Skip to content

Actions: The-Monk/llama.cpp

Actions

All workflows

Actions

Loading...
Loading

Showing runs from all workflows
54 workflow runs
54 workflow runs

Filter by Workflow

⌥ + click/return to exclude

Filter by Event

⌥ + click/return to exclude

Filter by Status

⌥ + click/return to exclude

Filter by Branch

⌥ + click/return to exclude

Filter by Actor

⌥ + click/return to exclude
Merge branch single-copy (T399): one-copy Q2_0 (NK_Q2_0_W2ONLY) with … Python Type-Check #11: Commit d38e723 pushed by The-Monk
1d 0h 0m 2s rock8-release
rock8-release
1d 0h 0m 2s
Merge branch n1-default (T412): N1 native int8 WMMA prefill default-o… Python Type-Check #10: Commit 74995fe pushed by The-Monk
1d 0h 0m 4s rock8-release
rock8-release
1d 0h 0m 4s
CUDA/HIP: use hardware v_perm_b32 for Q1_0 vec_dot on AMD Python Type-Check #9: Commit d576c95 pushed by The-Monk
1d 0h 0m 1s hip-q1_0-dp4a
hip-q1_0-dp4a
1d 0h 0m 1s
CUDA/HIP: use hardware v_perm_b32 for Q1_0 vec_dot on AMD Python check requirements.txt #6: Commit d576c95 pushed by The-Monk
1d 0h 0m 2s hip-q1_0-dp4a
hip-q1_0-dp4a
1d 0h 0m 2s
CUDA/HIP: use hardware v_perm_b32 for Q1_0 vec_dot on AMD Check Pre-Tokenizer Hashes #6: Commit d576c95 pushed by The-Monk
1d 0h 0m 2s hip-q1_0-dp4a
hip-q1_0-dp4a
1d 0h 0m 2s
CUDA/HIP: use hardware v_perm_b32 for Q1_0 vec_dot on AMD Update Operations Documentation #4: Commit d576c95 pushed by The-Monk
1d 0h 0m 1s hip-q1_0-dp4a
hip-q1_0-dp4a
1d 0h 0m 1s
CUDA/HIP: use hardware v_perm_b32 for Q1_0 vec_dot on AMD Copilot Setup Steps #3: Commit d576c95 pushed by The-Monk
2m 45s hip-q1_0-dp4a
hip-q1_0-dp4a
2m 45s
hip: TQ1_0 load width is a measured null -- correct the record Python Type-Check #8: Commit b4c0fce pushed by The-Monk
1d 0h 0m 2s rock8-release
rock8-release
1d 0h 0m 2s
rock8: MTP graft tool + ternary speculative decoding recipe (Q2_0 +21%) Python Type-Check #7: Commit d417f71 pushed by The-Monk
1d 0h 0m 2s rock8-release
rock8-release
1d 0h 0m 2s
rock8: MTP graft tool + ternary speculative decoding recipe (Q2_0 +21%) .github/workflows/build-cann.yml #23: Commit d417f71 pushed by The-Monk
Failure rock8-release
rock8-release
Failure
rock8: Strix Halo (gfx1151) tester notes .github/workflows/build-cann.yml #22: Commit 7ca56e5 pushed by The-Monk
Failure rock8-release
rock8-release
Failure
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Python Type-Check #6: Commit 7469ae6 pushed by The-Monk
1d 0h 0m 1s rock8-release
rock8-release
1d 0h 0m 1s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Copilot Setup Steps #2: Commit 7469ae6 pushed by The-Monk
1m 51s rock8-v1
rock8-v1
1m 51s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Python Type-Check #5: Commit 7469ae6 pushed by The-Monk
1d 0h 0m 2s rock8-v1
rock8-v1
1d 0h 0m 2s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Update Operations Documentation #3: Commit 7469ae6 pushed by The-Monk
1d 0h 0m 2s rock8-v1
rock8-v1
1d 0h 0m 2s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Check Pre-Tokenizer Hashes #5: Commit 7469ae6 pushed by The-Monk
1d 0h 0m 2s rock8-v1
rock8-v1
1d 0h 0m 2s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… Python check requirements.txt #5: Commit 7469ae6 pushed by The-Monk
1d 0h 0m 2s rock8-v1
rock8-v1
1d 0h 0m 2s
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… .github/workflows/build-cann.yml #21: Commit 7469ae6 pushed by The-Monk
Failure rock8-release
rock8-release
Failure
rock8: all-AMD compatibility -- multi-arch fatbin builds clean gfx103… .github/workflows/build-cann.yml #20: Commit 7469ae6 pushed by The-Monk
Failure rock8-release
rock8-release
Failure
ggml-cuda: SWAR unpack for q1_0/q2_0 on the HIP path (+30% prefill, +… Python check requirements.txt #4: Commit 61dc603 pushed by The-Monk
1d 0h 0m 2s hip-q1q2-swar
hip-q1q2-swar
1d 0h 0m 2s
ggml-cuda: SWAR unpack for q1_0/q2_0 on the HIP path (+30% prefill, +… Python Type-Check #4: Commit 61dc603 pushed by The-Monk
1d 0h 0m 2s hip-q1q2-swar
hip-q1q2-swar
1d 0h 0m 2s
ggml-cuda: SWAR unpack for q1_0/q2_0 on the HIP path (+30% prefill, +… Check Pre-Tokenizer Hashes #4: Commit 61dc603 pushed by The-Monk
1d 0h 0m 1s hip-q1q2-swar
hip-q1q2-swar
1d 0h 0m 1s
ggml-cuda: SWAR unpack for q1_0/q2_0 on the HIP path (+30% prefill, +… .github/workflows/build-cann.yml #19: Commit 61dc603 pushed by The-Monk
Failure hip-q1q2-swar
hip-q1q2-swar
Failure
mmq: SWAR tile unpack for q1_0/q2_0 -- kills the VCC select chains an… .github/workflows/build-cann.yml #18: Commit da53e9c pushed by The-Monk
Failure roc10
roc10
Failure
gated-delta-net: DPP wave reduction -- ds_bpermute off the token recu… .github/workflows/build-cann.yml #17: Commit ab69299 pushed by The-Monk
Failure roc10
roc10
Failure