TinyLLaVA TinyLLaVA_Factory issues

TinyLLaVA / TinyLLaVA_Factory

A Framework of Small-scale Large Multimodal Models

https://arxiv.org/abs/2402.14289

Apache License 2.0

658 stars 68 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

llama.cpp support

#36 TheBobbyliu closed 7 months ago
0
Saving vision tower in checkpointing

#35 zhangliang-04 closed 8 months ago
2
Did you train the whole LLM in the pretraining stage of share recipe？

#34 Yang-bug-star opened 8 months ago
3
请问phi-2是像llava-phi那样预先在文本指令上微调过的吗，还是直接拿基座模型连接视觉编码器进行预训练呢？

#33 Yang-bug-star opened 8 months ago
1
Question about commercial usage availability

#32 TheBobbyliu closed 8 months ago
1
Fine-tuning TinyLLaVA-1.5B on custom text-VQA dataset.

#31 soham-joshi closed 7 months ago
10
No module named 'tinyllava.eval.m4c_evaluator'

#30 soham-joshi closed 8 months ago
7
Inference seems can't work

#29 JalorOo closed 8 months ago
1
Gradio demo

#28 Lealand23231513 closed 8 months ago
0
Gradio demo

#27 Lealand23231513 closed 8 months ago
0
Changes to commit：

#26 Lealand23231513 closed 8 months ago
1
fix: delete the print function in phi.py

#25 JalorOo closed 8 months ago
0
Use merge_lora_weights.py Error:AttributeError: 'TinyLlavaPhiConfig' object has no attribute 'attention_bias'

#24 mengjiexu opened 8 months ago
3
Can Tiny Llama 1.5B lora finetune in one 24GB gpu?

#23 JalorOo closed 8 months ago
8
cls not defined!

#22 JalorOo closed 8 months ago
2
ModuleNotFoundError: No module named 'tinyllava'

#21 JalorOo closed 8 months ago
3
WARNING: tokenization mismatch: 0 vs

#20 G-JWLee closed 8 months ago
1
what the difference between v1 and v1.1?

#19 Eric-is-good opened 8 months ago
1
Is it convenient to provide some JSON files

#18 wangfengjuan closed 8 months ago
8
Why is the forward method of the Vision Encoder decorated with @torch.no_grad?

#17 hongming21 closed 8 months ago
2
Need training script

#16 dddraxxx opened 8 months ago
4
可以给个微调的脚本吗？

#15 LianghuiGuo closed 8 months ago
4
Are there ways to finetune Tiny-LLaVA 3.1B with LoRA?

#14 TheBobbyliu closed 8 months ago
1
the loss is nan when pre-training tinyllama using share recipe

#13 xushilin1 opened 8 months ago
1
Need train bash script, I got Error While training with llava1.5 bash script.

#12 mengjiexu closed 8 months ago
1
Checkpoints using base method and LLaVA dataset

#11 G-JWLee opened 8 months ago
1
Is the SigLip model required just to perform inference?

#10 plamb-viso closed 8 months ago
1
The result about llava-phi

#9 xushilin1 closed 8 months ago
11
Error when inferencing TinyLLaVA-1.5B

#8 eternalding closed 8 months ago
6
question about `unfreeze_mm_vision_tower`

#7 xushilin1 closed 9 months ago
1
Error about save model

#6 xushilin1 closed 8 months ago
3
Update README.md

#5 eltociear closed 9 months ago
0
【TinyLLaVA-3.1B Inference】

#4 Luo-Z13 closed 9 months ago
15
Dataset usage

#3 LinB203 closed 9 months ago
1
Good work!

#2 BlueBlueFF closed 9 months ago
2
Add some introduce on pretrain and sft data

#1 lucasjinreal closed 9 months ago
5