Shang Chieh Tseng 22e77e0dde Unload models from VRAM after use to free GPU memory ...

- Add unloadModel() method to LLMJudge class
- CLI calls unloadModel() after judging completes
- Workflows unload gemma3:4b after inference tests
- Uses Ollama API with keep_alive:0 to trigger unload

2025-12-17 16:51:12 +08:00

..

build.yml

Change workflow defaults: judge_mode=dual, judge_model=gemma3:12b

2025-12-17 16:43:38 +08:00

full-pipeline.yml

Unload models from VRAM after use to free GPU memory

2025-12-17 16:51:12 +08:00

inference.yml

Unload models from VRAM after use to free GPU memory

2025-12-17 16:51:12 +08:00

runtime.yml

Change workflow defaults: judge_mode=dual, judge_model=gemma3:12b

2025-12-17 16:43:38 +08:00