All issues

Search results

Consider using pre-built Flash Attention kernels via kernels

Status: Open.

#1582In FlagOpen/FlagEmbedding;

· sayakpaul opened on May 28, 2026

bge-reranker-v2-minicpm-layerwise微调报错

Status: Open.

#1581In FlagOpen/FlagEmbedding;

· yvway opened on May 19, 2026

Silent trust_remote_code=True enable in FlagAutoModel.from_finetuned via basename-only registry lookup

Status: Open.

#1579In FlagOpen/FlagEmbedding;

· beanduan22 opened on May 7, 2026

mac上使用omlx部署报错

Status: Open.

#1578In FlagOpen/FlagEmbedding;

· liumingliang0305-sudo opened on May 7, 2026

Embedder for m3-memory

Status: Open.

#1577In FlagOpen/FlagEmbedding;

· skynetcmd opened on May 6, 2026

rerank微调数据格式

Status: Open.

#1576In FlagOpen/FlagEmbedding;

· yvway opened on Apr 24, 2026

unable to use with transformers==5.2.0

Status: Open.

#1569In FlagOpen/FlagEmbedding;

· liu-yx17 opened on Mar 30, 2026

Activation Beacon，关于Loss定义的疑问：平均 token loss 是否会稀释真正依赖 beacon memory 的位置？

Status: Open.

#1565In FlagOpen/FlagEmbedding;

· Strandding opened on Mar 20, 2026

这行注释掉会导致微调模型时flashattention2报错(fla2只支持fp16、bf16，关还关不掉)，赶紧修复了吧。

Status: Open.

#1564In FlagOpen/FlagEmbedding;

· yy9996 opened on Mar 11, 2026

📋 Documentation Enhancement Suggestion

Status: Open.

#1562In FlagOpen/FlagEmbedding;

· croviatrust opened on Mar 4, 2026

question: calculate input_tokens about bge-reranker-v2-gemma

Status: Open.

#1559In FlagOpen/FlagEmbedding;

· llyycchhee opened on Dec 29, 2025

[Question] Seeking Spring Boot Starter or Best Practice for BGE Models in a RAG Project

Status: Open.

#1555In FlagOpen/FlagEmbedding;

· wangyz351 opened on Dec 25, 2025