Search results

Bug Report of monkeypatch.py

Status: Open.

#63 In Zefan-Cai/KVCache-Factory;

· viva-rock opened on Jan 22, 2026

[INFO] 给后来者的提醒 / Reminder for Future Users

Status: Open.

#62 In Zefan-Cai/KVCache-Factory;

· relic-yuexi opened on Oct 1, 2025

Question about Figure 2 in PyramidKV paper?

Status: Open.

#61 In Zefan-Cai/KVCache-Factory;

· dsj96 opened on Sep 22, 2025

BUG report: Persistent kv_seq_len breaks core pruning logic: kv_cluster.update_kv skipped after first batch

Status: Open.

#59 In Zefan-Cai/KVCache-Factory;

· vlev02 opened on Jul 14, 2025

Question about attention weight range selection in PyramidKV

Status: Open.

#58 In Zefan-Cai/KVCache-Factory;

· xiongsiheng opened on Jun 7, 2025

Potential Problem in Selecting Indices

Status: Open.

#57 In Zefan-Cai/KVCache-Factory;

· Lueci4er opened on Apr 24, 2025

About the efficiency of CAM implementation

Status: Open.

#55 In Zefan-Cai/KVCache-Factory;

· qqtang-code opened on Mar 24, 2025

Code error in reproducing the result of PyramidKV paper.

Status: Open.

#54 In Zefan-Cai/KVCache-Factory;

· yankeee-0411 opened on Mar 10, 2025

Can you provide a script to test the decoding latency and peak memory usage?

Status: Open.

#53 In Zefan-Cai/KVCache-Factory;

· CR400AF-A opened on Feb 19, 2025

对于Sdpa 与Flash Attention显存计算问题

Status: Open.

#52 In Zefan-Cai/KVCache-Factory;

· Couteaux123 opened on Jan 19, 2025

Regarding the Implementation Details of H2O

Status: Open.

#49 In Zefan-Cai/KVCache-Factory;

· masn1310 opened on Jan 4, 2025

提升不是很明显，在Meta-Llama-3-8B-Instruct上，与原论文的效果不符，能解释一下原因吗

Status: Open.

#46 In Zefan-Cai/KVCache-Factory;

· qqtang-code opened on Dec 30, 2024