Skip to content

unsloth: pin #253 (no CUDA graph cache count cap) in place of #241 #256

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
danielhanchen merged 1 commit into from
Oct 8, 2026
Merged

unsloth: pin #253 (no CUDA graph cache count cap) in place of #241 #256

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion scripts/unsloth/feature-checks.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,7 +99,7 @@
"unslothai#152": "per-run mmap of a context's tensors; a memory-layout change with no observable output difference",
"unslothai#157": "cudaMemcpyDefault in the ggml_cuda_cpy 2D fast path; needs a CUDA host",
"unslothai#158": "ROCm_Host compute buffer type on HIP integrated GPUs; needs a ROCm host",
"unslothai#241": "CUDA graph cache key, macOS row prefetch and the F16 glm5-next indexer cache under a quantized -ctk (the NextN draft graph is upstream since ggml-org#29928); the first needs a CUDA host, the second has no observable output difference, the last needs a real GLM-5-Next model",
"unslothai#253": "carries #241 (CUDA graph cache key, macOS row prefetch and the F16 glm5-next indexer cache under a quantized -ctk (the NextN draft graph is upstream since ggml-org#29928)) and drops the CUDA graph cache count cap so tensor split keeps replaying its 2 * n_layers + 1 graphs per device; the first needs a CUDA host, the second has no observable output difference, the last needs a real GLM-5-Next model; the cap removal also needs a multi-GPU CUDA host",
"unslothai#243": "glm5-next compute buffer size; a memory-layout change with no observable output difference, and the size only shows on a CUDA host with a GLM-5-Next model",
"unslothai#251": "--moe-cache-mib auto sizes the ggml-org#29887 MoE expert cache from the fit; it only acts when a MoE model spills on a single GPU, the flag is off by default, and no runner in the pipeline has a GPU"
}
Expand Down
2 changes: 1 addition & 1 deletion scripts/unsloth/pr-set.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,7 +30,7 @@
"https://github.com/unslothai/llama.cpp/pull/176/commits/e2519ebc09468b477847a7529834a8da998ed2da",
"https://github.com/unslothai/llama.cpp/pull/239/commits/bec216444552a20bc39bf2156fa5396154970608",
"https://github.com/unslothai/llama.cpp/pull/240/commits/36175a8325e24303704cfbab1b7d8a0bd30500a5",
"https://github.com/unslothai/llama.cpp/pull/241/commits/9dd797259031034c547c0d0dc426776b003731cc",
"https://github.com/unslothai/llama.cpp/pull/253/commits/20dd977eae5e1a3b0633bba85f03d8b09692379f",
"https://github.com/unslothai/llama.cpp/pull/243/commits/f67cec2bd6806d5294e2abc5d62a16405dd3eb74",
"https://github.com/unslothai/llama.cpp/pull/251/commits/13f4c714dff3bf7aa1a83be7490102b7706244ee"
]
Expand Down