Qwen3.5-2B-JP-v0.2-GGUF

keitokei1994/Qwen3.5-2B-JP-v0.2 を GGUF へ変換したリポジトリです。

Files

• Qwen3.5-2B-JP-v0.2.Q8_0.gguf
• Qwen3.5-2B-JP-v0.2.Q4_K_M.gguf
• Qwen3.5-2B-JP-v0.2.BF16-mmproj.gguf

Source Model

• Base model: Qwen/Qwen3.5-2B
• Fine-tuned model: keitokei1994/Qwen3.5-2B-JP-v0.2
• Dataset: DataPilot/Zero_SFT_Ja_v3.5
• Max sequence length during fine-tuning: 8192

Notes

• Quantizations prepared here are at least Q8_0 and Q4_K_M.
• The mmproj file was emitted by the exporter alongside the GGUF files.