Shoku Yanagawa comments

Results 7 comments of


                                            Shoku Yanagawa

How to generate or convert vocab.json, merges.txt, and config.json to match huggingface/transformers requirements ?

Thank you for quick reply. There is German GPT-2 repo in the huggingface/transformers. https://huggingface.co/anonymous-german-nlp/german-gpt2 This repo might be your fork.

How to generate or convert vocab.json, merges.txt, and config.json to match huggingface/transformers requirements ?

I know huggingface/tokenizers generates merges.txt and vocab.json. https://github.com/huggingface/tokenizers However I doubt is this solution. Anyway I feel your repo is active. I will push pull request when I need. Best...

Some example build fails on Arm64 environment(Armbian Focal) with Libtorch1.10.1 source build

Solved rebuild 1.10.0.a0 from source tree. It works. tch-rs example build is OK. USE_MKLDNN=OFF flag is also important.

NotImplementedError: Cannot copy out of meta tensor; no data!

I have same trouble. peft v0.4.0.dev0

NotImplementedError: Cannot copy out of meta tensor; no data!

Version: 0.9.2

vicuna-13b v1.1 Strange answer

Thisi problem is related #433, not a Japanese specific. reason: special_token_map.json is empty. Now it's fixed. You can close the issue. Vicuna 13-b-1.1 speakes Japanese fluently even though we expect...

midi.sequencer has no attribute SequencerHardware

I have same kind of problem. python 3.10 I have checked #129 and the related topic "Use relative imports for the sequencer as well".