Update README.md
Browse files
README.md
CHANGED
|
@@ -7,4 +7,25 @@ language:
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
pipeline_tag: automatic-speech-recognition
|
| 10 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
metrics:
|
| 8 |
- wer
|
| 9 |
pipeline_tag: automatic-speech-recognition
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
training on full commonvoice
|
| 13 |
+
|
| 14 |
+
The WERs are:
|
| 15 |
+
|
| 16 |
+
| decoding method | chunk size | test | comment | decoding mode |
|
| 17 |
+
| -------------------- | ---------- | ------ | -------------------- | -------------------- |
|
| 18 |
+
| greedy search | 640ms | 10.90 | --epoch 30 --avg 9 | simulated streaming |
|
| 19 |
+
| modified beam search | 640ms | 10.55 | --epoch 30 --avg 9 | simulated streaming |
|
| 20 |
+
| fast beam search | 640ms | 10.75 | --epoch 30 --avg 9 | simulated streaming |
|
| 21 |
+
|
| 22 |
+
training on full librispeech then finetune on full commonvoice
|
| 23 |
+
|
| 24 |
+
The WERs are:
|
| 25 |
+
|
| 26 |
+
| decoding method | chunk size | test | comment | decoding mode |
|
| 27 |
+
| -------------------- | ---------- | ------ | -------------------- | -------------------- |
|
| 28 |
+
| greedy search | 640ms | 10.57 | --epoch 29 --avg 9 | simulated streaming |
|
| 29 |
+
| modified beam search | 640ms | 10.19 | --epoch 29 --avg 9 | simulated streaming |
|
| 30 |
+
| fast beam search | 640ms | 10.25 | --epoch 29 --avg 9 | simulated streaming |
|
| 31 |
+
|