icefall

icefall copied to clipboard

Empty Hypotheses with n_best-oracle Decoding in ctc_decode.py

3

When I decode using the `n_best-oracle` method in `ctc_decode.py` with my dataset, I consistently get empty hypotheses, resulting in a 100% WER. I trained zipformer model on my own dataset...

moadel2002

I used to compile it in windows but still an error presenting

8

```bash mab@DESKTOP-1SOLNEE MINGW64 ~ $ cd ~/tmp/icefall/egs/ljspeech/TTS/ ``` ```bash mab@DESKTOP-1SOLNEE MINGW64 ~/tmp/icefall/egs/ljspeech/TTS (master) $ ./prepare.sh 2024-11-21 18:19:47 (prepare.sh:25:main) dl_dir: /c/Users/mab/tmp/icefall/egs/ljspeech/TTS/download 2024-11-21 18:19:47 (prepare.sh:28:main) Stage -1: build monotonic_align lib running build_ext...

mablue

I'm trying to train a streaming zipformer model with 2 custom datasets, first one with 160 hours of training data and second one with 37 hours training data. For the...

spacetronics

whisper load_model

3

I have a question about the model loading of whisper。https://github.com/k2-fsa/icefall/blob/master/egs/wenetspeech/ASR/whisper/train.py line 752 : > model = whisper.load_model(params.model_name, "cpu") We can see the official code provided by whisper。There are two ways...

spiderlx

So that you can deploy the exported rknn model using https://github.com/k2-fsa/sherpa-onnx/pull/1908 on rockchip NPUs, e.g., on RK3588.

csukuangfj

RuntimeError: CUDA error: device-side assert triggered

1

torch是2.3.0 K2: Version: 1.24.4.dev20241029+cuda12.1.torch2.3.0 lhotse：Version: 1.30.0.dev0+git.13c7616.clean 在使用mdcc粤语数据集，执行prepare.sh的第十二步训练rnn_lm语言模型时，出现错误 Traceback (most recent call last): File "/root/autodl-tmp/.autodl/icefall/egs/mdcc/ASR/../../../icefall/rnn_lm/train.py", line 689, in main() File "/root/autodl-tmp/.autodl/icefall/egs/mdcc/ASR/../../../icefall/rnn_lm/train.py", line 682, in main run(rank=0, world_size=1, args=args) File "/root/autodl-tmp/.autodl/icefall/egs/mdcc/ASR/../../../icefall/rnn_lm/train.py", line...

JiangSYKY

Decoding using Domain Language Model in sherpa

3

We have trained a zipformer model with 3500 hours .We built a domain specific language model using n gram model using build_lm.py script. We found the performance same with and...

bsshruthi22

How to delete the Range operator node in ONNX export

5

Description: We noticed that in the process of exporting the PyTorch model to the ONNX format, the generated ONNX model contains a Range operator. The locations of these nodes are...

Buer-Chen

arpa2fst.py output an empty G_3_gram.fst.txt without any error messages

5

Hi, I use prepare_lm.sh to build HLG. However, the output file G_3_gram.fst.txt is empty. I can convert vword.3gram.th1e-7.arpa to G.fst by using arpa2fst in kaldi. but, a txt format G.fst...

MeeElves

Hi, i'm trying to train a streaming zipformer transducer model for Indonesian language. I use Colab to train and i need help on how should i train this model. There's...

spacetronics