Skip to content

Actions: gaugarg-nv/llama.cpp

Actions

Python check requirements.txt

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
python-check-requirements.yml
23 workflow runs
23 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
Fix trailing space Python check requirements.txt #25: Commit de63db6 pushed by gaugarg-nv
5m 24s tp_fit_support
tp_fit_support
5m 24s
Enable Auto parameter fitting in split-mode tensor Python check requirements.txt #24: Commit f46c497 pushed by gaugarg-nv
5m 53s tp_fit_support
tp_fit_support
5m 53s
Python check requirements.txt Python check requirements.txt #23: by gaugarg-nv
7m 17s b9037
b9037
7m 17s
Write a readme on Multi-GPU usage in llama.cpp Python check requirements.txt #22: Commit f3f02e1 pushed by gaugarg-nv
8m 15s multi-gpu-read-me
multi-gpu-read-me
8m 15s
Fix recurrent state serialization for partial reads and writes Python check requirements.txt #21: Commit 92f0041 pushed by gaugarg-nv
4m 50s fix_cache_serialization
fix_cache_serialization
4m 50s
Fix delayed AllReduce on Gemma-4 MoE Python check requirements.txt #20: Commit 4ce8fde pushed by gaugarg-nv
5m 26s gemma4_perf
gemma4_perf
5m 26s
cache subgraph splits when cgraph is unchanged Python check requirements.txt #19: Commit 4bfe714 pushed by gaugarg-nv
5m 15s tp_cpu_overhead
tp_cpu_overhead
5m 15s
Change CMake to use NCCL on Windows Python check requirements.txt #18: Commit 68fb0ba pushed by gaugarg-nv
5m 59s nccl_windows
nccl_windows
5m 59s
Change CMake to use NCCL on Windows Python check requirements.txt #17: Commit a92f245 pushed by gaugarg-nv
4m 46s nccl_windows
nccl_windows
4m 46s
Use the new kernel only for nblocks_stream_k_raw > 4 * ntiles_dst to … Python check requirements.txt #16: Commit 25ef2df pushed by gaugarg-nv
6m 8s fa_opt
fa_opt
6m 8s
Change CMake to use NCCL on Windows Python check requirements.txt #15: Commit 23e1600 pushed by gaugarg-nv
6m 55s nccl_windows
nccl_windows
6m 55s
Reduce the number of stream-k blocks to reduce the overhead of the fl… Python check requirements.txt #14: Commit 38a6935 pushed by gaugarg-nv
4m 50s reduce_stream_k_block
reduce_stream_k_block
4m 50s
Cherry-pick changes from @am17an PR https://github.com/ggml-org/llama… Python check requirements.txt #13: Commit 5d282ad pushed by gaugarg-nv
6m 5s moe_opt
moe_opt
6m 5s
Optimize MOE GEMV kernel for BS > 1. Python check requirements.txt #12: Commit 7cfb67e pushed by gaugarg-nv
5m 2s moe_opt
moe_opt
5m 2s
Increase per-thread work if the K-dimension is small Python check requirements.txt #11: Commit cfbbfb2 pushed by gaugarg-nv
5m 0s small_k_optimization
small_k_optimization
5m 0s
Use a single warp per element instead of a single block per element i… Python check requirements.txt #10: Commit 4f20a44 pushed by gaugarg-nv
6m 26s small_k_optimization
small_k_optimization
6m 26s
Use a single warp per element instead of a single block per element i… Python check requirements.txt #9: Commit 2afd6a3 pushed by gaugarg-nv
4m 41s small_k_optimization
small_k_optimization
4m 41s
Change CMake to use NCCL on Windows Python check requirements.txt #8: Commit 6c90394 pushed by gaugarg-nv
4m 45s nccl_windows
nccl_windows
4m 45s
Improve CUDA graph capture Python check requirements.txt #7: Commit a874a2c pushed by gaugarg-nv
5m 9s cuda_graph_fix
cuda_graph_fix
5m 9s
Fix the seg fault without NCCL Python check requirements.txt #6: Commit 10385e8 pushed by gaugarg-nv
5m 33s minor_fixes
minor_fixes
5m 33s
Fix Issue !19219 Python check requirements.txt #5: Commit 027ef82 pushed by gaugarg-nv
5m 31s revert_scale_queue_env
revert_scale_queue_env
5m 31s
[CUDA] Reduce CPU-side stalls due to the CUDA command buffer being full Python check requirements.txt #4: Commit d3298dc pushed by gaugarg-nv
3m 46s pp_perf_improve
pp_perf_improve
3m 46s
Prefer vector flash decoding kernel for Gemma models Python check requirements.txt #3: Commit f7d07dd pushed by gaugarg-nv
3m 8s gemma_flash_attention
gemma_flash_attention
3m 8s