Yong Hoon ShinandYe Qi 99ef59cf7f [Llama4] Enable attention temperature tuning by default for long context (>32k) (#16439) ...

Signed-off-by: Ye (Charlotte) Qi <yeq@meta.com>
Co-authored-by: Ye (Charlotte) Qi <yeq@meta.com>

2025-04-10 21:26:07 -07:00

..

adapter_commons

[Misc] Add SPDX-License-Identifier headers to python source files (#12628)

2025-02-02 11:58:18 -08:00

assets

[Misc] Allow using OpenCV as video IO fallback (#15055)

2025-04-01 15:55:13 +00:00

attention

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

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

benchmarks

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

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

compilation

Fix the torch version parsing logic (#15857)

2025-04-10 07:37:47 -07:00

core

[FEAT]Support reset prefix cache by specified device (#15003)

2025-03-19 10:54:41 -07:00

device_allocator

[core] Add tags parameter to wake_up() (#15500)

2025-04-02 01:59:27 -07:00

distributed

[Misc] Merge the logs of pp layers partitions (#16225)

2025-04-08 00:18:15 -07:00

engine

[V1] Set structured output backend to auto by default (#15724)

2025-04-10 17:53:26 +00:00

entrypoints

[misc] use tqdm.auto where appropriate (#16290)

2025-04-09 21:54:54 -07:00

executor

[core] Add tags parameter to wake_up() (#15500)

2025-04-02 01:59:27 -07:00

inputs

[V1] Remove legacy input registry (#15673)

2025-03-27 23:34:34 -07:00

logging_utils

[Misc] Add SPDX-License-Identifier headers to python source files (#12628)

2025-02-02 11:58:18 -08:00

lora

[Bugfix] LoRA : Fix the order in which the kernels process LoRAs (#16040)

2025-04-06 14:04:50 +00:00

model_executor

[Llama4] Enable attention temperature tuning by default for long context (>32k) (#16439)

2025-04-10 21:26:07 -07:00

multimodal

[VLM] Avoid unnecessary dummy multimodal data during processing (#16416)

2025-04-10 19:32:14 +00:00

platforms

[TPU][V1] Disable per-request seed/Generator (#16172)

2025-04-10 17:05:44 -04:00

plugins

[Misc] Add SPDX-License-Identifier headers to python source files (#12628)

2025-02-02 11:58:18 -08:00

profiler

[Misc] Add SPDX-License-Identifier headers to python source files (#12628)

2025-02-02 11:58:18 -08:00

prompt_adapter

[Misc] set single whitespace between log sentences (#13771)

2025-02-25 10:26:12 +08:00

reasoning

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

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

spec_decode

[V1][Usage] Refactor speculative decoding configuration and tests (#14434)

2025-03-22 19:28:10 -10:00

third_party

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

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

transformers_utils

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

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

triton_utils

[Frontend] Remove custom_cache_manager (#13791)

2025-03-20 00:13:50 +08:00

usage

[Misc] Upgrade to Python 3.9 typing for additional directories (#14492)

2025-03-08 17:35:50 +00:00

v1

[TPU][V1] Disable per-request seed/Generator (#16172)

2025-04-10 17:05:44 -04:00

vllm_flash_attn

[V1] Enable V1 Fp8 cache for FA3 in the oracle (#15191)

2025-03-23 15:07:04 -07:00

worker

[misc] use tqdm.auto where appropriate (#16290)

2025-04-09 21:54:54 -07:00

__init__.py

[misc] instruct pytorch to use nvml-based cuda check (#15951)

2025-04-03 01:02:58 +08:00

_custom_ops.py

[ROCM][KERNEL] Paged attention for V1 (#15720)

2025-04-02 19:48:00 -07:00

_ipex_ops.py

[Kernel][CPU] CPU MLA (#14744)

2025-03-25 09:34:59 +00:00

beam_search.py

Update deprecated Python 3.8 typing (#13971)

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

config.py

[V1] Set structured output backend to auto by default (#15724)

2025-04-10 17:53:26 +00:00

connections.py

Update deprecated Python 3.8 typing (#13971)

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

env_override.py

[misc] instruct pytorch to use nvml-based cuda check (#15951)

2025-04-03 01:02:58 +08:00

envs.py

[Core] Upgrade to xgrammar 0.1.18, add cache size limit (#16283)

2025-04-08 19:13:22 -07:00

forward_context.py

[Bugfix] Remove num_tokens_across_dp (#14302)

2025-03-05 23:55:55 +00:00

jsontree.py

[VLM] Limit multimodal input cache by memory (#14805)

2025-03-15 02:52:05 -07:00

logger.py

Update deprecated Python 3.8 typing (#13971)

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

logits_process.py

Update deprecated Python 3.8 typing (#13971)

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

outputs.py

[Bugfix][V0] Multi-sequence logprobs streaming edge case (#15259)

2025-03-21 22:35:37 -07:00

pooling_params.py

[Misc] Add SPDX-License-Identifier headers to python source files (#12628)

2025-02-02 11:58:18 -08:00

py.typed

Add py.typed so consumers of vLLM can get type checking (#1509)

2023-10-30 14:50:47 -07:00

sampling_params.py

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

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

scalar_type.py

[NVIDIA] Support nvfp4 quantization (#12784)

2025-02-12 19:51:51 -08:00

scripts.py

[Frontend] Move CLI code into vllm.cmd package (#12971)

2025-02-12 23:12:21 -08:00

sequence.py

[Bugfix] EAGLE output norm bug (#14464)

2025-03-15 06:50:33 +00:00

test_utils.py

Use smaller embedding model when not testing model specifically (#13891)

2025-02-28 00:50:43 -08:00

tracing.py

Update deprecated Python 3.8 typing (#13971)

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

utils.py

Fix the torch version parsing logic (#15857)

2025-04-10 07:37:47 -07:00

version.py

[Metrics] Hide deprecated metrics (#15458)

2025-04-02 07:37:19 -07:00