mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-09-17 04:54:17 +00:00
50 lines
1.2 KiB
Markdown
50 lines
1.2 KiB
Markdown
## Results
|
|
|
|
### GigaSpeech BPE training results (Conformer-CTC)
|
|
|
|
#### 2022-04-06
|
|
|
|
The best WER, as of 2022-04-06, for the gigaspeech is below
|
|
(using HLG decoding + n-gram LM rescoring + attention decoder rescoring):
|
|
|
|
| | Dev | Test |
|
|
|-----|-------|-------|
|
|
| WER | 11.93 | 11.86 |
|
|
|
|
Scale values used in n-gram LM rescoring and attention rescoring for the best WERs are:
|
|
| ngram_lm_scale | attention_scale |
|
|
|----------------|-----------------|
|
|
| 0.3 | 1.5 |
|
|
|
|
|
|
To reproduce the above result, use the following commands for training:
|
|
|
|
```
|
|
cd egs/gigaspeech/ASR/conformer_ctc
|
|
./prepare.sh
|
|
export CUDA_VISIBLE_DEVICES="0,1,2,3,4,5,6,7"
|
|
./conformer_ctc/train.py \
|
|
--max-duration 120 \
|
|
--num-workers 1 \
|
|
--world-size 8 \
|
|
--exp-dir conformer_ctc/exp_500 \
|
|
--lang-dir data/lang_bpe_500
|
|
```
|
|
|
|
and the following command for decoding
|
|
|
|
```
|
|
./conformer_ctc/decode.py \
|
|
--epoch 19 \
|
|
--avg 8 \
|
|
--method attention-decoder \
|
|
--num-paths 1000 \
|
|
--exp-dir conformer_ctc/exp_500 \
|
|
--lang-dir data/lang_bpe_500 \
|
|
--max-duration 20 \
|
|
--num-workers 1
|
|
```
|
|
|
|
The tensorboard log for training is available at
|
|
<https://tensorboard.dev/experiment/rz63cmJXSK2fV9GceJtZXQ/>
|