Update README.md
Browse files
README.md
CHANGED
|
@@ -38,7 +38,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 38 |
# Whisper Large GA-EN Speech Translation
|
| 39 |
|
| 40 |
This model is a fine-tuned version of [openai/whisper-large](https://huggingface.co/openai/whisper-large) on the IWSLT-2023, FLEURS, BiteSize, SpokenWords, Tatoeba, and Wikimedia dataset.
|
| 41 |
-
|
|
|
|
|
|
|
| 42 |
- Loss: 1.1742
|
| 43 |
- Bleu: 30.16
|
| 44 |
- Chrf: 50.72
|
|
|
|
| 38 |
# Whisper Large GA-EN Speech Translation
|
| 39 |
|
| 40 |
This model is a fine-tuned version of [openai/whisper-large](https://huggingface.co/openai/whisper-large) on the IWSLT-2023, FLEURS, BiteSize, SpokenWords, Tatoeba, and Wikimedia dataset.
|
| 41 |
+
The datasets are augmented in two ways: noise augmentation, and truncating low-magnitude samples.
|
| 42 |
+
The best model checkpoint (this version) based on ChrF is at step 2900, epoch 0.6349,
|
| 43 |
+
and it achieves the following results on the evaluation set:
|
| 44 |
- Loss: 1.1742
|
| 45 |
- Bleu: 30.16
|
| 46 |
- Chrf: 50.72
|