mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-08-09 18:12:19 +00:00
8838fe0bd2
* Add soft links in pruned_transducer_stateless7 for CommonVoice * Add python files * Update prepare.sh * Update normalization * Fix for soft links * Add some docs * Add export * Update egs/commonvoice/ASR/RESULTS.md Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * Add export for onnx --------- Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com>
1.6 KiB
1.6 KiB
Results
GigaSpeech BPE training results (Pruned Stateless Transducer 7)
pruned_transducer_stateless7
See #997 for more details.
Number of model parameters: 70369391, i.e., 70.37 M
The best WER, as of 2023-04-17, for Common Voice English 13.0 (cv-corpus-13.0-2023-03-09/en) is below:
Results are:
| Dev | Test | |
|---|---|---|
| greedy search | 9.96 | 12.54 |
| modified beam search | 9.86 | 12.48 |
To reproduce the above result, use the following commands for training:
export CUDA_VISIBLE_DEVICES="0,1,2,3"
./pruned_transducer_stateless7/train.py \
--world-size 4 \
--num-epochs 30 \
--start-epoch 1 \
--use-fp16 1 \
--exp-dir pruned_transducer_stateless7/exp \
--max-duration 550
and the following commands for decoding:
# greedy search
./pruned_transducer_stateless7/decode.py \
--epoch 30 \
--avg 5 \
--decoding-method greedy_search \
--exp-dir pruned_transducer_stateless7/exp \
--bpe-model data/en/lang_bpe_500/bpe.model \
--max-duration 600
# modified beam search
./pruned_transducer_stateless7/decode.py \
--epoch 30 \
--avg 5 \
--decoding-method modified_beam_search \
--beam-size 4 \
--exp-dir pruned_transducer_stateless7/exp \
--bpe-model data/en/lang_bpe_500/bpe.model \
--max-duration 600
Pretrained model is available at https://huggingface.co/yfyeung/icefall-asr-cv-corpus-13.0-2023-03-09-en-pruned-transducer-stateless7-2023-04-17
The tensorboard log for training is available at https://tensorboard.dev/experiment/j4pJQty6RMOkMJtRySREKw/