Aamir Mirza

When uploading wav file with A-law codec ( sample rate 8000 ) I get a read error unknown file format, I have covert the file to wav 16-bit PCM format...

Based on the paper from Bloomberg GPT, the reference here https://arxiv.org/pdf/2303.17564v1.pdf They mention sequence length of more than 2048 can be used during inference https://paperswithcode.com/method/alibi Can the LLAMA model be...

Have you considered Cerebras-GPT-13B

1

Would this be suitable as the base model https://huggingface.co/cerebras/Cerebras-GPT-13B seems to be free of any licensing limitations

Loading the model on multiple GPUs

3

I have two 4090 24GB, if possible please provide an extra argument to demo.py to either load the model on CPU or 2 or more GPU and another argument to...

Issues with image loading and accelerate

5

FYI , when starting the demo file I get the following message torchvision/io/image.py:13: UserWarning: Failed to load image Python extension: libtorch_cuda_cu.so: cannot open shared object file: the other one is...

Transformers with memory , infinite chat

3

Can this be integrated https://github.com/wbbeyourself/SCM4LLMs the Authors claim that it can be .

Might I suggest that you use deep speed for inference in 16bit mode, that way you can scale to 2 24GB GPU or even CPU here is the link https://www.deepspeed.ai/tutorials/inference-tutorial/

Training Script stablity 3B and 7B

6

`# Developed by Aamir Mirza # create a conda virtual environment python 3.9 # install PyTorch 1.13.1 ( not 2.0) # conda install pytorch==1.13.1 torchvision==0.14.1 torchaudio==0.13.1 pytorch-cuda=11.7 -c pytorch -c...

Pytorch version of the vision model

4

If possible please provide a pytorch version of the vision model.

Feature Request Open AI Clip

1

Please add a feature to infer and train OpenAI clip ( Clip is in Hugging Face) , there is a down stream usage for zero shot image classification.