jgcb00

To notice : This work fix the issue on v1.8.11, but doesn't seem to fix the issue on 1.8.19, I didn't check why

Same for me ! The first send works well but as soon as the following one, which is the next line of code, there is the same issue, no connection...

"When training a 65B-parameter model, our code processes around 380 tokens/sec/GPU on 2048 A100 GPU with 80GB of RAM. This means that training over our dataset containing 1.4T tokens takes...

Hi, same here, I am stuck because of that, could you please merge it and publish a new version ? Regards !

To anyone interested, I took the liberty of creating a new python lib : https://github.com/jgcb00/fasttext-langdetect-wheel Will delete it when the merge goes through

To solve the issue I did : ```bash git clone https://github.com/adamerose/PandasGUI.git cd PandasGUI nano pandasgui/constants.py ``` Commented the 3 last line : ```python #SHORTCUT_PATH = os.path.join(os.getenv('APPDATA'), 'Microsoft/Windows/Start Menu/Programs/PandasGUI.lnk', ) #PY_INTERPRETTER_PATH...

Hi, We release an easy to install, easy to use, laser package. it supports Laser2 - Laser3, you can choose your hardware, and it support multiprocessing out of the box....

Hi, I wanted to try your implementation but when I try to convert the model, i run into an issue : ``` text-generation-server quantize tiiuae/falcon-40b-instruct falcon-40b-instruct-gptq --trust-remote-code ``` error :...

So it doesn't work after fixing it Now I have : ``` Token indices sequence length is longer than the specified maximum sequence length for this model (2782307 > 2048)....

But I don't get it, we cannot convert the model but then it will easily run on one signle A100 ? I will try with the device cpu see if...