Skip to content

Actions: mudler/vllm.cpp

Actions

containers

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
containers.yml
519 workflow runs
519 workflow runs

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
feat(LTX25-UPSAMPLER-ARMS): port the upsampler's dims=2 arm, and mirror model.py:47's else containers #2528: Pull request #2582 synchronize by mudler
Queued row/LTX25-UPSAMPLER-ARMS
row/LTX25-UPSAMPLER-ARMS
Queued
feat(LTX25-UPSAMPLER-ARMS): port the upsampler's dims=2 arm, and mirror model.py:47's else containers #2527: Pull request #2582 synchronize by mudler
57s row/LTX25-UPSAMPLER-ARMS
row/LTX25-UPSAMPLER-ARMS
57s
feat(LTX25-AUDIO-RESAMPLE): resample the take instead of refusing it, and the window is hann containers #2526: Pull request #2585 opened by localai-org-maint-bot
Queued row/LTX25-AUDIO-RESAMPLE
row/LTX25-AUDIO-RESAMPLE
Queued
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2525: Pull request #2361 synchronize by ghazni101
Queued ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
Queued
feat(LTX25-UPSAMPLER-ARMS): port the upsampler's dims=2 arm, and mirror model.py:47's else containers #2524: Pull request #2582 opened by localai-org-maint-bot
1h 5m 13s row/LTX25-UPSAMPLER-ARMS
row/LTX25-UPSAMPLER-ARMS
1h 5m 13s
fix(FIX-UNALIGNED-CONSUMERS-2540): read a borrowed 16-bit weight off a byte cursor, so the odd base costs nothing containers #2523: Pull request #2581 opened by localai-org-maint-bot
1h 13m 6s row/FIX-UNALIGNED-CONSUMERS-2540
row/FIX-UNALIGNED-CONSUMERS-2540
1h 13m 6s
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8 containers #2522: Pull request #2363 synchronize by ghazni101
1h 17m 43s ghazni101:row/ROCM-HW-DP4A
ghazni101:row/ROCM-HW-DP4A
1h 17m 43s
feat(BACKEND-ROCM): register kFusedNormRope, so GLM-5.3 stops taking the path a keep-quant MLA checkpoint cannot survive containers #2521: Pull request #2572 synchronize by mudler
1h 26m 23s row/BACKEND-ROCM-FUSED-NORM-ROPE
row/BACKEND-ROCM-FUSED-NORM-ROPE
1h 26m 23s
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8 containers #2520: Pull request #2363 synchronize by ghazni101
1h 7m 26s ghazni101:row/ROCM-HW-DP4A
ghazni101:row/ROCM-HW-DP4A
1h 7m 26s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2519: Pull request #2361 synchronize by ghazni101
2h 36m 16s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
2h 36m 16s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2518: Pull request #2361 synchronize by ghazni101
9m 46s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
9m 46s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2517: Pull request #2361 synchronize by ghazni101
19m 56s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
19m 56s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2516: Pull request #2361 synchronize by ghazni101
5m 13s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
5m 13s
diag(MODEL-MM-QWEN4-EXP): the MoE expert selections DIFFER, and a third of the boundaries are exact bf16 ties (#2552) containers #2515: Pull request #2556 synchronize by mudler
2h 42m 46s row/MOEDIV-q4exp-router-2552
row/MOEDIV-q4exp-router-2552
2h 42m 46s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2514: Pull request #2361 synchronize by ghazni101
23m 46s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
23m 46s
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8 containers #2513: Pull request #2363 synchronize by ghazni101
2h 0m 15s ghazni101:row/ROCM-HW-DP4A
ghazni101:row/ROCM-HW-DP4A
2h 0m 15s
feat(BACKEND-ROCM): register kFusedNormRope, so GLM-5.3 stops taking the path a keep-quant MLA checkpoint cannot survive containers #2512: Pull request #2572 synchronize by mudler
2h 29m 9s row/BACKEND-ROCM-FUSED-NORM-ROPE
row/BACKEND-ROCM-FUSED-NORM-ROPE
2h 29m 9s
MODEL-MM-GLM53-FLASH W9c-3b: the --device cuda SIGSEGV is a KV-page residency defect, not a mixed-residency one containers #2511: Pull request #2483 synchronize by mudler
3h 18m 0s row/MODEL-MM-GLM53-FLASH-DEVSEGV
row/MODEL-MM-GLM53-FLASH-DEVSEGV
3h 18m 0s
feat(BACKEND-ROCM): register kFusedNormRope, so GLM-5.3 stops taking the path a keep-quant MLA checkpoint cannot survive containers #2510: Pull request #2572 opened by localai-org-maint-bot
4m 15s row/BACKEND-ROCM-FUSED-NORM-ROPE
row/BACKEND-ROCM-FUSED-NORM-ROPE
4m 15s
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8 containers #2509: Pull request #2363 synchronize by ghazni101
29m 45s ghazni101:row/ROCM-HW-DP4A
ghazni101:row/ROCM-HW-DP4A
29m 45s
perf(ROCm): replace software Dp4a with hardware v_dot4_i32_iu8 containers #2508: Pull request #2363 synchronize by localai-org-maint-bot
6m 19s ghazni101:row/ROCM-HW-DP4A
ghazni101:row/ROCM-HW-DP4A
6m 19s
perf(BACKEND-ROCM): cooperate across Q8_K activation blocks containers #2507: Pull request #2472 synchronize by localai-org-maint-bot
2h 57m 57s VikashLoomba:row/BACKEND-ROCM-Q8K-COOP
VikashLoomba:row/BACKEND-ROCM-Q8K-COOP
2h 57m 57s
feat(rocm): fp8 KV cache prefill fast path via SharedK kernel containers #2506: Pull request #2361 synchronize by ghazni101
55m 45s ghazni101:row/fp8-prefill-fastpath
ghazni101:row/fp8-prefill-fastpath
55m 45s
ci(ENG-CI-ORACLE-GOLDENS): run the oracle-goldens digests on a lane, not only on a developer box containers #2505: Pull request #2555 synchronize by mudler
2h 53m 1s row/ENG-CI-ORACLE-GOLDENS
row/ENG-CI-ORACLE-GOLDENS
2h 53m 1s
fix(BACKEND-ROCM): ask the device that will RUN a placed expert tower whether it may keep its blocks containers #2504: Pull request #2562 synchronize by mudler
42m 53s row/BACKEND-ROCM-IQ-EXPERT-RESIDENCY
row/BACKEND-ROCM-IQ-EXPERT-RESIDENCY
42m 53s