-
问题:训练时loss总是为0;
原因:经过排查,发现计算 cost = criterion(preds, text, preds_size, length) 时,如果preds在gpu上,则loss为0,preds 在 cpu上,则loss正常。
我的conda运行环境:pytorch-0.4.1、cudatoolkit-9.0、cudnn-7.1.2、warp-ctc为https://git…
-
![Screenshot 2019-06-17 at 4 44 56 PM](https://user-images.githubusercontent.com/49031948/59600617-57872880-911f-11e9-8538-988770b61b18.png)
I am getting the above warning while running "make"
Aft…
-
"N:\AI\GPT-SoVITS\runtime\python.exe" tools/asr/fasterwhisper_asr.py -i "N:\AI\GPT-SoVITS\output\slicer_opt" -o "output/asr_opt" -s medium.en -l en -p float16
If you want use h5py dataset, please pip…
-
from ._warp_ctc import lib
ImportError: /usr/local/lib/python2.7/dist-packages/warpctc_pytorch/_warp_ctc.so: undefined symbol: state
-
IDE:AURIX DS
compiler:CCTC ( tasking company)
```
ctc W562: ["..\littlefs-2.2.1\lfs_util.h" 141/26] unary minus applied to unsigned value
ctc W560: ["../littlefs-2.2.1/lfs.c" 361/24] possible …
-
Hi,
This is very interesting project.
I have a question on the CTC beam search (with charlm) you're currently implementing. Are you planning to use RNN based character language model or n-gram b…
-
https://fhirbase.aidbox.app/getting-started#creating-and-initializing-database
fhirbase --host localhost -p 5432 -d fhirbase -U postgres -W postgres --fhir=3.3.0 init
produces this error:
Error…
-
### Prerequisite
- [X] I have searched [Issues](https://github.com/open-mmlab/mmocr/issues) and [Discussions](https://github.com/open-mmlab/mmocr/discussions) but cannot get the expected help.
- [X] …
-
I ran diarize_parallel on a wav file, and I get the following error:
```
Model was trained with pyannote.audio 0.0.1, yours is 3.1.1. Bad things might happen unless you revert pyannote.audio to 0.…
-
Help, Help!
I add some layers in both model_cz.prototxt and tiny.prototxt, when I train the model by python train.py, one error occur: valueError: cannot reshape array of size 6204 into shape(22,1…