Commit Verlauf

Autor SHA1 Nachricht Datum
  Patrick Devine 4cc3be3035 Move envconfig and consolidate env vars (#4608) vor 11 Monaten
  Daniel Hiltgen b37b496a12 Wire up load progress vor 1 Jahr
  Jeffrey Morgan 38255d2af1 Use flash attention flag for now (#4580) vor 1 Jahr
  Sam e15307fdf4 feat: add support for flash_attn (#4120) vor 1 Jahr
  Patrick Devine d1692fd3e0 fix the cpu estimatedTotal memory + get the expiry time for loading models (#4461) vor 1 Jahr
  Daniel Hiltgen 853ae490e1 Sanitize the env var debug log vor 1 Jahr
  Patrick Devine 6845988807 Ollama `ps` command for showing currently loaded models (#4327) vor 1 Jahr
  jmorganca 92ca2cca95 Revert "only forward some env vars" vor 1 Jahr
  Daniel Hiltgen c4014e73a2 Fall back to CPU runner with zero layers vor 1 Jahr
  Jeffrey Morgan bb6fd02298 Don't clamp ctx size in `PredictServerFit` (#4317) vor 1 Jahr
  Michael Yang cf442cd57e fix typo vor 1 Jahr
  Michael Yang ce3b212d12 only forward some env vars vor 1 Jahr
  Michael Yang 58876091f7 log clean up vor 1 Jahr
  Daniel Hiltgen d0425f26cf Merge pull request #4294 from dhiltgen/harden_subprocess_reaping vor 1 Jahr
  Bruce MacDonald cfa84b8470 add done_reason to the api (#4235) vor 1 Jahr
  Daniel Hiltgen 84ac7ce139 Refine subprocess reaping vor 1 Jahr
  Daniel Hiltgen 920a4b0794 Merge remote-tracking branch 'upstream/main' into pr3702 vor 1 Jahr
  Daniel Hiltgen ee49844d09 Merge pull request #4153 from dhiltgen/gpu_verbose_response vor 1 Jahr
  Daniel Hiltgen bee2f4a3b0 Record GPU usage information vor 1 Jahr
  Daniel Hiltgen 72700279e2 Detect noexec and report a better error vor 1 Jahr
  Daniel Hiltgen 380378cc80 Use our libraries first vor 1 Jahr
  Jeffrey Morgan ed740a2504 Fix `no slots available` error with concurrent requests (#4160) vor 1 Jahr
  Jeffrey Morgan 1b0e6c9c0e Fix llava models not working after first request (#4164) vor 1 Jahr
  Daniel Hiltgen f56aa20014 Centralize server config handling vor 1 Jahr
  Mark Ward 321d57e1a0 Removing go routine calling .wait from load. vor 1 Jahr
  Mark Ward ba26c7aa00 it will always return an error due to Kill() discarding Wait() errors vor 1 Jahr
  Mark Ward 63c763685f log when the waiting for the process to stop to help debug when other tasks execute during this wait. vor 1 Jahr
  Mark Ward 948114e3e3 fix sched to wait for the runner to terminate to ensure following vram check will be more accurate vor 1 Jahr
  Jeffrey Morgan 7aa08a77ca llm: dont cap context window limit to training context window (#3988) vor 1 Jahr
  Jeffrey Morgan bb31def011 return code `499` when user cancels request while a model is loading (#3955) vor 1 Jahr