Aamir Mirza

> I have never really liked the SDRClassifier algorithm because it requires a lot of training before it can reliably classify (it uses a classical NN). I have noticed that...

thank you it works, however, it would be better is this was done internally. Mostly on account of people like me posting the obvious.

> We have an own implementation of RDSE using lookup table, which is calculated very quickly only one time by initialization. So that we do not need any multithreaded! Yes...

I guess to keep this on your radar for future work, if there is nothing further to be added please close this issue

> Hi, I want to fine-tune the 7b model, am I supposed to download the provided checkpoint and fine-tune it as shown in this repo: https://github.com/EleutherAI/gpt-neox#using-custom-data . Would they be...

Here you go the full training script `# Developed by Aamir Mirza # create a conda virtual environment python 3.9 # install PyTorch 1.13.1 ( not 2.0) # conda install...

I have shard the script works for 3B as well as 7 B On Wed, May 3, 2023 at 2:21 AM Shaila Zaman ***@***.***> wrote: > @aamir-gmail could you please...

Same works for 3B and 7B On Mon, May 22, 2023 at 2:08 PM xuantoan02 ***@***.***> wrote: > Hi, I want to fine-tune the 7b model, am I supposed to...

> Are you able to train 7B using dual RTX3090s? Do you think you could setup a notebook on Colab? Thank you!!!!!! Yes I can, the script provided does it...

Send me the output you are getting on python consiole On Fri, May 19, 2023 at 4:15 AM snirbenyosef ***@***.***> wrote: > i'm using V100 on aws. 2 GPU 8...