icefall
icefall copied to clipboard
k2-fsa
Hi guys, Recently, we have been trying to get starting timestamps for each of non-blank tokens using Zipformer2-Pruned-transducer, but found the time is weirdly inaccurate. We tested on a 16000Hz...
I am training a transformer LM ,and at some intermediate batch i am gettting loss nan and ppl nan. with hooks -ValueError: The sum of module.input_embedding.output is not finite:
I'm finetuning the model in zipformer. When I finetune with 100hrs of data, there was no issue but when I finetune the model with 3000hrs of data I'm facing the...
I am trying to run the Zipformer model using my custom dataset. For that the steps that I have followed are given below:- - I have prepared the data by...
Hi Dan and team, I read your recent article on CR-CTC. Is that your top recommended approach right now? Thanks, AlexG
I am trying to run Zipformer model using my custom dataset. For that the steps that I have followed are given below:- 1. I have prepared the data by running...
Hi everyone, I have some very noisy audio data with a lot of added noise, distortion from saturation, and bit quantization errors. I’d prefer not to spend time training a...
Hello All, we are training a zip former model for about 3400 hours of Tamil data. This is in reference with https://github.com/k2-fsa/icefall/issues/1751 We have NVIDIA A6000 50GB GPU. Getting the...
In my practice on aishell -conformer_ctc-asr-task, I found that the script only implemented single machine - multi gpus, which is inconvenient for our gpusevrers. So I modified train.py, hope can...
