Arnav Chavan comments

Results 10 comments of


                                            Arnav Chavan

OpenAPI Key Request Issue in Danswer when Changing Environment to Ollama

I am also facing a similar issue!

About accuracy of iMAML implementation

Did you complete the implementation and were you able to reproduce the original results?

Evaluation of lit-llama models

This is great, this opens up lit-llama based finetuned models to be published on the hugging face leaderboard!

Evaluation of lit-llama models

Yes, but it should be pretty simpler I guess, I just have one concern - the time taken for evaluation seems to be higher than other huggingface models. Can you...

Evaluation of lit-llama models

Any headsup on how I can implement higher batch sizes? I will try to implement it over the weekend! Update - It works directly just by hardcoding any batch size...

Phi3 has a nearly constant DPO loss of 0.69xx

I tried both the variants - 1. Having pad token = eos token 2. Using L218 in run_dpo.py to add an explicit [PAD] token. In both the cases the issue...

Phi3 has a nearly constant DPO loss of 0.69xx

I am using transformers 4.42.4 and trl 0.9.6. I tried using the DPOTrainer from TRL and still faced the same issue. @WeiXiongUST if you get time can you please try...

Details of LoRA of pruned models.

Thanks alot, any idea on the GPU hours you need for this run?

Details of LoRA of pruned models.

Thanks for your reply! Did you perform a zero-shot evaluation on the LoRA fine-tuned+pruned models? This is practically very important as models undergo fine-tuning before deployment.

Details of LoRA of pruned models.

Please let me know if you have the checkpoint someplace, would save some time for me. I am interested in seeing the role of peft in sparse models.