search

YouAreSpecialToMe / QST

Quantized Side Tuning: Fast and Memory-Efficient Tuning of Quantized Large Language Models
Apache License 2.0
36 stars 2 forks source link

Question about GLUE dataset #5

Closed qtyandhasee closed 3 weeks ago

qtyandhasee commented 1 month ago

Hello! Thank you for your excellent code. In the paper, it is mentioned that fine-tuning was done on the GLUE dataset, but the load_data function in the code does not include GLUE. How can the results on the GLUE dataset be reproduced? Thank you very much!

YouAreSpecialToMe commented 1 month ago

Hi, you can use "QSTLlamaForSequenceClassification" or "QSTOPTForSequenceClassification" and the huggingface GLUE dataset to test QST's performance.

qtyandhasee commented 1 month ago

Hi, you can use "QSTLlamaForSequenceClassification" or "QSTOPTForSequenceClassification" and the huggingface GLUE dataset to test QST's performance.

Should I use the file "qst.py" or https://github.com/huggingface/transformers/blob/main/examples/pytorch/text-classification/run_glue.py

qtyandhasee commented 1 month ago

Do you train on GLUE dataset？

YouAreSpecialToMe commented 4 weeks ago

Hi, I've uploaded the glue_qst.py to conduct experiments on the GLUE benchmark and you can run the glue_qsy.sh to run the script.

qtyandhasee commented 3 weeks ago

Hi, I've uploaded the glue_qst.py to conduct experiments on the GLUE benchmark and you can run the glue_qsy.sh to run the script.

Thank you！