Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Aug 25, 2023

• Add copilot-codex completions endpoint for drop-in copilot usage

  abetlen

  committed
  5de8009

  Copy full SHA for 5de8009

Commits on Aug 8, 2023

• Add mul_mat_q option

  abetlen

  committed
  d015bdb

  Copy full SHA for d015bdb

Commits on Jul 24, 2023

• Add temporary rms_norm_eps parameter

  abetlen

  committed
  11dd2bf

  Copy full SHA for 11dd2bf

Commits on Jul 20, 2023

• Fix compatibility with older python versions

  abetlen

  committed
  28a1117

  Copy full SHA for 28a1117

• Merge pull request #481 from c0sogi/main

  Show description for 365d9a4
  abetlen

  authored
  365d9a4

  Copy full SHA for 365d9a4

Commits on Jul 18, 2023

• expose RoPE param to server start

  

  shutup

  committed
  5ed8bf1

  Copy full SHA for 5ed8bf1

Commits on Jul 16, 2023

• Added `RouteErrorHandler` for server

  c0sogi

  committed
  1551ba1

  Copy full SHA for 1551ba1

Commits on Jul 14, 2023

• fix: tensor_split should be optional list

  abetlen

  committed
  118b7f6

  Copy full SHA for 118b7f6

• Resolve merge conflicts

  

  Shouyi Wang

  committed
  579f526

  Copy full SHA for 579f526

• bugfix: pydantic v2 fields

  abetlen

  committed
  de4cc5a

  Copy full SHA for de4cc5a

Commits on Jul 9, 2023

• Add tensor split

  

  Shouyi Wang

  committed
  9f21f54

  Copy full SHA for 9f21f54

Commits on Jul 8, 2023

• Upgrade fastapi to 0.100.0 and pydantic v2

  abetlen

  committed
  52753b7

  Copy full SHA for 52753b7

Commits on Jul 7, 2023

• Add setting to control request interruption

  abetlen

  committed
  57d8ec3

  Copy full SHA for 57d8ec3

• Add interruptible streaming requests for llama-cpp-python server. Closes #183

  abetlen

  committed
  4c7cdcc

  Copy full SHA for 4c7cdcc

Commits on Jun 22, 2023

• server: pass seed param from command line to llama

  lexin4ever

  authored
  282698b

  Copy full SHA for 282698b

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 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