dianxin556

Results 10 comments of dianxin556

编译cuda出错

遇到同样问题，期待能解决～

the range of loss when training

why my training loss is so big ,that is begined with about 2000000...

problem in finetune stage

hi, i use your code to train my own model, but in funtune stage i set train_bn=False, the training loss began to increase rapidly and then decrese very slowly. and...

How to use my own midi data as input?

Hi, > Hi dianxin556, > > You may consult this README (from my colleague's repo) for the conversion of MIDI files into REMI event format: https://github.com/YatingMusic/compound-word-transformer/blob/main/dataset/Dataset.md > > The `corpus2events.py`...

assert len(set(num_channels_list)) == 1

I also meet the same problem , so how to solve it?

DML EP cannot load some onnx files.

oh， I meet the same problem! 2022-09-05 14:46:03.2726621 [E:onnxruntime:, inference_session.cc:1530 onnxruntime::InferenceSession::Initialize::::operator ()] Exception during initialization:

make错误

hello,这个问题解决了吗？我也遇到了这个问题

import error

I meet the same error in MacOS. And i have tried all the methods upstairs。

loc loss nan

when i get this error, i print the values of loss_t, loc_t and loc_p. and i found the min(loc_t) in smooth_l1_loss is -inf, this result in inf in loss_t, but...

Loss balancing parameters and normalization

I have the same questions,