Juyoung Suk
Juyoung Suk
@lewtun Hi, I also want to take part in the translation project! It seems that @Taekyoon is not working on Chapter 3 anymore, so can I start with Chapter 3?
Ok! I'll wait till the end of this week and start working on chapter 3 if no response. @lewtun
@JasmondL I was able to resolve the error with different `CUDA_VISIBLE_DEVICES` permutations: when I changed my command from `CUDA_VISIBLE_DEVICES=7,8,9,10 python -m ...` to `CUDA_VISIBLE_DEVICES=10,9,8,7 python -m ...` it worked without...
@ArmenAg Hello, I'm really interested in this paper! Is threre any plan to release the model and the code for HTLM? Thank you in advance!
Thank you for reporting this issue, @bittersweet1999. I will address it as soon as possible. In the meantime, please use version 0.1.17 as a workaround.
@bittersweet1999 I have modified the parsing function and verified its performance on the pattern mentioned above. However, I will leave this issue open as a thread to track any potential...
I'm really looking forward to this integration! Just out of curiosity, do you think using optimum or torch.compile as a generation backend is possible? @vwxyzjn
Hey @ilyalasy, thank you for the interest! Are you willing to help for this? I have very little background in quantization..