Closed HildaNya closed 2 years ago
Hi. There should be some output messages during training, e.g., global step=..., training loss=...
, according to the argument summary_steps
.
So is there any messages about CUDA environment or does your GPU support mixed-precision training? By default, NeurST uses mixed-precision training and you can switch to normal training via --dtype float32
.
Thanks for the tip. I'm still trying to figure out the issue. Another dumb question though: what is the path that training weights are automatically stored? Is it the same path as the model configuration,
/path_to_data/asr_st/asr_benchmark
? Sorry for the trivial question. I'm just trying to double-check all possible factors that could go wrong. Thanks!
Hi. You need to change this path according to your file system.
Thanks for all the help. Turns out, it was running, just EXTREMELY slowly because I was using CPU instead of GPU.
Hi, just one more suggestion. Turn on --dtype float32
option when using CPU because CPU does not support mixed precision computation.
You are a life-saver.
Hi. I'm working on the ASR training step from the Must-C example. After executing
the training process became stuck for days.
Looking at the output, it seems like it got stuck at "Training for 200000 steps...Saving model configurations to directory:.." One of the last lines of output is this
I'm not sure if it's encountered an error or if it's just slow (due to GPU incompatibility). So I'm looking for general ideas. Also, just to double-check, in the middle of training, is there supposed to be output messages on the progress or just completely silent until training is finished?
Thanks!