Fangjun Kuang
d4d4f281ec
Revert "Replace deprecated pytorch methods ( #1814 )" ( #1841 )
...
This reverts commit 3e4da5f78160d3dba3bdf97968bd7ceb8c11631f.
2024-12-18 16:49:57 +08:00
Li Peng
3e4da5f781
Replace deprecated pytorch methods ( #1814 )
...
* Replace deprecated pytorch methods
- torch.cuda.amp.GradScaler(...) => torch.amp.GradScaler("cuda", ...)
- torch.cuda.amp.autocast(...) => torch.amp.autocast("cuda", ...)
* Replace `with autocast(...)` with `with autocast("cuda", ...)`
Co-authored-by: Li Peng <lipeng@unisound.ai>
2024-12-16 10:24:16 +08:00
zr_jin
a1ade8ecb7
fixed failed assertion in the xbmu_ambo31 recipe ( #1816 )
2024-11-29 16:36:02 +08:00
zr_jin
eb132da00d
additional instruction for the grad_scale is too small error ( #1550 )
2024-03-14 11:33:49 +08:00
zr_jin
ef658d691e
fixes for init value of diagnostics.TensorDiagnosticOptions ( #1269 )
...
* fixes for `diagnostics`
Replace `2 ** 22` with `512` as the default value of `diagnostics.TensorDiagnosticOptions`
also black formatted some scripts
* fixed formatting issues
2023-09-24 17:06:47 +08:00
Fangjun Kuang
7b0afbdc16
Remove cur_batch_idx ( #1102 )
2023-05-30 14:49:54 +08:00
Senyan Li
e6a6727012
Add Tibetan Amdo dialect xbmu_amdo31 in egs ( #706 )
...
* add egs/xbmu_amdo31
* fix xbmu_amdo31/ASR/pruned_transducer_stateless5/train.py
* fix xbmu_amdo31/ASR/pruned_transducer_stateless5/asr_datamodule.py
* fix xbmu_amdo31/ASR/prepare.sh
* add RESULTS.md and README.md
* dix pruned_transducer_stateless5 decode.py
* add transducer stateless7
* fix transducer_stateless7
* fix RESULTS.md error
* Add pruned_transducer_stateless7 validation set results
2022-12-03 23:50:49 +08:00
