Nicolò Lucchesi 6116ca8cd7 [Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132) ...

Signed-off-by: NickLucche <nlucches@redhat.com>
Signed-off-by: wallashss <wallashss@ibm.com>
Co-authored-by: wallashss <wallashss@ibm.com>

2025-01-27 13:38:35 -08:00

..

__init__.py

[Speculative decoding 7/9] Speculative decoding end-to-end correctness tests. (#3951)

2024-04-23 08:02:36 +00:00

conftest.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_compatibility.py

[Bugfix] Fix for Spec model TP + Chunked Prefill (#10232)

2024-11-26 09:11:16 -08:00

test_eagle_correctness.py

[Core] Deprecating block manager v1 and make block manager v2 default (#8704)

2024-10-17 11:38:15 -05:00

test_integration_dist_tp2.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_integration_dist_tp4.py

[ci] fix broken distributed-tests-4-gpus (#11937)

2025-01-11 09:07:24 +08:00

test_integration.py

[Core] Deprecating block manager v1 and make block manager v2 default (#8704)

2024-10-17 11:38:15 -05:00

test_logprobs.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_medusa_correctness.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_mlp_correctness.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_multistep_correctness.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_ngram_correctness.py

[Feature] [Spec decode]: Enable MLPSpeculator/Medusa and prompt_logprobs with ChunkedPrefill (#10132)

2025-01-27 13:38:35 -08:00

test_seed.py

[Core] Deprecating block manager v1 and make block manager v2 default (#8704)

2024-10-17 11:38:15 -05:00