[Cherry-Pick][BugFix] Support redundant expert for eplb (#5918)#5922

Closed

xiaoxiaohehe001 wants to merge 36 commits intoPaddlePaddle:developfrom

xiaoxiaohehe001:re_eplb

Commits on Nov 25, 2025

Commits on Nov 26, 2025

Commits on Nov 27, 2025

Commits on Nov 28, 2025

[Cherry-Pick][Feature] support flash_mask_attention backend(#5134) (#5256)
lizhenyun01
authored
[Cherry-pick][XPU][CI] Set pip index URL to Tsinghua mirror (#5277) (#5281)
plusNew001
authored
Update load_weight_utils.py (#5285)
yuanlehome
authored

Commits on Nov 29, 2025

fix mm to_dict bug (#5299)
kevincheng2
authored

Commits on Dec 2, 2025

[Optimization] 1.fix tp+ep moe_forward; 2.set max_prefill_batch=env.MAX_PREFILL_NUM (#5316)
carryyu
authored

Commits on Dec 3, 2025

supports mtp split_kv_attn (#5344)
carryyu
authored

Commits on Dec 5, 2025

[Cherry-Pick][BugFix] Fix async download(#5349) (#5347)
kevincheng2
authored

Commits on Dec 9, 2025

[Others] Maintain the mtp branch temporarily. (#5447)
carryyu
authored

Commits on Dec 10, 2025

Commits on Dec 11, 2025

[CI] disable test_cuda_graph_dynamic_subgraph.py in unit_test
EmmonsCurse
authored

Commits on Dec 15, 2025

[Optimize][Cherry-pick] Robust stabilty for PD deployment #5338 (#5395)

rainyfly
and
heavengate
authored

Commits on Dec 16, 2025

Commits on Dec 17, 2025

[Cherry-Pick][BugFix] fix speculate_limit_thinking_content_length #5590 (#5591)

yuanlehome
and
EmmonsCurse
authored

Commits on Dec 23, 2025

Commits on Dec 24, 2025

Commits on Dec 26, 2025

Commits on Dec 30, 2025

[Feature] two chunk overlap (#5754)

zhoutianzi666
and
yangjianfengo1
authored

Commits on Dec 31, 2025

fix quant (#5837)
Sunny-bot1
authored

Commits on Jan 5, 2026

add del to decrease peak memory (#5862)
zhoutianzi666
authored

Commits on Jan 7, 2026

[Cherry-Pick] Support redundant expert for eplb
xiaoxiaohehe001
committed