Commits

Branch selector

 main

User selector

All users

All time

Commit History

Commits on May 7, 2026

• Update build-and-release.yaml

  quimodotcom

  authored
  bd482ec

  Copy full SHA for bd482ec

Commits on May 4, 2026

• fix(_internals): use n_tokens0 offset when enabling last-token logits in add_sequence (#2205)

  Show description for 90e8df9
  Anai-Guo

  authored
  90e8df9

  Copy full SHA for 90e8df9

Commits on May 2, 2026

• fix(ci): skip unsupported Windows CUDA versions (#2204)

  abetlen

  authored
  14d7846

  Copy full SHA for 14d7846

• fix(ci): install CUDA CCCL headers for wheel builds (#2203)

  abetlen

  authored
  bc6ff9f

  Copy full SHA for bc6ff9f

• fix(ci): pass CUDA compiler arg for Windows detection (#2202)

  abetlen

  authored
  04a3638

  Copy full SHA for 04a3638

• fix(ci): pass CUDA unsupported compiler flag during detection (#2201)

  abetlen

  authored
  2bfd80c

  Copy full SHA for 2bfd80c

• chore: bump version to 0.3.22 (#2200)

  abetlen

  authored
  9cf0ce7

  Copy full SHA for 9cf0ce7

• feat(ci): re-enable Windows CUDA wheels (#2198)

  Show description for d2113a1
  abetlen

  authored
  d2113a1

  Copy full SHA for d2113a1

• feat: Update llama.cpp to ggerganov/llama.cpp@63d93d173 (#2197)

  abetlen

  authored
  587d94a

  Copy full SHA for 587d94a

Commits on Apr 27, 2026

• fix(docs): update mkdocstrings inventories config (#2195)

  abetlen

  authored
  c6dc905

  Copy full SHA for c6dc905

• fix(ci): Scope CPU release wheel selectors by OS (#2194)

  abetlen

  authored
  d2bcbac

  Copy full SHA for d2bcbac

• fix(ci): Repair py3 CPU release wheels (#2193)

  abetlen

  authored
  195cc59

  Copy full SHA for 195cc59

• chore: bump version to 0.3.21 (#2192)

  abetlen

  authored
  c8075d1

  Copy full SHA for c8075d1

• fix(ci): Build one arm64 py3 release wheel (#2191)

  Show description for 511b3f4
  abetlen

  authored
  511b3f4

  Copy full SHA for 511b3f4

• feat: Update llama.cpp to ggerganov/llama.cpp@f53577432 (#2189)

  Show description for d87bf08
  abetlen

  authored
  d87bf08

  Copy full SHA for d87bf08

Commits on Apr 13, 2026

• feat: Update llama.cpp to ggerganov/llama.cpp@227ed28e1 (#2182)

  abetlen

  authored
  1b1a320

  Copy full SHA for 1b1a320

Commits on Apr 8, 2026

• feat: Update llama.cpp to ggerganov/llama.cpp@3bd9aa1f9 (#2176)

  Show description for 1bcc5bc
  abetlen

  authored
  1bcc5bc

  Copy full SHA for 1bcc5bc

Commits on Apr 3, 2026

• chore: bump version to 0.3.20 (#2171)

  abetlen

  authored
  02d6bee

  Copy full SHA for 02d6bee

• fix(misc): replace deprecated llama.cpp references (#2170)

  Show description for 08e088c
  abetlen

  authored
  08e088c

  Copy full SHA for 08e088c

• feat: Update llama.cpp to ggerganov/llama.cpp@f49e9178767d557a522618b16ce8694f9ddac628 (#2169)

  abetlen

  authored
  100b275

  Copy full SHA for 100b275

Commits on Mar 30, 2026

• feat(server): add model-load chat_template_kwargs (#2168)

  abetlen

  authored
  7257ba9

  Copy full SHA for 7257ba9

Commits on Mar 29, 2026

• ci: publish release wheels as py3-none (#2166)

  Show description for 7613aca
  abetlen

  authored
  7613aca

  Copy full SHA for 7613aca

• fix(ci): publish distinct manylinux and musllinux cpu wheels (#2165)

  Show description for fcd932a
  abetlen

  authored
  fcd932a

  Copy full SHA for fcd932a

Commits on Mar 25, 2026

• Bump version to 0.3.19 (#2162)

  abetlen

  authored
  f54421b

  Copy full SHA for f54421b

• feat: Update llama.cpp to ggerganov/llama.cpp@c0159f9c1f874da15e94f371d136f5920b4b5335 (#2161)

  Show description for c670222
  abetlen

  authored
  c670222

  Copy full SHA for c670222

• fix: handle embedding models without KV memory (#2160)

  Show description for ac59e5a
  abetlen

  authored
  ac59e5a

  Copy full SHA for ac59e5a

• fix(ci): reduce CUDA binary wheel size only including cubins for current arches and one PTX target for forward compatibility (#2158)

  Show description for 5f9c231
  abetlen

  authored
  5f9c231

  Copy full SHA for 5f9c231

Commits on Mar 24, 2026

• chore: bump version (#2157)

  abetlen

  authored
  d6f46a5

  Copy full SHA for d6f46a5

• feat: expose attention_type parameter in Llama.__init__ (#2143)

  Show description for 7b38c31
  

  3 peopleauthored
  7b38c31

  Copy full SHA for 7b38c31

• fix(ci): docker build workflow (#2156)

  Show description for ccc6bc0
  abetlen

  authored
  ccc6bc0

  Copy full SHA for ccc6bc0

• fix(ci): cuda wheel workflow (#2155)

  Show description for 909ebf1
  abetlen

  authored
  909ebf1

  Copy full SHA for 909ebf1

• fix(ci): release wheel workflow (#2154)

  Show description for f0391c5
  abetlen

  authored
  f0391c5

  Copy full SHA for f0391c5

Commits on Mar 23, 2026

• chore: Bump version (#2153)

  abetlen

  authored
  a6b1807

  Copy full SHA for a6b1807

• fix: Qwen 3.5 support (#2152)

  Show description for 11e7a55
  abetlen

  authored
  11e7a55

  Copy full SHA for 11e7a55

• ci: add riscv64 wheel builds to release workflow (#2139)

  Show description for e1f8ac0
  

  gounthar

  and

  abetlen

  authored
  e1f8ac0

  Copy full SHA for e1f8ac0