-
Most visitors need to modify the ASR, and duplicating them just to be able to modify ASR will add more code, slower compilation, etc.
I think the AST->ASR does not modify AST, so having that visito…
-
```python
from numpy import empty, float32
def main():
a: f32[10000] = empty(10000, dtype=float32)
b: f32[10000]
b[:] = a[:]
main()
```
gives:
```console
$ lpython a.py
Inte…
-
# Task Name
Code-switching refers to the phenomenon where a speaker alternates between two or more languages or dialects within a conversation, sentence, or phrase. This presents a significant chal…
-
Hi there,
I am trying to load the downloaded Hubert Large with 960hr finetuning from here: https://github.com/pytorch/fairseq/tree/main/examples/hubert
I downloaded the model, stored the checkpo…
-
I'm wondering if its on the roadmap to add support for OpenAI's whisper. It could possibly be done using a packaged docker container something like this: https://github.com/ahmetoner/whisper-asr-webse…
-
使用流式ASR服务时，语音识别效果差，如识别结果为“晴初是身份证”，这种问题应该怎么优化呢，换模型吗？
调用方式如下：
from paddlespeech.server.bin.paddlespeech_server import ServerExecutor
streaming_asr_server = ServerExecutor()
streaming_asr_server(conf…
-
Debug: audio_data shape: (76117,), sample_rate: 16000.0
Debug: audio_tensor shape: torch.Size([1, 76117])
Debug: Exception occurred Calculated padded input size per channel: (0). Kernel size: (1). K…
-
Hi, I am running stage 3 of [egs/librimix/ConvTasNet/run.sh](https://github.com/asteroid-team/asteroid/blob/master/egs/librimix/ConvTasNet/run.sh#L119).
I used ```--compute_wer 1 --eval_mode max``` …
-
```diff
--- a/src/libasr/ASR.asdl
+++ b/src/libasr/ASR.asdl
@@ -343,6 +343,8 @@ expr
ttype
= Integer(int kind, dimension* dims)
+ | BitVector(int kind, dimension* dims)
+ | Modul…
-
### Model description
I'm trying to figure out how to create tflite models(int8/float32 ) for OpenAI->Whisper ASR model (Tiny.en.pt)
Somehow below generated tflite file getting crashed while running…