Commit
•
bf3485e
1
Parent(s):
bcae806
First commit
Browse files
README.md
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Introduction
|
| 2 |
+
|
| 3 |
+
This repo contains models for speech recognition in ncnn format,
|
| 4 |
+
which is converted from
|
| 5 |
+
<https://huggingface.co/ptrnull/icefall-asr-conv-emformer-transducer-stateless2-zh>.
|
| 6 |
+
|
| 7 |
+
Please see
|
| 8 |
+
<https://github.com/k2-fsa/sherpa-ncnn>
|
| 9 |
+
for usage.
|
| 10 |
+
|
| 11 |
+
It supports both English as well as Chinese. That is, you can have
|
| 12 |
+
both English and Chinese in a single utterance.
|
| 13 |
+
|
| 14 |
+
Please visit <https://huggingface.co/spaces/k2-fsa/automatic-speech-recognition>
|
| 15 |
+
to try the model in your browser without installing anything.
|
| 16 |
+
|
| 17 |
+
|