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

..

test_activation.py

[Feature] use phi permute/unpermute & rm swiglu (#6361)

2026-03-12 02:01:57 -07:00

test_append_attention_with_output.py

[Others]get_block_shape_and_split_kv_block clean code (#5123)

2025-11-20 16:40:04 +08:00

test_append_attention.py

[Feature]Supports SWA based on appendattn (#6547)

2026-03-01 19:02:08 +08:00

test_attention_layer.py

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

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

test_batched_count_greater_than.py

[Optimization] Optimization for gather_logprob by 10GB (#5817)

2025-12-30 15:33:34 +08:00

test_deepgemm_fused_moe.py

[BugFix] add ut for fused_moe_degemm (#6840)

2026-03-16 12:22:18 +08:00

test_ep_moe_expert_dispatch_fp8.py

[Others] remove template NUM_EXPERTS_PER_RANK in permute_x_fp8_kernel (#5620)

2026-01-04 11:21:15 +08:00

test_ffn.py

[Models] Add forward_meta to moe models' forward function (#5138)

2025-12-04 13:26:58 +08:00

test_flash_attention_versions_determinism.py

[Feature] Add Deterministic Inference Support (#6476)

2026-02-26 19:31:51 -08:00

test_flash_attn_func.py

fix fa4 test (#6408)

2026-02-10 10:57:21 +08:00

test_fp8_ue8m0.py

fix deepgemm import (#6451)

2026-02-11 20:10:01 +08:00

test_fused_moe_cutlass_backend.py

[CI] 【Hackathon 10th Spring No.23】fastdeploy/model_executor/layers/moe/fused_moe_cutlass_backend.py 单测补充 (#6209)

2026-02-28 19:29:02 +08:00

test_fused_moe_triton_backend.py

[CI] 【Hackathon 10th Spring No.24】功能模块 fastdeploy/model_executor/layers/moe/fused_moe_triton_backend.py 单测补充 (#6208)

2026-03-09 14:24:08 +08:00

test_fusedmoe.py

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

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

test_guided_decoding.py

[Others]update paddleformer 1.0.0 (#6496)

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

test_min_sampling.py

Add stable ci (#3460)

2025-08-20 08:57:17 +08:00

test_moba_attention_backend.py

[KVCache] support unified cache backend (#4903)

2025-11-12 14:54:52 +08:00

test_native_paddle_backend.py

[Docs] Add License in Unittest (#4957)

2025-11-12 10:44:09 +08:00

test_plas_attention.py

[Executor] Default use CUDAGraph (#3594)

2025-10-21 14:25:45 +08:00

test_qkvg_parallel_linear.py

[Feature] support qkv&gate linear fusion (#6455)

2026-02-24 15:20:29 +08:00

test_quantized_linear.py

[Optimization] Support FA2/FA3/FA4 with attn_mask_q (#6354)

2026-02-05 14:39:00 +08:00

test_repetition_early_stopper.py

Add stable ci (#3460)

2025-08-20 08:57:17 +08:00

test_sampler.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

test_speculative_sampler.py

[Optimization][Speculative Decoding]Fuse padding sampling params (#6765)

2026-03-12 05:05:15 -07:00

test_w4a8_moe.py

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

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

test_w4afp8_moe.py

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

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