gapsong

I have another question regarding the paper. In which format are the adapters? Do we use NF4 for the adapters or do we have bfloat16 and just convert it to...

@thistleknot - Offering help with QA-LoRA adapter merge process! Since PEFT doesn't support adapter merging with quantized models yet, I've implemented a custom solution. Successfully replicated the QA-LoRA paper and...

The tutorial from the readme is a bit confusing and it is not working out of the box

Hi @NanoCode012, @ved1beta, @winglian, Thanks for the work on this. Here is the important change. The code is a bit messy...but I hope you get the main idea. https://github.com/ModelCloud/GPTQModel/compare/main...gapsong:GPTQModel:qzero_unquantized I...

> @gapsong > > > I noticed the qzero values are currently being quantized during the save process. > > Could you share where this is happening in peft? @NanoCode012...

@QingshuiL Thank you! It works, but why?

@Qubitium where did you fix it? I have still the same error

@Qubitium I just pulled the newest version from github and installed via pip install -e .

`Traceback (most recent call last): File "/home/nudel/miniconda3/envs/peft_testing/lib/python3.11/runpy.py", line 198, in _run_module_as_main return _run_code(code, main_globals, None, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/nudel/miniconda3/envs/peft_testing/lib/python3.11/runpy.py", line 88, in _run_code exec(code, run_globals) File "/home/nudel/.vscode-server/extensions/ms-python.debugpy-2025.10.0-linux-x64/bundled/libs/debugpy/adapter/../../debugpy/launcher/../../debugpy/__main__.py", line 71, in cli.main()...

@Qubitium I'm encountering a version-specific issue with GPTQModel. When I use an older version of transformers, I can successfully quantize and cache the model. However, a problem arises during loading,...