- python 3.8
- pytorch 2.01
- opencv-python 4.8.0.76
- spack 3.6.1
How to run:
- Select 4 or 9 portrait photos you like.
- Run the Main.py
- You will get the processed photo like the third photo in the workflow and the vivid text with emoji😀.
- Feel free to upload the content to Twitter or any social media platform.
Training
The new bert will be updating soon
Attention
We use a program I that designed(Accessing Mirror Sites with Crawlers) to enter a fake chatgpt to achive call LLM
Validation
In image caption, we use clip to compute a similarity between image and word to control the keyword generating.
- python 3.8
- pytorch 2.01
- opencv-python 4.8.0.76
- spack 3.6.1
How to run:
- Select 4 or 9 portrait photos you like.
- Run the Main.py
- You will get the processed photo like the third photo in the workflow and the vivid text with emoji😀.
- Feel free to upload the content to Twitter or any social media platform.
Training
The new bert will be updating soon
Attention
We use a program I that designed(Accessing Mirror Sites with Crawlers) to enter a fake chatgpt to achive call LLM
Validation
In image caption, we use clip to compute a similarity between image and word to control the keyword generating.
- Sentiment Analysis Net is on ./sentimentAnalysisNet, follow the instruction on the readme in the folder to train the net.
- Topic Categorization Net is on ./Bert_classification, follow the instruction on the readme in the folder to train the net.