InternVL

InternVL copied to clipboard

baochi0212

[Bug]推理时模型加载报错

1

### Checklist - [X] 1. I have searched related issues but cannot get the expected help. - [X] 2. The bug has not been fixed in the latest version. -...

striver666

[Bug] OCR-related task return unicode instead of UTF-8 character

5

### Checklist - [X] 1. I have searched related issues but cannot get the expected help. - [X] 2. The bug has not been fixed in the latest version. -...

KohakuBlueleaf

Feat: support balanced dataset to speed-up training

1

Balanced Dynamic Mini-Batch for paper [OmniBal: Towards Fast Instruct-tuning for Vision-Language Models via Omniverse Computation Balance](https://arxiv.org/abs/2407.20761). Original Repo: https://github.com/ModelTC/OmniBal

yqyao

added hf compatibility

1

snova-guhae

多卡推理internVL2-40B报错：RuntimeError: CUDA error: device-side assert triggered

1

您好~使用多卡推理时报错如下：请问怎么解决额？谢谢 "index out of bounds"` failed. ../aten/src/ATen/native/cuda/IndexKernel.cu:92: operator(): block: [1,0,0], thread: [21,0,0] Assertion `-sizes[i]

muximuxi

[Feature] 请问如何在internvl模型代码上实现mixture-of- experts版本呢

1

### Motivation 我尝试加入MOE，但是在hidden_states经过F.Linear(embed_dim, num_of_experts)这个线性层后就变成了nan，假设num_of_experts为4，请问这是怎么回事呢，菜菜希望得到解答，万分感谢！ ### Related resources _No response_ ### Additional context _No response_

zweiqi

[Bug] Apply fix https://github.com/OpenGVLab/InternVL/pull/127 to the huggingface pretrained model

1

### Checklist - [x] 1. I have searched related issues but cannot get the expected help. - [X] 2. The bug has not been fixed in the latest version. -...

xyg-coder

internvl2 video data train CUDA out of memory

1

我正在对internvl2使用视频数据进行full finetune，显卡为单张32G V100，报错torch.cuda.OutOfMemoryError: CUDA out of memory. torchrun /cache/InternVL/internvl_chat/internvl/train/internvl_chat_finetune.py \ --model_name_or_path /cache/MODELS/internvl2-4B \ --conv_style "phi3-chat" \ --output_dir /cache/InternVL/OUTPUTS/internvl_chat_v1_5_phi3_3_8b_dynamic_res_finetune_debug_load_2nd \ --meta_path /cache/InternVL/internvl_chat/shell/data/internvl_1_2_finetune_7k.json \ --overwrite_output_dir True \ --force_image_size 448 \ --max_dynamic_patch 1...

gaotiexinqu

[Bug] InternVL2-Llama3-76B-AWQ infer result is wrong

9

### Checklist - [X] 1. I have searched related issues but cannot get the expected help. - [x] 2. The bug has not been fixed in the latest version. -...

AmazDeng