Files

T

Nick Hill 276b837dc4 [ModelRunner V2][BugFix] Free all model refs on shutdown (#47483 )

Signed-off-by: Nick Hill <nickhill123@gmail.com>

2026-07-02 19:32:48 -07:00

metrics

[Model Runner V2] Support num NaNs in logits (#30187 )

2025-12-09 10:00:49 -08:00

[ModelRunner V2] Support realtime embeddings (#46762 )

2026-06-26 19:42:27 -07:00

model_states

[ModelRunner V2] Fix Mamba2 crash on non-spec-decode (#47428 )

2026-07-02 07:05:16 -07:00

pool

[Frontend] Re-enable running MaxSim on GPU (#38620 )

2026-04-03 00:03:13 +08:00

sample

[Spec Decode] DSpark (#46995 )

2026-07-01 08:30:24 -07:00

spec_decode

[Bugfix][Model Runner V2][Spec Decode] Fix int32 offset overflow in block verification kernels (#47383 )

2026-07-02 07:32:38 -07:00

__init__.py

GPU Model Runner V2 (#25266 )

2025-11-21 08:20:55 -08:00

async_utils.py

[Platform] Replace torch.cuda.Event with torch.Event (#47140 )

2026-06-30 08:39:59 -07:00

attn_utils.py

[Model] Support Unlimited OCR (#46564 )

2026-06-27 23:09:18 -07:00

block_table.py

[PERF] Fuse multi-group block table staged writes (#44944 )

2026-06-16 10:53:27 -07:00

buffer_utils.py

[Core] Ensure memory is pinned prior to async h2d copy (#45424 )

2026-06-20 20:02:24 -07:00

cp_utils.py

[Model Runner V2] Minor simplification for DCP (#34786 )

2026-02-18 11:04:53 -08:00

cudagraph_utils.py

[Optimization] Skip DP padding tokens in MoE (#46428 )

2026-06-23 14:54:46 -07:00

dp_utils.py

[Model Runner V2][Bugfix] Fix MRV2 LoRA warmup (#35536 )

2026-06-15 13:17:23 -07:00

eplb_utils.py

[EPLB] Mask padding in EPLB load recording (#38128 )

2026-06-28 19:43:58 -07:00

input_batch.py

[ModelRunner V2] Simplify recent UnlimitedOCR-related changes (#46975 )

2026-06-29 09:56:17 -07:00

kv_connector.py

[Core] Cleanup KVConnector handling with PP + fix MRV2 (#43732 )

2026-05-28 13:12:03 -07:00

lora_utils.py

[Model Runner V2][Bugfix] Fix MRV2 LoRA warmup (#35536 )

2026-06-15 13:17:23 -07:00

model_runner.py

[ModelRunner V2][BugFix] Free all model refs on shutdown (#47483 )

2026-07-02 19:32:48 -07:00

pp_utils.py

[Platform] Replace torch.cuda.Event with torch.Event (#47140 )

2026-06-30 08:39:59 -07:00

README.md

GPU Model Runner V2 (#25266 )

2025-11-21 08:20:55 -08:00

shutdown.py

[MRV2] Add shutdown() method (#41297 )

2026-05-02 21:06:30 -07:00

states.py

[ModelRunnerV2] Avoid pipeline parallel bubbles (#42187 )

2026-06-02 14:02:01 -07:00

structured_outputs.py

[ModelRunner V2] Misc minor simplifications and optimizations (#33467 )

2026-02-01 22:17:14 +00:00

warmup.py

[ModelRunner V2] Warmup cross-attn properly in encoder-decoder case (#47308 )

2026-07-01 12:36:48 -07:00

README.md

[Experimental] Model Runner V2

This directory contains the new model runner which is under active development. Ping Woosuk Kwon for any changes.