Issues

• [Security] Arbitrary code execution: hardcoded trust_remote_code=True in TinyLlavaConfig._load_text_config

  documentationenhancement
  TinyLLaVA/TinyLLaVA_Factory#212 · Filter by author 3em0 opened on Sep 24, 2026

• Training

  bug
  TinyLLaVA/TinyLLaVA_Factory#211 · Filter by author ymliu-felix opened on Jun 16, 2026
  4

  

• [Bug] Hardcoded slicing in forward method causes data loss for SigLIP vision towers

  bug
  TinyLLaVA/TinyLLaVA_Factory#203 · Filter by author jhjangjh opened on Jan 24, 2026
  1

  

• 模型格式转换

  documentation
  TinyLLaVA/TinyLLaVA_Factory#202 · Filter by author ymliu-felix opened on Jan 23, 2026

• AttributeError: 'LlamaConfig' object has no attribute 'embd_pdrop'

  question
  TinyLLaVA/TinyLLaVA_Factory#201 · Filter by author xiaoma-ux opened on Dec 23, 2025

• 如何更改LLM的损失函数

  question
  TinyLLaVA/TinyLLaVA_Factory#200 · Filter by author wxxhaoshuai opened on Oct 17, 2025

• tinyllava/training_recipe/qlora_recipe.py设置矛盾

  bug
  TinyLLaVA/TinyLLaVA_Factory#198 · Filter by author LuShuaie opened on Sep 25, 2025
  

• tinyllava/training_recipe/lora_recipe.py使用lora时，并没有真正使用lora

  question
  TinyLLaVA/TinyLLaVA_Factory#197 · Filter by author LuShuaie opened on Sep 25, 2025

• Qwen2-7B for pre-training and fine-tuning

  documentation
  TinyLLaVA/TinyLLaVA_Factory#190 · Filter by author fjwang3 opened on Sep 9, 2025
  1

• 关于TextVQA和GQA的测试结果以及LoRA微调的问题

  documentation
  TinyLLaVA/TinyLLaVA_Factory#189 · Filter by author zhengjin11 opened on Sep 8, 2025
  9

• load pretrained model for loss computation

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#188 · Filter by author ymliu-felix opened on Aug 27, 2025
  1

• partially-tune vision tower

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#185 · Filter by author RaptorMai opened on Aug 14, 2025
  4

• Request for Alternative Download Links for Dataset (Baidu Unavailable in My Region)

  documentation
  TinyLLaVA/TinyLLaVA_Factory#184 · Filter by author TheDarkKnight-21th opened on Jul 7, 2025

• Is there options for not using flash-attn?

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#182 · Filter by author Gjj-114514 opened on Jun 30, 2025

• What if I want to apply larger LLM base with more parameters?

  bug
  TinyLLaVA/TinyLLaVA_Factory#181 · Filter by author Hoyant-Su opened on Jun 9, 2025
  1

  

• vision_feature_select_strategy

  bug
  TinyLLaVA/TinyLLaVA_Factory#180 · Filter by author WuYeeh opened on May 22, 2025
  1

  

• instruction finetune error

  bug
  TinyLLaVA/TinyLLaVA_Factory#178 · Filter by author WuYeeh opened on May 3, 2025
  10

• siglip2

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#177 · Filter by author sailfish009 opened on Apr 12, 2025
  1

• Evalution for Qwen2

  documentation
  TinyLLaVA/TinyLLaVA_Factory#176 · Filter by author kris-singh opened on Apr 6, 2025
  1

• 关于强化学习的问题

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#175 · Filter by author ctgushiwei opened on Mar 28, 2025
  1

• Problem on custom_fintune

  bug
  TinyLLaVA/TinyLLaVA_Factory#174 · Filter by author ZhihuaGao opened on Mar 26, 2025
  3

• 请问有没有不使用flash_attn的解决方案

  enhancement
  TinyLLaVA/TinyLLaVA_Factory#173 · Filter by author hzh12131108 opened on Mar 25, 2025

• base conv_mode for /scripts/eval/mme.sh

  documentation
  TinyLLaVA/TinyLLaVA_Factory#172 · Filter by author bastian1209 opened on Mar 5, 2025
  1

• For the 'share' training recipe, the adapter_config.json file is missing when LoRA is used in both the pretraining and fine-tuning processes.

  bug
  TinyLLaVA/TinyLLaVA_Factory#170 · Filter by author Lucky-Snowball opened on Jan 17, 2025

• The size of tensor a (1290) must match the size of tensor b (778) at non-singleton dimension 3

  bug
  TinyLLaVA/TinyLLaVA_Factory#169 · Filter by author mrpeerat opened on Jan 16, 2025
  2