Janus icon indicating copy to clipboard operation
Janus copied to clipboard

Janus-Series: Unified Multimodal Understanding and Generation Models

Results 156 Janus issues
Sort by recently updated
recently updated
newest added

Run the service with command: 【python3 app_januspro.py】 Python version is above 3.10, patching the collections module. /Library/Frameworks/Python.framework/Versions/3.12/lib/python3.12/site-packages/transformers/models/auto/image_processing_auto.py:590: FutureWarning: The image_processor_class argument is deprecated and will be removed in v4.42. Please...

16g跑7b就差一点点,我真是吐了,吊模型。

It would be amazing if there is a fine tuning script to use on this amazing model!

Hello, awesome work on Janus-Pro! This PR enables Janus-Pro deployment on Replicate through two channels: - Web interface: [deepseek-ai/janus-pro-7b](https://replicate.com/deepseek-ai/janus-pro-7b) - API endpoint: [deepseek-ai/janus-pro-7b/api](https://replicate.com/deepseek-ai/janus-pro-7b/api) Additionally, the demo page has been migrated...

I have implemented some code related to understanding fine-tuning and used sample from inference.py as a reference. Feedback and suggestions are welcome! ```python import torch import torch.nn as nn import...

我尝试过,传入images为None或者[], 都会报错。完全没有办法吗?

I have been trying to understand this and only need to generate let's say 6 images. But I always end up getting shape invalid for input.

HI, can you provide the finetune coder for image generation ? Or the image process pipeline to get the image tokens for next token prediction loss calculation.

我需要在一个question中对多张图进行联合推理,以下是我的部分代码: content_images = [] for idx in range(len(image_files)): # content.append({"type": "image", "image": image_files[idx]}) content_images.append("data:image,{}".format(image_to_base64(image_files[idx], max_size=1024))) conversation = [ { "role": "", "content": f"\n{query}", "images": content_images, }, {"role": "", "content": ""},...

This Docker implementation makes it easy for people to run the server and a helpful web UI with a single command: 这个 Docker 实现让人们可以用一个命令轻松运行服务器和一个有用的 Web UI: ```bash docker run -it...