Skip to content

v1.4.1

Latest

Choose a tag to compare

@Honguan Honguan released this 16 Oct 16:36
· 1 commit to main since this release
14a0e44

依第一次真的 GPU 實測(RTX 4060 Laptop 8 GB、Windows 11,10 項中 9 項通過)做的修正。

修正

  • eval gpu:VRAM 峰值改成「最高減最低讀數」;量不到 VRAM 變化時只提出警告,不再判為失敗(實際裝置以引擎記錄為準)。這次唯一的失敗就是這個誤判。
  • eval gpu 的 Whisper 檢查改用真的英文語音(LibriSpeech 片段,需要 pyarrow),要求 WER ≤ 0.5;embedding 檢查改用一對一的問答,要求 recall@1 ≥ 0.8。
  • llm-finetune 0.1.2、llm-preference 0.1.1:最低 VRAM 宣告從 8 GB 改成 6 GB(0.5B QLoRA 實測 1.4–1.6 GB),6 GB 的顯示卡也能用;gradient checkpointing 改用 use_reentrant=False,不再出現警告。
  • text-classification 0.1.1:記錄 loss 改用 loss.item()。

實測結果:https://honguan.github.io/autoforge/GPU/

升級並重新驗證

pip install -U autoforge-trainer pyarrow
autoforge eval gpu

完整變更紀錄:https://github.com/Honguan/autoforge/blob/main/CHANGELOG.md