Dongximing comments

Results 7 comments of


                                            Dongximing

for word embedding in RNN model

thank u for your kindly reply. I think I got it.

Example for LMGradientAttribution is missing.

Hi aobo, if you are busy, could you tell me which model layer need I take as a parameters in the LayerIntegratedGradients ?

Example for LMGradientAttribution is missing.

Thanks, I saw the result, and analysis the code, the final results are computed on log_softmax. and is that means if a contribution in this way, [-10,20,-20](Assuming). the token_1 and...

[QUESTION] which script to use to launch llama 2 fine tuning?

same question please !!!!!!

[Frontend] Add option for LLMEngine to return model hidden states.

please merge it, thanks, really need it !

eval context

also, I extracted the data from the hallucination result(https://huggingface.co/datasets/vectara/leaderboard_results), which consists of the deepseek R1 result and I used the model(vectara/hallucination_evaluation_model) to evaluate the result and find the mean hallucination...

Can not reproduce the results of EAGLE3-DeepSeek-R1-Distill-LLaMA-8B

Thanks for your work! Could you please provide a script when using the Deepseek-R1 model?:)