.4206 | 71.1391 |
|
70 |
+
| 0.387 | 1.4 | 1300 | 28.48 | 44.44 | 1.4083 | 69.4282 |
|
71 |
+
| 0.3714 | 1.51 | 1400 | 28.53 | 44.93 | 1.3989 | 68.1675 |
|
72 |
+
| 0.3695 | 1.62 | 1500 | 26.13 | 43.65 | 1.4049 | 76.9923 |
|
73 |
|
74 |
|
75 |
### Framework versions
|