21063c11c7
[CI/Build] drop support for Python 3.8 EOL ( #8464 )
...
Signed-off-by: Aaron Pham <contact@aarnphm.xyz >
2024-11-06 07:11:55 +00:00
bbc3619dc8
[Core] Make encoder-decoder inputs a nested structure to be more composable ( #9604 )
...
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk >
2024-11-05 10:07:31 +08:00
4fdc581f9e
[core] simplify seq group code ( #9569 )
...
Co-authored-by: Zhuohan Li <zhuohan123@gmail.com >
2024-10-24 00:16:44 -07:00
d11bf435a0
[MISC] Consolidate cleanup() and refactor offline_inference_with_prefix.py ( #9510 )
2024-10-18 14:30:55 -07:00
051eaf6db3
[Model] Add user-configurable task for models that support both generation and embedding ( #9424 )
2024-10-18 11:31:58 -07:00
81ede99ca4
[Core] Deprecating block manager v1 and make block manager v2 default ( #8704 )
...
Removing the block manager v1. This is the initial piece of prefix-caching-centric design. In order to achieve prefix-caching-centric design, we need to simplify the code path so that we only use v2 block manager (which has much higher performance on prefix caching).
2024-10-17 11:38:15 -05:00
f3a507f1d3
[Core] Add an environment variable which needs to be set explicitly to allow BlockSpaceManagerV1 ( #9149 )
2024-10-10 14:17:17 +08:00
18b296fdb2
[core] remove beam search from the core ( #9105 )
2024-10-07 05:47:04 +00:00
cb3b2b9ba4
[Bugfix] Fix incorrect updates to num_computed_tokens in multi-step scheduling ( #9038 )
...
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com >
2024-10-06 12:48:11 -07:00
5bf8789b2a
[Bugfix] Block manager v2 with preemption and lookahead slots ( #8824 )
2024-09-29 09:17:45 +08:00
fc3afc20df
Fix tests in test_chunked_prefill_scheduler which fail with BlockManager V2 ( #8752 )
2024-09-24 21:26:36 -07:00
ee777d9c30
Fix test_schedule_swapped_simple in test_scheduler.py ( #8780 )
2024-09-24 21:26:18 -07:00
88577ac928
Fix tests in test_scheduler.py that fail with BlockManager V2 ( #8728 )
2024-09-24 04:43:13 +00:00
e3580537a4
[Performance] Enable chunked prefill and prefix caching together ( #7753 )
2024-08-28 00:36:31 -07:00
2eedede875
[Core] Asynchronous Output Processor ( #7049 )
...
Co-authored-by: Alexander Matveev <alexm@neuralmagic.com >
2024-08-26 20:53:20 -07:00
2deb029d11
[Performance][BlockManagerV2] Mark prefix cache block as computed after schedule ( #7822 )
2024-08-26 11:24:53 -07:00
3ac50b47d0
[MISC] Add prefix cache hit rate to metrics ( #7606 )
2024-08-19 11:52:07 -07:00
ff7ec82c4d
[Core] Optimize SPMD architecture with delta + serialization optimization ( #7109 )
2024-08-18 17:57:20 -07:00
baa240252e
[Core] Fix edge case in chunked prefill + block manager v2 ( #7380 )
2024-08-09 23:48:49 +00:00
782e53ab59
[Bugfix][fast] Fix the get_num_blocks_touched logic ( #6849 )
2024-08-08 10:43:30 -07:00
fd95e026e0
[Core] Subclass ModelRunner to support cross-attention & encoder sequences (towards eventual encoder/decoder model support) ( #4942 )
...
Co-authored-by: Andrew Feldman <afeld2012@gmail.com >
Co-authored-by: Nick Hill <nickhill@us.ibm.com >
2024-08-06 16:51:47 -04:00
c8a7e93273
[core][scheduler] simplify and improve scheduler ( #6867 )
2024-07-31 23:51:09 -07:00
42c7f66a38
[Core] Support dynamically loading Lora adapter from HuggingFace ( #6234 )
...
Co-authored-by: Antoni Baum <antoni.baum@protonmail.com >
2024-07-22 15:42:40 -07:00
9ed82e7074
[Misc] Small perf improvements ( #6520 )
2024-07-19 12:10:56 -07:00
3476ed0809
[Core] Optimize block_manager_v2 vs block_manager_v1 (to make V2 default) ( #5602 )
2024-07-01 20:10:37 -07:00
0e9164b40a
[mypy] Enable type checking for test directory ( #5017 )
2024-06-15 04:45:31 +00:00
1b8a0d71cf
[Core][Bugfix]: fix prefix caching for blockv2 ( #5364 )
...
Signed-off-by: Lei Wen <wenlei03@qiyi.com >
Co-authored-by: Lei Wen <wenlei03@qiyi.com >
2024-06-14 17:23:56 -07:00
847cdcca1c
[CI] Upgrade codespell version. ( #5381 )
2024-06-12 10:06:14 -07:00
10c38e3e46
[Misc]: Implement CPU/GPU swapping in BlockManagerV2 ( #3834 )
2024-06-03 13:37:11 -07:00
b1c255630d
[Core] Avoid the need to pass None values to Sequence.inputs ( #5099 )
2024-05-29 16:05:01 -07:00
4238bc82f2
[Core] Cross-attention KV caching and memory-management (towards eventual encoder/decoder model support) ( #4837 )
2024-05-29 16:09:13 +00:00
5ae5ed1e60
[Core] Consolidate prompt arguments to LLM engines ( #4328 )
...
Co-authored-by: Roger Wang <ywang@roblox.com >
2024-05-28 13:29:31 -07:00
d4f3985907
[Core] Sliding window for block manager v2 ( #4545 )
...
Co-authored-by: Ruth Evans <ruthevans@Ruths-MacBook-Pro.local >
2024-05-28 11:07:07 +09:00
e64fde4b01
[Core][Bugfix]: fix prefix caching for blockv2 ( #4764 )
...
Co-authored-by: Lei Wen <wenlei03@qiyi.com >
2024-05-24 10:07:09 -07:00
e7c46b9527
[Scheduler] Warning upon preemption and Swapping ( #4647 )
...
Co-authored-by: Robert Shaw <114415538+robertgshaw2-neuralmagic@users.noreply.github.com >
2024-05-13 23:50:44 +09:00
350f9e107f
[CI/Build] Move test_utils.py to tests/utils.py ( #4425 )
...
Since #4335 was merged, I've noticed that the definition of ServerRunner in the tests is the same as in the test for OpenAI API. I have moved the class to the test utilities to avoid code duplication. (Although it only has been repeated twice so far, I will add another similar test suite in #4200 which would duplicate the code a third time)
Also, I have moved the test utilities file (test_utils.py) to under the test directory (tests/utils.py), since none of its code is actually used in the main package. Note that I have added __init__.py to each test subpackage and updated the ray.init() call in the test utilities file in order to relative import tests/utils.py.
2024-05-13 23:50:09 +09:00
fcc2994be6
[CI] Nits for bad initialization of SeqGroup in testing ( #4748 )
2024-05-10 18:01:01 -04:00
20cfcdec99
[Core][Optimization] change python dict to pytorch tensor for blocks to swap ( #4659 )
2024-05-08 12:07:05 -07:00
469f85c782
[Core][Optimization] change copy-on-write from dict[int, list] to list ( #4648 )
2024-05-07 11:06:32 -07:00
63575bc2e1
[Core][Optimization] change python dict to pytorch tensor ( #4607 )
2024-05-06 21:30:27 -07:00
0f8a91401c
[Core] Ignore infeasible swap requests. ( #4557 )
2024-05-02 14:31:20 -07:00
24750f4cad
[Core] Enable prefix caching with block manager v2 enabled ( #4142 )
...
Co-authored-by: Lei Wen <wenlei03@qiyi.com >
Co-authored-by: Sage Moore <sagemoore@utexas.edu >
2024-05-01 11:20:32 -07:00
050f285ff6
[Core] Scheduling optimization 2 ( #4280 )
2024-04-23 08:02:11 +00:00
ad8d696a99
[Core] Scheduler perf fix ( #4270 )
2024-04-22 21:11:06 +00:00
e95cd87959
[Speculative decoding 6/9] Integrate speculative decoding with LLMEngine ( #3894 )
2024-04-16 13:09:21 -07:00
67b4221a61
[Core][5/N] Fully working chunked prefill e2e ( #3884 )
2024-04-10 17:56:48 -07:00
e7c7067b45
[Misc] [Core] Implement RFC "Augment BaseExecutor interfaces to enable hardware-agnostic speculative decoding" ( #3837 )
2024-04-09 11:44:15 -07:00
18de883489
[Chunked Prefill][4/n] Chunked prefill scheduler. ( #3853 )
2024-04-05 10:17:58 -07:00
3dcb3e8b98
[3/N] Refactor scheduler for chunked prefill scheduling ( #3550 )
2024-04-03 14:13:49 -07:00
eb69d68804
[Misc] [CI/Build] Speed up block manager CPU-only unit tests ~10x by opting-out of GPU cleanup ( #3783 )
2024-04-02 00:49:51 +00:00
