Skip to content

Activity

Showing most recent first

Vulkan: Add device-specific blacklist for coopmat for the AMD proprie…

bmtwlpushed 597 commits to numathreadscheduling • 70392f1…b56f079 • 
on Jan 5

Vulkan: Add device-specific blacklist for coopmat for the AMD proprie…

bmtwlpushed 1629 commits to numabackend • 8425001…b56f079 • 
on Jan 5

ggml : add AVX512DQ requirement for AVX512 builds (ggml-org#9622)

bmtwlpushed 152 commits to numathreadscheduling • 4db0478…70392f1 • 
on Sep 24, 2024

cuda : fix defrag with quantized KV (ggml-org#9319)

bmtwlcreated numathreadscheduling • 4db0478 • 
on Sep 5, 2024

Merge branch 'ggerganov:master' into numamovepages

bmtwlpushed 3 commits to numamovepages • 169ebe3…95a9a9e • 
on May 7, 2024

Merge branch 'ggerganov:master' into numamovepages

bmtwlpushed 9 commits to numamovepages • cac347c…169ebe3 • 
on May 7, 2024

gguf-split: add --no-tensor-first-split (ggml-org#7072)

bmtwlpushed 257 commits to numabackend • b06c16e…8425001 • 
on May 4, 2024

Deleted branch

bmtwldeleted oldnumaflags • 
on May 4, 2024

Merge branch 'ggerganov:master' into numamovepages

bmtwlpushed 16 commits to numamovepages • 3ed9c2f…cac347c • 
on May 4, 2024

changed memory allocation to split each tensor into equal sized numa-…

bmtwlpushed 3 commits to numamovepages • 2bc1d64…3ed9c2f • 
on May 4, 2024

Add in a tensor init step that uses move_pages() and mbind() to force…

bmtwlpushed 3 commits to numamovepages • f364eb6…2bc1d64 • 
on May 1, 2024

switch to using localizedDescription (ggml-org#7010)

bmtwlcreated numamovepages • f364eb6 • 
on Apr 30, 2024

nix: fix blas support (ggml-org#6281)

bmtwlpushed 2 commits to numabackend • 43139cc…b06c16e • 
on Mar 25, 2024

flake.lock: Update (ggml-org#6266)

bmtwlpushed 8 commits to numabackend • a0e584d…43139cc • 
on Mar 25, 2024

imatrix : fix wname for mul_mat_id ops (ggml-org#6271)

bmtwlpushed 57 commits to numabackend • d8b009a…a0e584d • 
on Mar 24, 2024

Remove undeed header file. (ggml-org#6158)

bmtwlpushed 1 commit to numabackend • d0d5de4…d8b009a • 
on Mar 20, 2024

gguf-split: split and merge gguf per batch of tensors (ggml-org#6135)

bmtwlpushed 17 commits to numabackend • c47cf41…d0d5de4 • 
on Mar 19, 2024

ggml : add AVX512F SIMD (ggml-org#6088)

bmtwlpushed 66 commits to numabackend • 77d1ac7…c47cf41 • 
on Mar 16, 2024

server : print chat template info

bmtwlpushed 21 commits to numabackend • 6cdabe6…77d1ac7 • 
on Mar 9, 2024

llama-bench : add embeddings option (ggml-org#5924)

bmtwlpushed 9 commits to numabackend • 652ca2b…6cdabe6 • 
on Mar 8, 2024

compare-llama-bench.py : remove mul_mat_q (ggml-org#5892)

bmtwlcreated numabackend • 652ca2b • 
on Mar 6, 2024

Deleted branch

bmtwldeleted master • 
on Mar 6, 2024

Merge branch 'ggerganov:master' into master

bmtwlcreated oldnumaflags • ff09606 • 
on Mar 6, 2024

Deleted branch

bmtwldeleted numaflags • 
on Feb 29, 2024

Deleted branch

bmtwldeleted numamirror • 
on Feb 29, 2024

Merge branch 'ggerganov:master' into master

bmtwlpushed 66 commits to master • e3e245c…ff09606 • 
on Feb 28, 2024

examples : do not assume BOS when shifting context (ggml-org#5622)

bmtwlpushed 10 commits to numamirror • 6560bed…89febfe • 
on Feb 21, 2024

Merge branch 'ggerganov:master' into numaflags

bmtwlpushed 11 commits to numaflags • fff7ec0…967e6af • 
on Feb 21, 2024

Merge branch 'ggerganov:master' into master

bmtwlpushed 11 commits to master • de77c7a…e3e245c • 
on Feb 21, 2024

Merge branch 'ggerganov:master' into numaflags

bmtwlpushed 32 commits to numaflags • e944c86…fff7ec0 • 
on Feb 20, 2024