tlg 38e1523d7e feat: proper VRAM cleanup and admin clear-vram endpoint ...

- gc.collect() + torch.cuda.empty_cache() in unload for reliable VRAM release
- POST /admin/clear-vram endpoint unloads all models and reports GPU memory
- VRAMManager.clear_all() method for programmatic VRAM cleanup

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

2026-04-05 21:03:39 +02:00

..

config

feat: project scaffolding with config files and test fixtures

2026-04-04 07:23:14 +02:00

docs/superpowers

Add llmux implementation plan (30 tasks)

2026-04-03 22:43:37 +02:00

llmux

feat: proper VRAM cleanup and admin clear-vram endpoint

2026-04-05 21:03:39 +02:00

scripts

fix: use LLMUX_SRC env var for Dockerfile path in pod creation script

2026-04-05 13:05:38 +02:00

tests

feat: API routes for models, chat, transcription, speech, and admin

2026-04-05 10:04:45 +02:00

Dockerfile

fix: proper VRAM cleanup on model unload + CUDA alloc config

2026-04-05 17:59:23 +02:00

requirements.txt

fix: Dockerfile uses explicit pip install, skip pre-installed packages

2026-04-05 14:10:07 +02:00