snowfall
snowfall copied to clipboard
k2-fsa
Moved to https://github.com/k2-fsa/icefall
Guys, Does anyone want to take on the task of getting expected times per phone in n-best lists? FINALLY submitting this to the right repo. Would be something like: -...
Previously, `ans.phones` was always the same as `ans.labels` after `k2.compose` if `inner_labels=phones` was requested. https://github.com/k2-fsa/k2/pull/667 fixed this bug and we need to re-run the training script since the columns of...
egs/librispeech/asr/simple_v1/run.sh python3 ./mmi_mbr_train.py 2021-02-04 22:15:14,543 INFO [mmi_mbr_train.py:552] epoch 0, learning rate 0.001 Traceback (most recent call last): File "./mmi_mbr_train.py", line 617, in main() File "./mmi_mbr_train.py", line 564, in main global_batch_idx_train=global_batch_idx_train,...
In future can we have more detail in RESULTS.md? Right now I only see results for MMI+MBR, it would be nice to have results for CTC and MMI as well....
egs/aishell/asr/simple_v1/path.sh I see that I need to set up the PATH environment where the script will run. But I can't find 'env.sh' in master branch at the location : $SNOWFALL_ROOT/tools/...
I just want to make people aware of this issue I create in Lhotse, regarding padding. https://github.com/lhotse-speech/lhotse/issues/189
We are going to want a smaller-than-normal decoding graph for training purposes. The creation process will be similar to the regular one except: - use a unigram language model (later...
Hi, can you share the e2e inference code from audio file? Input an audio, just return the texts.
Hello, I am reading train.py and decode.py. For me, It is difficult to know how to combine the nenural net log-softmax outputs and fsa. Could you provide some papers or...
Hi, these examples are excellent, and can we training on multi-GPUs with K2/lhotse? Just like the DDP in pytorch?