rungjoo / stable-style-transformer Goto Github PK
View Code? Open in Web Editor NEWStable Style Transformer with Classifier for Text Style Transfer (INLG 2020)
Stable Style Transformer with Classifier for Text Style Transfer (INLG 2020)
안녕하세요 텍스트 스타일 변화를 연령대 별로 해보고 싶어서 aihub에서 연령대 별 명령어음성 데이터(https://aihub.or.kr/aidata/33337) 를 다운받아 학습을 진행해보고 있는데요, 말투가 큰 차이가 나지 않아서 reconstruction_loss가 줄어들지 않는 것인지, 아니면 해당 모델에서 학습을 진행할 때 주의할 점이 있을까요?
koGPT2 토크나이저를 사용해주고, vocab file을 huggingface에서 확인하여 pad_idx, start_idx, end_idx를 수정하였습니다.
전체 학습과정의 20% 정도를 진행한다음 텐서보드로 loss를 확인해보니 11 근처에서 진동하는 형태를 띄고 있습니다.
안녕하세요. 좋은 연구 공개해주셔서 감사합니다.
학습하려고 했는데, 80시간이 넘는 시간이 걸리는 것 같습니다.
혹시 학습된 모델 가중치를 공유해주실 수 있을까요?
감사합니다.
hi,
Thanks for your sharing.
Could you tell me how to use the byte pair encoding, or which tool you are using? I want to use the BPE for another dataset.
great job!
But I cannot run it successfully ><~
python 3.7
torch 1.6.0
File "train.py", line 159, in main gen_cls_loss.backward(retain_graph=True) File "/home/euphoria/anaconda3/lib/python3.7/site-packages/torch/tensor.py", line 185, in backward torch.autograd.backward(self, gradient, retain_graph, create_graph) File "/home/euphoria/anaconda3/lib/python3.7/site-packages/torch/autograd/__init__.py", line 127, in backward allow_unreachable=True) # allow_unreachable flag RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [256, 50259]], which is output 0 of TBackward, is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
Hello,
Please consider releasing you outputs for yelp and amazon so that people can easily use it to compare as a baseline.
Issue with dis_model
in generation_model/yelp/train.py
While training the generator as per this:
dis_model
or the sys.path.insert
refer to in line# 20:21 respectively?gen_model
instead?Kindly help, thanks.
Hello, first of all, thank you very much for sharing the project with us.
I faced the following problem in the train generator step.
Traceback (most recent call last):
File "train.py", line 162, in
main()
File "train.py", line 141, in main
gen_cls_loss.backward() # retain_graph=True
File "/usr/local/lib/python3.8/dist-packages/torch/_tensor.py", line 487, in backward
torch.autograd.backward(
File "/usr/local/lib/python3.8/dist-packages/torch/autograd/init.py", line 197, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [256, 50259]], which is output 0 of AsStridedBackward0, is at version 2; expected version 1 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).
There are people who solved it using clone() but I couldn't. Please your help. Thanks have a good day.
Good job!
Hello, I encountered the following problems while running the model. Can you please take a look at it if you have time. Thank you.
python = 3.5.4
pytorch = 1.2.0
File "/home/fer/anaconda3/envs/py3/lib/python3.5/site-packages/requests/adapters.py", line 512, in send
raise SSLError(e, request=request)
requests.exceptions.SSLError: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:719)
File "train.py", line 13, in
gpt_tokenizer = GPT2Tokenizer.from_pretrained('gpt2')
File "/home/fer/anaconda3/envs/py3/lib/python3.5/site-packages/transformers/tokenization_utils.py", line 282, in from_pretrained
return cls._from_pretrained(*inputs, **kwargs)
File "/home/fer/anaconda3/envs/py3/lib/python3.5/site-packages/transformers/tokenization_utils.py", line 367, in _from_pretrained
raise EnvironmentError(msg)
OSError: Couldn't reach server at '{}' to download vocabulary files.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.