Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Jul 25, 2023

• Change tensor_split from array to pointer

  

  Shouyi Wang

  committed
  426dbfe

  Copy full SHA for 426dbfe

Commits on Jul 24, 2023

• Add temporary rms_norm_eps parameter

  abetlen

  committed
  11dd2bf

  Copy full SHA for 11dd2bf

• Add rms_eps_norm

  abetlen

  committed
  8cd64d4

  Copy full SHA for 8cd64d4

• add support for llama2 70b

  bretello

  committed
  0f09f10

  Copy full SHA for 0f09f10

• Revert "Merge pull request #521 from bretello/main"

  Show description for 401309d
  abetlen

  committed
  401309d

  Copy full SHA for 401309d

• Merge pull request #521 from bretello/main

  Show description for 07f0f3a
  abetlen

  authored
  07f0f3a

  Copy full SHA for 07f0f3a

• Update llama.cpp

  abetlen

  committed
  d8a3ddb

  Copy full SHA for d8a3ddb

• Update llama.cpp

  abetlen

  committed
  985d559

  Copy full SHA for 985d559

• raise exception when `llama_load_model_from_file` fails

  bretello

  committed
  8be7d67

  Copy full SHA for 8be7d67

Commits on Jul 21, 2023

• Update llama.cpp

  abetlen

  committed
  b83728a

  Copy full SHA for b83728a

Commits on Jul 20, 2023

• Update llama.cpp

  abetlen

  committed
  01435da

  Copy full SHA for 01435da

• Fix compatibility with older python versions

  abetlen

  committed
  28a1117

  Copy full SHA for 28a1117

• Revert ctypes argtype change

  abetlen

  committed
  d10ce62

  Copy full SHA for d10ce62

• Merge pull request #481 from c0sogi/main

  Show description for 365d9a4
  abetlen

  authored
  365d9a4

  Copy full SHA for 365d9a4

• Update llama_cpp.py - Fix c_char_p to Array[c_char_p] and c_float to Array[c_float]

  viniciusarruda

  authored
  a855147

  Copy full SHA for a855147

• Now the last token sent when `stream=True`

  ctejada85

  committed
  0756a2d

  Copy full SHA for 0756a2d

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 15, 2023

• Re-order Llama class params

  abetlen

  committed
  8ab098e

  Copy full SHA for 8ab098e

• Fix context_params struct layout

  abetlen

  committed
  e4f9db3

  Copy full SHA for e4f9db3

• Merge branch main into custom_rope

  abetlen

  committed
  f0797a6

  Copy full SHA for f0797a6

Commits on Jul 10, 2023

• Add bindings for custom_rope

  randoentity

  committed
  3f8f276

  Copy full SHA for 3f8f276

Commits on Jul 9, 2023

• bugfix: truncate completion max_tokens to fit context length by default

  abetlen

  committed
  a86bfdf

  Copy full SHA for a86bfdf

• bugfix: pydantic settings missing / changed fields

  abetlen

  committed
  6f70cc4

  Copy full SHA for 6f70cc4

Commits on Jul 8, 2023

• Merge branch 'main' into add_unlimited_max_tokens

  abetlen

  authored
  5d756de

  Copy full SHA for 5d756de

• Merge pull request #453 from wu-qing-157/main

  Show description for b8e0bed
  abetlen

  authored
  b8e0bed

  Copy full SHA for b8e0bed

• bugfix: fix compatibility bug with openai api on last token

  abetlen

  committed
  d6e6aad

  Copy full SHA for d6e6aad

• Format

  abetlen

  committed
  4f2b5d0

  Copy full SHA for 4f2b5d0

• perf: convert pointer to byref

  abetlen

  committed
  34c505e

  Copy full SHA for 34c505e

• Upgrade fastapi to 0.100.0 and pydantic v2

  abetlen

  committed
  52753b7

  Copy full SHA for 52753b7

Commits on Jul 7, 2023

• perf: avoid allocating new buffers during sampling

  abetlen

  committed
  11eae75

  Copy full SHA for 11eae75

• perf: assign to candidates data structure instead

  abetlen

  committed
  a14d8a9

  Copy full SHA for a14d8a9

• fix indexing token_logprobs after sorting

  wu-qing-157

  committed
  9e61661

  Copy full SHA for 9e61661

• 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

Pagination

Previous

Next