icefall
icefall copied to clipboard
k2-fsa
I am curious about the reasoning to do nothing when an exception is triggered in `BalancerFunction`. https://github.com/k2-fsa/icefall/blob/23913f6afdea59caf703e3ac715852810cd246ad/egs/librispeech/ASR/zipformer/scaling.py#L798-L801 In my case, a CUDA OOM exception was ignored in one of my...
UnboundLocalError: local variable 'loss' referenced before assignment ``` python def compute_validation_loss( params: AttributeDict, model: nn.Module, valid_dl: torch.utils.data.DataLoader, world_size: int = 1, ) -> MetricsTracker: """Run the validation process. The validation...
Hi, I'm doing 80k hours training using the default zipformer of wenetspeech recipe. The token size is about 6800 (Chinese char + English bpe) and GPU memory is 32G. The...
Hi Guys, amazing works with the icefall recipes. I am quite new to using the recipes and having a hard time creating a custom dataset using lhotse that I have...
In parse_hyp_and_timestamp() after tokens = sp.id_to_piece(res.hyps[i]) I have such tokens: ▁', '', '-', 'mo', 'du', 'le', then words = sp.decode_pieces(tokens).split() gives me two words instead of one '⁇', '-module' but...
I was wondering if there are any existing recipes for the [HAT](https://arxiv.org/abs/2003.07705) model. It is a straightforward change by modeling the blank distribution as a Bernoulli distribution, and was shown...
When you export non-streaming zipformer2 model to onnx using export-onnx script, the result model causes many warnings in sherpa: ``` 2023-09-25 22:18:11.589877696 [W:onnxruntime:, graph.cc:3553 CleanUnusedInitializersAndNodeArgs] Removing initializer '/encoder/5/encoder/encoder_pos/Constant_37_output_0'. It is...
I'm looking for hints about how to correctly replicate the exact normalization process that is applied during simulated streaming with `decode.py` in `streaming_decode.py` for the streaming zipformer. I have been...
This PR shows how to use Lhotse Shar data format to train models in Icefall. I only added the necessary adaptation to the basic TDNN LSTM recipe as I don't...
