yuanqian_zhao

Quac Dataset

5

When will you support evaluation on the quac dataset? I found the results of the Llama2 paper difficult to reproduce. Especially regarding how to segment the answer for the base...

Does the GPTQ INT8 quantization method ensure that weight values originally at zero remain zero after the quantization process?

This PR mainly involves the following aspects: - QLoRA overall logic: - First, quantize the model parameter files. - Set the int4 field in the model's config to enable QLoRA...

Want to know how to train using AWQ+LoRA? Currently, I haven't found any research supporting the training of the LoRa module with AWQ.

When will some of the Leaderboard evaluation results be made available?

1

as title

I would like to know about chat models. When I use AWQ for calibration, I do not want to compute the loss for the instructions, but only for the responses....

Will MLA that used in DeepSeek-V2 (https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat and https://arxiv.org/abs/2405.04434) be supported by activation smooth method?

https://huggingface.co/openbmb/MiniCPM-2B-sft-bf16 I wonder how well can OmniQuant do on those sota SLMs?

The link to model zoo

1

hi! where is the pre-trained ABQ-LLM model zoo?

Custom datasets support for chat models

1

for chat models, calibration datasets' input_ids&attn_masks should be passed in