COMMITS
May 11, 2026
M
Update README.md (#280)
Miles committed
January 24, 2026
M
Merge pull request #272 from Tiiny-AI/Tiiny
Miles committed
S
Launch Tiiny
syx committed
M
Update README.md (#271)
Miles committed
July 28, 2025
W
Update README.md (#262)
wwwisman committed
July 27, 2025
W
Update README.md (#261)
wwwisman committed
W
Update README.md (#260)
wwwisman committed
W
Support SmallThinker (#259)
wwwisman committed
February 19, 2025
Z
Update README.md (#239)
zeyu committed
A
Release code for OPT models (#245)
AliceRayLu committed
September 6, 2024
July 1, 2024
M
add convert-hf-to-powerinfer-gguf.py to CMakeLists.txt (#205)
Matthew Croughan committed
June 27, 2024
J
Update README (#204)
Jeremy Song committed
June 11, 2024
J
minor in README(#193)
Jeremy Song committed
J
add news (#192)
Jeremy Song committed
May 20, 2024
L
Support AMD ROCm(#187)
Lavent Lee committed
April 7, 2024
H
fix: offload ffn norm weights (#179)
Holden X committed
April 2, 2024
H
Remove axpy dense op (#177)
Holden X committed
H
hotfix: failed to build ffn with LLM_GATE_SEQ (#178)
Holden X committed
March 28, 2024
H
Support Bamboo LM sparse and dense inference (#171)
Holden X committed
March 27, 2024
H
Optimize `mul_mat_sparse` for INT4 quantized weights (#174)
Holden X committed
B
Use `mul_mat_transpose` at axpy op for large batch (#168)
Begunner committed
March 17, 2024
H
Full-GPU computational graph and CUDA refactoring (#153)
Holden X committed
March 14, 2024
R
Add prosparse to README (#149)
Raincleared_Song committed
March 7, 2024
H
fix: unknown host compiler flags passed to nvcc (#161)
Holden X committed
February 20, 2024
H
Fix compiling issue under git worktrees (#146)
Holden X committed
February 18, 2024
H
Fix CMake requirement in README (#145)
Holden X committed
February 1, 2024
M
fix(ggml): add macro definitions to allow compiling to rocm (#139)
Mack Wang committed
Y
Fix a bug of memory leak on Windows platform (#138)
Yifan Tan committed
January 30, 2024
H
Docs: add detailed instruction on downloading HF models (#136)
Holden X committed