Xiaojiang Liu
Xiaojiang Liu
Failed to sign up at CSDN. Could you provide a plain download address on Github? Thanks
感谢 @ChenYilong 提出非常有价值的建议，下个版本会做出优化。
Thanks for report. Would you provide any detail information or demo project about this?
You are free to customize what to delete. Making it not to delete library dir may solve your problem.
Guess SwiftyUserDefaults rely on some folders under Library Directory. You need to find out what folder it is, and avoid to delete these folder.
maybe hold option-A would output å leads to this problem?
I has same problem with 65B 4-bit 128-group model inferenced with CUDA: Quantization: `CUDA_VISIBLE_DEVICES=0,1 python llama.py decapoda-research/llama-65b-hf c4 --wbits 4 --true-sequential --act-order --groupsize 128 --save llama65b-4bit-128g.pt` Inference: `CUDA_VISIBLE_DEVICES=0,1 python -i...
> I converted a model to 8-bit using AutoGPTQ and now it works both here and in what converted it.. so the quantization code is broken and not inference. could...
> I dont see `freeze_vision_encoder` parameter for pi0/pi05. Are are training the entire VLA (FFT) rather than just the action expert ? I think so. see learnable params (4B) in...
Hi @imstevenpmwork thanks for your great work on async inf. It helps me a lot and I would like to contribute. Could you check this PR when you have time?