Search results

量化方法的比较

Status: Open.

#67 In AkaliKong/MiniOneRec;

· AlphaMeoww opened on Mar 14, 2026

在RL训练的时候开启test_during_training=True会导致大量no valid token的wanring

Status: Open.

#66 In AkaliKong/MiniOneRec;

· Gavin-Cool opened on Mar 5, 2026

EvalSidDataset 和 SidSFTDataset 的 prompt 文本未对齐，想了解设计考量

Status: Open.

#65 In AkaliKong/MiniOneRec;

· cywuuuu opened on Mar 5, 2026

模型部署问题

Status: Open.

#64 In AkaliKong/MiniOneRec;

· yijingshihenxiule opened on Mar 3, 2026

能否提供更多baseline

Status: Open.

#63 In AkaliKong/MiniOneRec;

· cywuuuu opened on Feb 28, 2026

设置freeze_LLM=True 时 运行sft.py 报 NameError: original_vocab_size 未定义

Status: Open.

#61 In AkaliKong/MiniOneRec;

· xiaoxian-x opened on Feb 13, 2026

rqvae训练问题

Status: Open.

#60 In AkaliKong/MiniOneRec;

· sullicsullic opened on Feb 12, 2026

强化学习（RL）后模型性能下降

Status: Open.

#58 In AkaliKong/MiniOneRec;

· cywuuuu opened on Feb 1, 2026

测试的时候非常多UserWarning: No valid tokens found for hash_key

Status: Open.

#56 In AkaliKong/MiniOneRec;

· SuperTom99 opened on Jan 29, 2026

Fluctuations in Reinforcement Learning Rewards

Status: Open.

#55 In AkaliKong/MiniOneRec;

· fukioston opened on Jan 26, 2026

贴一个用BPE分词的BUG

Status: Open.

#54 In AkaliKong/MiniOneRec;

· lyzltysgithub opened on Jan 21, 2026

Which model is the priority choice for reproduction: Base or Instruct?

Status: Open.

#52 In AkaliKong/MiniOneRec;

· cywuuuu opened on Jan 14, 2026