Bobak Hashemi

compile error when installing versions 0.9.2 or 0.8.1.rc2

8

I ran into [this issue](https://github.com/huggingface/transformers/issues/15065) when installing `transformers[flax]` which seems to depend on older versions of `tokenizers`. I do have the latest `tokenizers` available installed via pip, but attempting to...

[WIP] Train Rankgen ranking model for RLHF

2

This PR is just a WIP to show the approach. As of now, I am training on the webgpt dataset, there are three main items I still need to accomplish...

Bth5032/78 blackcat trainer

3

#78 As per discussion with @theblackcat102 I built the rankgen trainer on top of their framework ([wandb](https://wandb.ai/bobakhashemi/reward-model/runs/2sgitozt?workspace=user-bobakhashemi)). The model seems to be training now in fp32. Apparently t5 has some...

Get model evaluation working on the reward model trainer

8

Based on #313 , we are having issues with model evaluation in the reward model trainer (code in model/ranking). It seems that the evaluation results are not being computed and...