download?
there's no link to download the gguf file?
python3 -c "from huggingface_hub import list_repo_files; print('\n'.join(list_repo_files('douyamv/Gemma-4-31B-JANG_4M-CRACK-GGUF')))"
.gitattributes
README.md
the github reposity is empty
Sorry for delay, the file is large. I will upload now.
Sorry for delay, the file is large. I will upload now.
Got it, I’ll wait patiently! I’ve actually conducted similar ablation studies using two-stage LoRA fine-tuning. It worked great on Qwen 2.5, but Qwen 3.5 has been tricky due to its CoT (Chain of Thought) nature. Even though I included tags in the dataset to represent the reasoning process, the results were inconsistent—sometimes the output is cluttered with tags, and other times the thinking process is missing entirely. I’m quite puzzled by this as well, so I’m really looking forward to seeing how others' fine-tuning results turn out.
done for min size model.
please upload other models Thanks.