Skip to content

Actions: CodeLinaro/llama.cpp

Actions

Check Pre-Tokenizer Hashes

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
pre-tokenizer-hashes.yml
23 workflow runs
23 workflow runs

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

workgroup tuning for Hamoa Check Pre-Tokenizer Hashes #23: Commit 74be5d2 pushed by shawngu-quic
52s moe-mxfp4-opt-2
moe-mxfp4-opt-2
52s
server : fix cancel pending task (#16467) Check Pre-Tokenizer Hashes #22: Commit d2ee056 pushed by max-krasnyansky
52s master
master
52s
opencl: only copy to cont for f32 and f16 tensors Check Pre-Tokenizer Hashes #21: Commit 48ba1fb pushed by lhez
58s copy-to-cont-new
copy-to-cont-new
58s
llama : add --no-host to disable host buffers (#16310) Check Pre-Tokenizer Hashes #20: Commit 3df2244 pushed by max-krasnyansky
51s master
master
51s
CI: reenable cdna in rocm docker builds (#16376) Check Pre-Tokenizer Hashes #19: Commit c8dedc9 pushed by shawngu-quic
52s master
master
52s
ci: Properly install rocwmma for hip builds (#16305) Check Pre-Tokenizer Hashes #18: Commit 1fe4e38 pushed by max-krasnyansky
51s master
master
51s
codeowners: add codeowners for opencl backend Check Pre-Tokenizer Hashes #17: Commit 9fd1450 pushed by lhez
54s codeowners
codeowners
54s
ggml : implement set_rows with i32 index (#16159) Check Pre-Tokenizer Hashes #16: Commit 3ecb2f6 pushed by lhez
54s master
master
54s
opencl: fix extra struct alloc Check Pre-Tokenizer Hashes #15: Commit 7f64bfd pushed by lhez
47s q8_0-mv
q8_0-mv
47s
opencl: fix concat crash on win arm64 with Adreno Check Pre-Tokenizer Hashes #14: Commit b84d6ba pushed by lhez
50s concat-win-arm64-fix
concat-win-arm64-fix
50s
opencl: fix concat crash on win arm64 with Adreno Check Pre-Tokenizer Hashes #13: Commit cad982a pushed by lhez
49s concat-win-arm64-fix
concat-win-arm64-fix
49s
opencl: fix extra struct alloc Check Pre-Tokenizer Hashes #12: Commit 066b21d pushed by lhez
49s q8_0-mv
q8_0-mv
49s
ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type (#15797) Check Pre-Tokenizer Hashes #11: Commit 360d653 pushed by lhez
51s master
master
51s
opencl: fix extra struct alloc Check Pre-Tokenizer Hashes #10: Commit 9e07767 pushed by lhez
36s q8_0-mv
q8_0-mv
36s
json : support enum values within allOf (#15830) Check Pre-Tokenizer Hashes #9: Commit 7057faf pushed by lhez
30s master
master
30s
opencl: support ne3 in get_rows Check Pre-Tokenizer Hashes #8: Commit 2e87a69 pushed by lhez
32s get-rows-ne3
get-rows-ne3
32s
opencl: add flattened q8_0 mv_id Check Pre-Tokenizer Hashes #7: Commit a302cc2 pushed by lhez
31s q8_0-mv
q8_0-mv
31s
CUDA: Accelerate MXFP4 table lookup using __byte_perm (#15451) Check Pre-Tokenizer Hashes #6: Commit 74f52f7 pushed by max-krasnyansky
30s master
master
30s
opencl: fix support ops condition for rms_norm Check Pre-Tokenizer Hashes #5: Commit cb52b57 pushed by lhez
37s rms-norm-support-condition
rms-norm-support-condition
37s
opencl: mark argsort unsupported if cols exceed workgroup limit Check Pre-Tokenizer Hashes #4: Commit 1a6e78f pushed by lhez
30s argsort-fix
argsort-fix
30s
HIP: disable sync warp shuffel operators from clr amd_warp_sync_funct… Check Pre-Tokenizer Hashes #3: Commit b049315 pushed by max-krasnyansky
29s master
master
29s
opencl: add reference mul_mv_id for mxfp4 Check Pre-Tokenizer Hashes #2: Commit 3427959 pushed by lhez
30s mul-mv-mxfp4
mul-mv-mxfp4
30s
opencl: add add_id Check Pre-Tokenizer Hashes #1: Commit 02721cc pushed by lhez
28s swiglu-oai-add-id
swiglu-oai-add-id
28s