Commit History

Автор SHA1 Съобщение Дата
  Patrick Devine 4cc3be3035 Move envconfig and consolidate env vars (#4608) преди 11 месеца
  Sang Park 4434d7f447 Correct typo in error message (#4535) преди 11 месеца
  Daniel Hiltgen ec231a7923 Remove VRAM convergence check for windows преди 11 месеца
  Patrick Devine 6845988807 Ollama `ps` command for showing currently loaded models (#4327) преди 11 месеца
  Daniel Hiltgen 4142c3ef7c Always use the sorted list of GPUs преди 11 месеца
  Jeffrey Morgan bb6fd02298 Don't clamp ctx size in `PredictServerFit` (#4317) преди 11 месеца
  Daniel Hiltgen 354ad9254e Wait for GPU free memory reporting to converge преди 11 месеца
  Jeffrey Morgan c9f98622b1 Skip scheduling cancelled requests, always reload unloaded runners (#4189) преди 1 година
  Jeffrey Morgan dfa2f32ca0 unload in critical section (#4187) преди 1 година
  Daniel Hiltgen f56aa20014 Centralize server config handling преди 1 година
  Jeffrey Morgan 942c979232 allocate a large enough kv cache for all parallel requests (#4162) преди 1 година
  Daniel Hiltgen 20f6c06569 Make maximum pending request configurable преди 1 година
  Mark Ward 63c763685f log when the waiting for the process to stop to help debug when other tasks execute during this wait. преди 1 година
  Mark Ward f4a73d57a4 fix runner expire during active use. Clearing the expire timer as it is used. Allowing the finish to assign an expire timer so that the runner will expire after no use. преди 1 година
  Daniel Hiltgen d6e3b64582 Fix concurrency for CPU mode преди 1 година
  Daniel Hiltgen 9b5a3c5991 Merge pull request #3914 from dhiltgen/mac_perf преди 1 година
  Jeffrey Morgan 00b0699c75 Reload model if `num_gpu` changes (#3920) преди 1 година
  Daniel Hiltgen b123be5b71 Adjust context size for parallelism преди 1 година
  Bryce Reitano 36a6daccab Restructure loading conditional chain преди 1 година
  Bryce Reitano 284e02bed0 Move ggml loading to when we attempt fitting преди 1 година
  Daniel Hiltgen 34b9db5afc Request and model concurrency преди 1 година