sunxin 33e01f22a8 [Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894) ...

* update sampling

* fix

* fix

* fix mtp

* fix test

2026-03-19 01:43:10 -07:00

..

ernie4_5_vl_processor

[Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894)

2026-03-19 01:43:10 -07:00

paddleocr_vl_processor

[Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894)

2026-03-19 01:43:10 -07:00

qwen3_vl_processor

[Processor]add qwen3vl prompt_token_ids support (#6764)

2026-03-11 15:08:56 +08:00

qwen_vl_processor

[Others]update paddleformer 1.0.0 (#6496)

2026-03-11 15:06:29 +08:00

v1

[Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894)

2026-03-19 01:43:10 -07:00

__init__.py

polish code with new pre-commit rule (#2923)

2025-07-19 23:19:27 +08:00

ernie4_5_processor.py

[Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894)

2026-03-19 01:43:10 -07:00

ernie4_5_tokenizer.py

[Others]update paddleformer 1.0.0 (#6496)

2026-03-11 15:06:29 +08:00

mm_data_processor.py

[Feature] consider multimodal model when dummy run (#6045)

2026-02-09 17:49:55 +08:00

preprocess.py

[BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533)

2026-03-16 21:32:43 +08:00

text_processor.py

[Feature][Sampling] Extend top-k_top-p sampling to all backends and unify greedy decoding with top_k=1 (#6894)

2026-03-19 01:43:10 -07:00

tokenzier_client.py

[BugFix][Optimization] Replace silent failures with catchable exceptions and informative error messages (#6533)

2026-03-16 21:32:43 +08:00

utils.py

[BugFix][DataProcessor] Add validate_model_path to fail fast on bad model path or unreachable network (#6713)

2026-03-08 12:36:32 +08:00