Commits

Branch selector

 llama-cpp-python-low-level

User selector

All users

Datepicker

All time

Commit History

Commits on Jun 1, 2023

• low_level_api_chat_cpp.py: fix default path_prefix arg value to match class default value

  

  Don Mahurin

  committed
  93278f8

  Copy full SHA for 93278f8

Commits on May 31, 2023

• Look for libllama in parent directory

  

  Don Mahurin

  committed
  e5dad2a

  Copy full SHA for e5dad2a

• fix "missing 1 required positional argument: 'min_keep'"

  

  Don Mahurin

  committed
  4ad62c4

  Copy full SHA for 4ad62c4

• Update llama.cpp

  

  abetlen

  authored and

  Don Mahurin

  committed
  60a7c76

  Copy full SHA for 60a7c76

• Fix llama_cpp and Llama type signatures. Closes #221

  

  abetlen

  authored and

  Don Mahurin

  committed
  fda33dd

  Copy full SHA for fda33dd

• Check for CUDA_PATH before adding

  

  abetlen

  authored and

  Don Mahurin

  committed
  601b192

  Copy full SHA for 601b192

• Fixd CUBLAS dll load issue in Windows

  

  aneeshjoy

  authored and

  Don Mahurin

  committed
  66c27f3

  Copy full SHA for 66c27f3

• Update llama.cpp

  

  abetlen

  authored and

  Don Mahurin

  committed
  aae6c03

  Copy full SHA for aae6c03

• Add winmode arg only on windows if python version supports it

  

  abetlen

  authored and

  Don Mahurin

  committed
  a83d117

  Copy full SHA for a83d117

• Update llama.cpp (remove min_keep default value)

  

  abetlen

  authored and

  Don Mahurin

  committed
  7609c73

  Copy full SHA for 7609c73

• Fix mlock_supported and mmap_supported return type

  

  abetlen

  authored and

  Don Mahurin

  committed
  59f80d2

  Copy full SHA for 59f80d2

• Fix obscure Wndows DLL issue. Closes #208

  

  abetlen

  authored and

  Don Mahurin

  committed
  3808a73

  Copy full SHA for 3808a73

• Fix return type

  

  abetlen

  authored and

  Don Mahurin

  committed
  6905884

  Copy full SHA for 6905884

• Fix: runtime type errors

  

  abetlen

  authored and

  Don Mahurin

  committed
  4885e55

  Copy full SHA for 4885e55

• Fix: types

  

  abetlen

  authored and

  Don Mahurin

  committed
  0c2fb05

  Copy full SHA for 0c2fb05

• Fix candidates type

  

  abetlen

  authored and

  Don Mahurin

  committed
  ff31330

  Copy full SHA for ff31330

• Fix llama_cpp types

  

  abetlen

  authored and

  Don Mahurin

  committed
  7862b52

  Copy full SHA for 7862b52

• Add return type annotations for embeddings and logits

  

  abetlen

  authored and

  Don Mahurin

  committed
  f20b34a

  Copy full SHA for f20b34a

• Add types for all low-level api functions

  

  abetlen

  authored and

  Don Mahurin

  committed
  731c712

  Copy full SHA for 731c712

• Allow model to tokenize strings longer than context length and set add_bos. Closes #92

  

  abetlen

  authored and

  Don Mahurin

  committed
  a439fe1

  Copy full SHA for a439fe1

• low_level_api_chat_cpp.py: Fix missing antiprompt output in chat.

  

  Don Mahurin

  committed
  b5531e1

  Copy full SHA for b5531e1

• Fix session loading and saving in low level example chat

  

  SagsMug

  authored and

  Don Mahurin

  committed
  fb79c56

  Copy full SHA for fb79c56

• Fix mirastat requiring c_float

  

  SagsMug

  authored and

  Don Mahurin

  committed
  0bf36a7

  Copy full SHA for 0bf36a7

• Fix lora

  

  SagsMug

  authored and

  Don Mahurin

  committed
  f8ba031

  Copy full SHA for f8ba031

• Wrong logit_bias parsed type

  

  SagsMug

  authored and

  Don Mahurin

  committed
  bbf6848

  Copy full SHA for bbf6848

• Rename postfix to suffix to match upstream

  

  SagsMug

  authored and

  Don Mahurin

  committed
  335cd8d

  Copy full SHA for 335cd8d

• Update low level examples

  

  SagsMug

  authored and

  Don Mahurin

  committed
  32cf013

  Copy full SHA for 32cf013

• Prefer explicit imports

  

  abetlen

  authored and

  Don Mahurin

  committed
  9e79465

  Copy full SHA for 9e79465

• Update llama.cpp (session version)

  

  abetlen

  authored and

  Don Mahurin

  committed
  d15578e

  Copy full SHA for d15578e

• Update sampling api

  

  abetlen

  authored and

  Don Mahurin

  committed
  c26e9bf

  Copy full SHA for c26e9bf

• Fix return types and import comments

  

  abetlen

  authored and

  Don Mahurin

  committed
  78531e5

  Copy full SHA for 78531e5

• Update llama.cpp

  

  abetlen

  authored and

  Don Mahurin

  committed
  d0031ed

  Copy full SHA for d0031ed

• Detect multi-byte responses and wait

  

  Mug

  authored and

  Don Mahurin

  committed
  441d308

  Copy full SHA for 441d308

• Also ignore errors on input prompts

  

  Mug

  authored and

  Don Mahurin

  committed
  36b3494

  Copy full SHA for 36b3494

• Update llama.cpp (llama_load_session_file)

  

  abetlen

  authored and

  Don Mahurin

  committed
  c8e6ac3

  Copy full SHA for c8e6ac3

Pagination

Previous

Next