Richard Mao comments

Repositories
Issues
Comments

Results 3 comments of


                                            Richard Mao

Can we initialize the solution of qpth with some approximated solution?

I raise this question since sometimes qpth can't provide accurate solution and will output warning.

Negative reward when serving ArmoRM-Llama3-8B-v0.1

Perhaps I got the reason... I built the model from AutoModel.from_pretrained rather than AutoModelForSequenceClassification.from_pretrained, and when I tried the example you gave, the model will output something like this: It's...

Negative reward when serving ArmoRM-Llama3-8B-v0.1

Thanks a lot! By the way, could you please provide an example that inferences for a batch of input and takes attention mask as an input? The example that you...