Hatsunama/Qwen3-4B-OBLITERATED-toolcall-lora

Short model card with 15/16 holdout and BFCL scores

Add holdout and BFCL eval writeup after recovery

Document no-call recovery and 15/16 holdout

Upload holdout + BFCL eval after no-call recovery

Store no-call recovery trainer

No-call recovery SFT after when-to-call patch

Upload when-to-call patch eval (holdout + BFCL subset)

When-to-call patch on tool-call LoRA (continue r=16)

Upload README.md with huggingface_hub

Upload README.md with huggingface_hub

GPU tool-calling LoRA on repaired Qwen3-4B-OBLITERATED (1 epoch, r=16)

Upload README.md with huggingface_hub

Upload README.md with huggingface_hub

Point adapter_config at Hatsunama repaired base

Document embedding repair base and short tool-call LoRA

CPU fallback tool-calling LoRA (r=8, 50 steps)

initial commit