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

基于qwen2-1.5B的训练

#135 riariam opened 1 week ago
1
Could not load library libcudnn_cnn_infer.so.8. Error: libcuda.so: cannot open shared object file: No such file or directory

#134 hyun95roh closed 1 week ago
1
[Feature Request] qwen2.5 and llava onevision

#133 sailfish009 opened 1 week ago
5
failure to build flash-attn

#132 hyun95roh closed 1 week ago
2
--tune_type_llm set to full but LLM parameters are not trainable

#131 YenCheHsiao closed 1 week ago
4
EnvironmentError: Accessing Gated Repo in train_openelm.sh

#130 YenCheHsiao closed 1 week ago
4
ModuleNotFoundError: No module named 'tinyllava'

#129 mrpeerat closed 2 weeks ago
1
VQA下游微调阶段GPU的显存占用波动很大

#128 DeguangChen opened 1 month ago
9
fix bug for lora finetune

#127 liboaccn closed 1 month ago
0
dependency conflict

#126 Nengzyue opened 1 month ago
1
Bugfix for load_pretrained_model and training_model_converse

#125 liboaccn closed 1 month ago
0
请问在训练qwen2时，需要做什么特殊处理吗？

#124 riariam opened 1 month ago
3
请问能做VQA任务吗

#123 DeguangChen opened 1 month ago
2
如何merge完整模型

#122 liboaccn closed 1 month ago
9
请问想用TinyLLaVA factory 基于 Qwen2 - 1.5B 进行训练，该如何进行

#121 baiyang2464 opened 1 month ago
4
loss does not decrease

#120 wangfengjuan opened 1 month ago
6
After reading "readme.md for qwen2"

#119 joony414 closed 1 month ago
2
损失函数的值不下降，一直维持在3到4

#118 MichealZhangxa opened 1 month ago
4
data_path能不能添加多个数据集

#117 xiaobanli opened 1 month ago
1
Request for Weights of TinyLLaVA Model Built on Tinyllama with Siglip

#116 phh713 opened 2 months ago
1
Export To Common Formats

#115 TalonBvV opened 2 months ago
1
Model type `tinyllava` not recognized by Transformers

#114 AaronChuh closed 2 months ago
1
surpport on mobile device

#113 perp opened 2 months ago
0
Unable to use GPUs 4,5,6,7 for training

#112 shashwat14 opened 2 months ago
4
About the Vision Tower.

#111 Sootung opened 3 months ago
1
qformer

#110 PangziZhang523 closed 2 months ago
1
Finetuning get: ValueError: not enough values to unpack (expected 2, got 0) in get_modality_length_grouped_indices

#109 sc268 opened 3 months ago
2
Update README.md

#108 huangleiBuaa closed 3 months ago
0
Add '--raise_error_at_min_scale False' flag in pretrain and finetune scripts to avoid minimum loss scale crash caused by bad batch.

#107 RobotiX101 closed 3 months ago
0
Tinyllava visualizer

#106 lingcco closed 3 months ago
0
在使用hf上下载TinyLLaVA-Phi-2-SigLIP-3.1B模型时，参数设置。

#105 eva10084 opened 3 months ago
0
Continue fine-tuning from previously trained model: error on inference

#104 simoneriggi opened 3 months ago
8
Does the Phi template work for Phi3？

#103 NyKxo1 opened 3 months ago
1
About share recipe. Should change "tune_type_vision_tower" to set to "partially-tune"?

#102 NyKxo1 opened 3 months ago
1
About "conv_version" and PretrainTemplate .

#101 NyKxo1 closed 2 months ago
1
TypeError: __init__() takes 1 positional argument but 2 were given

#100 RobotiX101 closed 3 months ago
0
Exception: Current Loss Scale Already at Minimum While Fine-Tuning with 8 A800-80G GPUs

#99 codefanw closed 3 months ago
2
add code for qwen2 0.5B

#98 bhneo closed 4 months ago
0
Can you reproduce the Llava results in this project?

#97 Yangr116 opened 4 months ago
2
Gemma template for integrating instruct tuned model

#96 Jayantverma2 closed 2 months ago
1
replace llm

#95 LilDevsy0117 opened 4 months ago
1
mof_mlp error

#94 Daming-W opened 4 months ago
1
distributed computing

#93 1764758458 opened 4 months ago
1
Results are not reproducible on Qwen1.5-1.8B

#92 Fantasy1120 opened 4 months ago
0
dataset seems imcomplete

#91 BroJunn closed 4 months ago
1
Fix `group_by_modality_length` in sync with LLaVA repo

#90 ggcr opened 4 months ago
0
Why `group_by_modality` is set to False on the `custom_finetune.sh` script?

#89 ggcr opened 5 months ago
3
Missing key "lm_head.weight" in GemmaForCausalLM when loading lora finetuned TinyLLaVA-Gemma-SigLIP-2.4B

#88 Yuki-Kokomi opened 5 months ago
5
Create a tinyllava with qwen2-0.5B-instruct

#87 bil-ash opened 5 months ago
0
really_cleaned_share-captioner_coco_lcs_sam_1246k_1107.json not found

#86 Wangman1 opened 5 months ago
2