icefall/egs/mdcc/ASR/RESULTS.md
zr_jin c3f6f28116
Zipformer recipe for Cantonese dataset MDCC (#1537)
* init commit

* Create README.md

* handle code switching cases

* misc. fixes

* added manifest statistics

* init commit for the zipformer recipe

* added scripts for exporting model

* added RESULTS.md

* added scripts for streaming related stuff

* doc str fixed
2024-03-13 10:01:28 +08:00

1.1 KiB

Results

Zipformer

See https://github.com/k2-fsa/icefall/pull/1537

./zipformer

normal-scaled model, number of model parameters: 74470867, i.e., 74.47 M
test valid comment
greedy search 7.45 7.51 --epoch 45 --avg 35
modified beam search 6.68 6.73 --epoch 45 --avg 35
fast beam search 7.22 7.28 --epoch 45 --avg 35

The training command:

export CUDA_VISIBLE_DEVICES="0,1,2,3"

./zipformer/train.py \
  --world-size 4 \
  --start-epoch 1 \
  --num-epochs 50 \
  --use-fp16 1 \
  --exp-dir ./zipformer/exp \
  --max-duration 1000 

The decoding command:

 ./zipformer/decode.py \
   --epoch 45 \
   --avg 35 \
   --exp-dir ./zipformer/exp \
   --decoding-method greedy_search  # modified_beam_search

The pretrained model is available at: https://huggingface.co/zrjin/icefall-asr-mdcc-zipformer-2024-03-11/