Update README.md
Browse files
README.md
CHANGED
|
@@ -54,7 +54,6 @@ The following hyperparameters were used during training:
|
|
| 54 |
| 0.8953 | 4.0 | 28 | 0.8911 | 52.9234 |
|
| 55 |
| 0.8159 | 5.0 | 35 | 0.8875 | 34.9798 |
|
| 56 |
|
| 57 |
-
| | WER |
|
| 58 |
| Model | WER (CV18) | WER (Gowejee) | WER (LOTUS-TRD) | WER (Thai Dialect) | WER (Elderly) | WER (Gigaspeech2) | WER (Fleurs) | WER (Distant Meeting) | WER (Podcast) |
|
| 59 |
|:----------------------------------------|:----------------------:|:-------------------------:|:----------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|
|
| 60 |
| whisper-large-v3 | 18.75 | 46.59 | 48.14 | 57.82 | 12.27 | 33.26 | 24.08 | 72.57 | 41.24 |
|
|
@@ -65,6 +64,24 @@ The following hyperparameters were used during training:
|
|
| 65 |
|
| 66 |
|
| 67 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
### Framework versions
|
| 69 |
|
| 70 |
- Transformers 4.44.2
|
|
|
|
| 54 |
| 0.8953 | 4.0 | 28 | 0.8911 | 52.9234 |
|
| 55 |
| 0.8159 | 5.0 | 35 | 0.8875 | 34.9798 |
|
| 56 |
|
|
|
|
| 57 |
| Model | WER (CV18) | WER (Gowejee) | WER (LOTUS-TRD) | WER (Thai Dialect) | WER (Elderly) | WER (Gigaspeech2) | WER (Fleurs) | WER (Distant Meeting) | WER (Podcast) |
|
| 58 |
|:----------------------------------------|:----------------------:|:-------------------------:|:----------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|:--------------------------:|
|
| 59 |
| whisper-large-v3 | 18.75 | 46.59 | 48.14 | 57.82 | 12.27 | 33.26 | 24.08 | 72.57 | 41.24 |
|
|
|
|
| 64 |
|
| 65 |
|
| 66 |
|
| 67 |
+
| Model | ASR-th (WER↓) | ASR-en (WER↓) | ASR-en (WER↓) | Emotion (Acc↑, F1↑)| Gender (Acc↑, F1↑) |
|
| 68 |
+
| | CV18 th 1000 | CV18 En 1000 | Librispeech En 1000 | ThaiSER Emotion | ThaiSER Gender |
|
| 69 |
+
|:----------------------------:|:------------------------:|:------------------------:|:------------------------------:|:------------------:|:--------------------:|
|
| 70 |
+
| Typhoon-Audio-Preview | 13.26 | 13.34 (partial result) | 5.07 (partial result) | 41.50, 33.48 | 96.20, 96.69 |
|
| 71 |
+
| DIVA | 69.15 (partial result) | 37.40 | 49.06 | 18.64, 8.16 | 47.50, 35.90 |
|
| 72 |
+
| Gemini-1.5-Pro | 16.49 | 12.94 | 25.83 | 26.00, 18.26 | 79.66, 77.32 |
|
| 73 |
+
| Pathumma-llm-audio-1.0.0 | 12.03 | 12.20 | 11.36 | 42.30, 36.88 | 90.30, 92.07 |
|
| 74 |
+
|
| 75 |
+
|
| 76 |
+
|
| 77 |
+
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
| 78 |
+
|:-------------:|:-----:|:----:|:---------------:|:-------:|
|
| 79 |
+
| No log | 1.0 | 7 | 0.9713 | 37.2984 |
|
| 80 |
+
| 1.1414 | 2.0 | 14 | 0.9285 | 34.4758 |
|
| 81 |
+
| 0.8953 | 3.0 | 21 | 0.9022 | 35.2823 |
|
| 82 |
+
| 0.8953 | 4.0 | 28 | 0.8911 | 52.9234 |
|
| 83 |
+
| 0.8159 | 5.0 | 35 | 0.8875 | 34.9798 |
|
| 84 |
+
|
| 85 |
### Framework versions
|
| 86 |
|
| 87 |
- Transformers 4.44.2
|