Chengji Yao 1eb3c2ed48 [DOC][TPU] Add core idea about avoiding recompilation after warmup (#16614) ...

Signed-off-by: Chengji Yao <chengjiyao@google.com>

2025-04-14 21:56:06 +00:00

..

attention

[Kernel] support merge_attn_states CUDA kernel, 3x speedup (#16173)

2025-04-11 06:50:50 -06:00

core

[V1][Spec Decode] KV cache slots for eagle heads (#16370)

2025-04-12 19:42:51 -07:00

engine

[Bugfix] Multi-modal caches not acting like LRU caches (#16593)

2025-04-14 09:24:16 -07:00

executor

[BugFix] fix some typos found by typos. (#16314)

2025-04-09 03:43:59 -07:00

metrics

[Metrics] Add bucket for request_latency, time_to_first_token and time_per_output_token (#15202)

2025-04-06 20:34:51 -07:00

sample

[Bugfix] Validate logit biases to prevent out of vocab ids crashing engine (#16529)

2025-04-12 20:19:19 +00:00

spec_decode

[V1][Spec Decode] Eagle Model loading (#16035)

2025-04-10 11:21:48 -07:00

stats

Update deprecated Python 3.8 typing (#13971)

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

structured_output

[Feature][V1] Add xgrammar to support minLength, maxLength with test (#16516)

2025-04-11 23:22:07 -07:00

worker

[DOC][TPU] Add core idea about avoiding recompilation after warmup (#16614)

2025-04-14 21:56:06 +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

[Bugfix] Avoid transferring cached multi-modal items from P0 to P1 (#16273)

2025-04-09 00:51:27 -07:00

serial_utils.py

[BugFix] Handle non-contiguous tensors properly when serializing (#16492)

2025-04-11 17:54:06 -07:00

utils.py

[V1][BugFix] Exit properly if engine core fails during startup (#16137)

2025-04-07 15:30:15 -07:00