AIbin c3aceb6bdc [Models][OP][Optimization] Support DeepSeek-v3.2 model, integrate DSA & Indexer architecture with FlashMLA/DeepGEMM (#6689) ...

* Support DeepSeek-v3.2 model, integrate DSA & Indexer architecture with FlashMLA/DeepGEMM

2026-03-10 15:05:14 +08:00

..

ernie4_5_vl

[Iluvatar] Support CudaGraph and optimize flash_attn_unpadded and fused_neox_rope_embedding (#6553)

2026-03-02 14:07:17 +08:00

paddleformers

fix paddleformers fallback (#6465)

2026-02-23 15:29:13 +08:00

paddleocr_vl

[Iluvatar] Support CudaGraph and optimize flash_attn_unpadded and fused_neox_rope_embedding (#6553)

2026-03-02 14:07:17 +08:00

qwen2_5_vl

[Iluvatar] Support CudaGraph and optimize flash_attn_unpadded and fused_neox_rope_embedding (#6553)

2026-03-02 14:07:17 +08:00

qwen3_vl

[loader]supoort wint2 backend (#6139)

2026-02-08 22:42:36 -08:00

__init__.py

[Models] Add Qwen3-VL Model Support (#5763)

2025-12-29 17:39:33 +08:00

adapters.py

[Feature] support reward model (#5301)

2025-12-02 14:55:31 +08:00

deepseek_v3.py

[Models][OP][Optimization] Support DeepSeek-v3.2 model, integrate DSA & Indexer architecture with FlashMLA/DeepGEMM (#6689)

2026-03-10 15:05:14 +08:00

ernie4_5_moe.py

[loader]supoort wint2 backend (#6139)

2026-02-08 22:42:36 -08:00

ernie4_5_mtp.py

Revert "Revert "[TSP] last_norm allgather move to model.py (#5924)" (#5961)" (#5972)

2026-01-09 15:58:22 +08:00

ernie_vl_rm.py

[Feature] support reward model (#5301)

2025-12-02 14:55:31 +08:00

glm4_moe.py

[Optimization] Enable BF16 gate computation for GLM and Qwen (#6457)

2026-02-26 21:08:46 -08:00

glm4_mtp.py

[RL] Support GLM MTP RL Model (#6267)

2026-02-04 20:14:35 +08:00

gpt_oss.py

[loader]supoort wint2 backend (#6139)

2026-02-08 22:42:36 -08:00

interfaces_base.py

Optimizing the performance of think length limit using custom operators (#4279)

2025-10-20 21:09:13 +08:00

model_base.py

[Models][OP][Optimization] Support DeepSeek-v3.2 model, integrate DSA & Indexer architecture with FlashMLA/DeepGEMM (#6689)

2026-03-10 15:05:14 +08:00

qwen2_rm.py

add qwen-2.5-7B-PRM/ernie-rm (#4319)

2025-10-20 15:31:03 +08:00

qwen2.py

[loader]supoort wint2 backend (#6139)

2026-02-08 22:42:36 -08:00

qwen3.py

[loader]supoort wint2 backend (#6139)

2026-02-08 22:42:36 -08:00

qwen3moe.py

clean nvfp4 related code (#6644)

2026-03-05 15:48:33 +08:00

tp_utils.py

[Others] Rename tensor_parallel_degree to tensor_model_parallel_size for paddleformers 0.4.1 (#5727)

2025-12-23 23:19:11 -08:00

utils.py

delete useless code (#4544)

2025-10-23 13:40:34 +08:00