Search results

[Bug] Hardcoded slicing in <code>forward</code> method causes data loss for SigLIP vision towers

Status: Open.

#203 In TinyLLaVA/TinyLLaVA_Factory;

· jhjangjh opened on Jan 24, 2026

模型格式转换

Status: Open.

#202 In TinyLLaVA/TinyLLaVA_Factory;

· ymliu-felix opened on Jan 23, 2026

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

Status: Open.

#201 In TinyLLaVA/TinyLLaVA_Factory;

· xiaoma-ux opened on Dec 23, 2025

如何更改LLM的损失函数

Status: Open.

#200 In TinyLLaVA/TinyLLaVA_Factory;

· wxxhaoshuai opened on Oct 17, 2025

tinyllava/training_recipe/qlora_recipe.py设置矛盾

Status: Open.

#198 In TinyLLaVA/TinyLLaVA_Factory;

· LuShuaie opened on Sep 25, 2025

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

Status: Open.

#197 In TinyLLaVA/TinyLLaVA_Factory;

· LuShuaie opened on Sep 25, 2025

Token indices sequence length is longer than the specified maximum sequence length for this model (2071 &gt; 2048).

Status: Open.

#192 In TinyLLaVA/TinyLLaVA_Factory;

· liuyafeng0609 opened on Sep 17, 2025

SAM数据集预训练问题

Status: Open.

#191 In TinyLLaVA/TinyLLaVA_Factory;

· liuyafeng0609 opened on Sep 15, 2025

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

Status: Open.

#190 In TinyLLaVA/TinyLLaVA_Factory;

· fjwang3 opened on Sep 9, 2025

关于TextVQA和GQA的测试结果以及LoRA微调的问题

Status: Open.

#189 In TinyLLaVA/TinyLLaVA_Factory;

· zhengjin11 opened on Sep 8, 2025

load pretrained model for loss computation

Status: Open.

#188 In TinyLLaVA/TinyLLaVA_Factory;

· ymliu-felix opened on Aug 27, 2025

partially-tune vision tower

Status: Open.

#185 In TinyLLaVA/TinyLLaVA_Factory;

· RaptorMai opened on Aug 14, 2025