iefgnoix b6be6f8d1e

[TPU] Support sliding window and logit soft capping in the paged attention kernel for TPU. (#15732) ...

Signed-off-by: Xiongfei Wei <isaacwxf23@gmail.com>

2025-04-03 14:23:28 -07:00

..

attention

[TPU] Support sliding window and logit soft capping in the paged attention kernel for TPU. (#15732)

2025-04-03 14:23:28 -07:00

core

[V1][Metrics] Initial speculative decoding metrics (#15151)

2025-04-01 10:45:04 -07:00

engine

[Bugfix][V1] Fix bug from putting llm_engine.model_executor in a background process (#15367)

2025-04-03 07:32:10 +00:00

executor

[V1] AsyncLLM data parallel (#13923)

2025-03-27 16:14:41 -07:00

metrics

[V1][Metrics] Initial speculative decoding metrics (#15151)

2025-04-01 10:45:04 -07:00

sample

[V1][TPU] TPU-optimized top-p implementation (avoids scattering). (#15736)

2025-04-02 17:18:08 -07:00

spec_decode

[V1][Minor] Enhance SpecDecoding Metrics Log in V1 (#15902)

2025-04-01 23:38:02 -07:00

stats

Update deprecated Python 3.8 typing (#13971)

2025-03-02 17:34:51 -08:00

structured_output

[V1] Fix json_object support with xgrammar (#15488)

2025-04-02 02:00:08 -07:00

worker

[Bugfix][V1] Fix bug from putting llm_engine.model_executor in a background process (#15367)

2025-04-03 07:32:10 +00:00

__init__.py

[V1] AsyncLLM Implementation (#9826)

2024-11-11 23:05:38 +00:00

kv_cache_interface.py

[V1] Implement sliding window attention in kv_cache_manager (#14097)

2025-04-01 00:33:17 -07:00

outputs.py

[V1] Aggregate chunked prompt logprobs in model runner (#14875)

2025-03-24 12:27:57 -04:00

request.py

[V1][Scheduler] Avoid calling _try_schedule_encoder_inputs for every request (#15778)

2025-03-30 14:10:42 -07:00

serial_utils.py

[V1] [Feature] Collective RPC (#15444)

2025-03-29 03:39:14 -07:00

utils.py

[V1] AsyncLLM data parallel (#13923)

2025-03-27 16:14:41 -07:00