A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 14:45 UTC
main
config check sleep mode support oot platforms (#16562)
Shuqiao Li committed over 1 year ago
[DOC][TPU] Add core idea about avoiding recompilation after warmup (#16614)
Chengji Yao committed over 1 year ago
[Hardware][TPU] Add torchvision to tpu dependency file (#16616)
Siyuan Liu committed over 1 year ago
[Model][VLM] Add Kimi-VL model support (#16387)
courage17340 committed over 1 year ago
s390x: Fix PyArrow build and add CPU test script for Buildkite CI (#16036)
Nishan Acharya committed over 1 year ago
Fix triton install condition on CPU (#16600)
Harry Mellor committed over 1 year ago
[TPU][V1] Fix exponential padding when `max-num-batched-tokens` is not a power of 2 (#16596)
Nicolò Lucchesi committed over 1 year ago
[Bugfix] Multi-modal caches not acting like LRU caches (#16593)
Cyrus Leung committed over 1 year ago
[fix]: Dockerfile.ppc64le fixes for opencv-python and hf-xet (#16048)
Md. Shafi Hussain committed over 1 year ago
[Misc] Update usage with mooncake lib for kv transfer (#16523)
shangmingc committed over 1 year ago
[Misc] refactor examples (#16563)
Reid committed over 1 year ago
[Misc] remove warning if triton>=3.2.0 (#16553)
DefTruth committed over 1 year ago
Improve configs - `SchedulerConfig` (#16533)
Harry Mellor committed over 1 year ago
[Core][V0] Enable regex support with xgrammar (#13228)
Russell Bryant committed over 1 year ago
[Benchmark][Bugfix] Fix SonnetDataset default values in benchmark_throughput.py (#16556)
Jennifer Zhao committed over 1 year ago
Enable PTPC FP8 for CompressedTensorsW8A8Fp8MoEMethod (triton fused_moe) (#16537)
Michael Goin committed over 1 year ago
[V1][Spec Decode] KV cache slots for eagle heads (#16370)
Lily Liu committed over 1 year ago
[Bugfix] Validate logit biases to prevent out of vocab ids crashing engine (#16529)
Ryan McConville committed over 1 year ago
[Perf] Optimize Preparing Inputs for GPU Model Runner (#16484)
SnowCharm committed over 1 year ago
Fix syntaxWarning: invalid escape sequence '\s' (#16532)
Jie Fu (傅杰) committed over 1 year ago