Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Jun 15, 2023

• Add low_vram to server settings

  abetlen

  committed
  1e20be6

  Copy full SHA for 1e20be6

• Format server options

  abetlen

  committed
  f7c5cfa

  Copy full SHA for f7c5cfa

• Merge branch 'main' of github.com:abetlen/llama_cpp_python into main

  abetlen

  committed
  9c41a3e

  Copy full SHA for 9c41a3e

• Merge pull request #351 from player1537-forks/th/add-logits-bias-parameter

  Show description for f568bae
  abetlen

  authored
  f568bae

  Copy full SHA for f568bae

• Add additional verbose logs for cache

  abetlen

  committed
  f27393a

  Copy full SHA for f27393a

Commits on Jun 11, 2023

• correction to add back environment variable support <3 docker

  Orfeous

  committed
  3129a0e

  Copy full SHA for 3129a0e

Commits on Jun 10, 2023

• fixes abetlen/llama-cpp-python #358

  Orfeous

  committed
  3ea3193

  Copy full SHA for 3ea3193

Commits on Jun 9, 2023

• Add support for logit_bias and logit_bias_type parameters

  player1537

  committed
  eb7645b

  Copy full SHA for eb7645b

Commits on Jun 8, 2023

• Fix cache implementation breaking changes

  abetlen

  committed
  0c42168

  Copy full SHA for 0c42168

Commits on Jun 6, 2023

• Added mirostat support for completions, chat completions API

  spirilis

  committed
  9b1c9e9

  Copy full SHA for 9b1c9e9

Commits on May 27, 2023

• Use async routes

  abetlen

  committed
  80066f0

  Copy full SHA for 80066f0

• Import unnused import

  abetlen

  committed
  c2b59a5

  Copy full SHA for c2b59a5

Commits on May 19, 2023

• feat: make embedding support list of string as input

  Show description for e783f1c
  chabotsi

  committed
  e783f1c

  Copy full SHA for e783f1c

• Bugfix: Stop sequences can be strings

  abetlen

  committed
  a8cd169

  Copy full SHA for a8cd169

• Use server sent events function for streaming completion

  abetlen

  committed
  dc39cc0

  Copy full SHA for dc39cc0

Commits on May 16, 2023

• Add model_alias option to override model_path in completions. Closes #39

  abetlen

  committed
  a335292

  Copy full SHA for a335292

Commits on May 14, 2023

• Update llama.cpp

  abetlen

  committed
  cdf5976

  Copy full SHA for cdf5976

Commits on May 12, 2023

• Only support generating one prompt at a time.

  abetlen

  committed
  8740ddc

  Copy full SHA for 8740ddc

• Revert "llama_cpp server: prompt is a string". Closes #187

  Show description for 8895b90
  abetlen

  committed
  8895b90

  Copy full SHA for 8895b90

Commits on May 9, 2023

• llama_cpp server: document presence_penalty and frequency_penalty, mark as supported

  Stonelinks

  committed
  02e8a01

  Copy full SHA for 02e8a01

Commits on May 8, 2023

• Fix: default repeat_penalty

  abetlen

  committed
  82d138f

  Copy full SHA for 82d138f

• Bugfix: not falling back to environment variables when default is value is set.

  abetlen

  committed
  29f094b

  Copy full SHA for 29f094b

• Show default value when --help is called

  abetlen

  committed
  0d6c600

  Copy full SHA for 0d6c600

• Use environment variable if parsed cli arg is None

  abetlen

  committed
  022e9eb

  Copy full SHA for 022e9eb

• Set repeat_penalty to 0 by default

  abetlen

  committed
  0d751a6

  Copy full SHA for 0d751a6

• Add openai frequency and presence penalty parameters. Closes #169

  abetlen

  committed
  65d9cc0

  Copy full SHA for 65d9cc0

• Bugfix for models endpoint

  abetlen

  committed
  a0b61ea

  Copy full SHA for a0b61ea

Commits on May 7, 2023

• Added cache size to settins object.

  abetlen

  committed
  14da46f

  Copy full SHA for 14da46f

• Add verbose flag to server

  abetlen

  committed
  627811e

  Copy full SHA for 627811e

• Fix mlock_supported and mmap_supported return type

  abetlen

  committed
  3fbda71

  Copy full SHA for 3fbda71

• Update cpu_count

  abetlen

  committed
  5a3413e

  Copy full SHA for 5a3413e

• Update settings fields and defaults

  abetlen

  committed
  1a00e45

  Copy full SHA for 1a00e45

• Revert "llama_cpp server: delete some ignored / unused parameters"

  Show description for 8675397
  abetlen

  committed
  8675397

  Copy full SHA for 8675397

• Revert "llama_cpp server: mark model as required"

  Show description for c382d8f
  abetlen

  committed
  c382d8f

  Copy full SHA for c382d8f

• Merge branch 'main' of github.com:abetlen/llama_cpp_python into better-server-params-and-fields

  abetlen

  committed
  d8fddcc

  Copy full SHA for d8fddcc