https://github.com/zsxkib/replicate-emotion2colour
text -> emotion -> colour
https://github.com/zsxkib/replicate-emotion2colour
Last synced: 6 months ago
JSON representation
text -> emotion -> colour
- Host: GitHub
- URL: https://github.com/zsxkib/replicate-emotion2colour
- Owner: zsxkib
- License: mit
- Created: 2023-06-05T19:45:19.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-06-07T18:21:13.000Z (over 2 years ago)
- Last Synced: 2025-04-01T13:08:36.827Z (6 months ago)
- Language: Jupyter Notebook
- Size: 13.7 KB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# **Emotion Visualization: Paint Your Mood with Color!** 🌈
Transform a sentence into vibrant colors representing emotions with our tool. Let us guide you through the journey from text to radiant color gradients. 🎨
## The Mechanics of It All 🤔
Our Python-based Cog Predictor effortlessly converts your textual input into captivating color gradients. Here's how it works:
- Your text is received, no matter its content.
- Our model, [`distilroberta-base` by `j-hartmann`](https://huggingface.co/j-hartmann/emotion-english-distilroberta-base), adeptly uncovers the emotional sentiments within your text.
- These detected emotions are then converted into colors, forming a visual representation as intricate as your feelings.
## Translating Emotions into Radiant Gradients 🎨
The model recognizes a range of emotions: anger 🤬, disgust 🤢, fear 😨, joy 😀, neutral 😐, sadness 😭, and surprise 😲. Each emotion corresponds to a unique color, forming a comprehensive palette to paint your emotional landscape. The intensity of the detected emotions influences the transition points within the gradient. The stronger the emotion, the more it shapes the color scheme.
## Your Journey to Emotion Visualization 🚀
### Building with Cog 🛠️
Before you start, ensure you have [Cog](https://github.com/replicate/cog) installed. Once that's done, building with Cog is a breeze.
Just run the following command in your terminal:
```zsh
cog build
```
### Executing a Prediction 🏃♀️
For illustration, let's explore how a prediction is executed. With an input like `"I'm so mad, I don't know why"`, we create this eye-catching gradient:
```
cog predict -i text="I'm so mad, i don't know why"
```
The script provides a detailed output:
```zsh
Running prediction...
================================================================================
sorted_emotions=[{'label': 'anger', 'score': 0.6365653872489929}, {'label': 'surprise', 'score': 0.22874169051647186}]
(#FF0000) top_emotion='anger'
(#FFA500) bottom_emotion='surprise'
================================================================================
Written output to output.png
```
As you can see, your part is easy: provide the text and watch the machine learning magic happen!
Once the prediction is complete, you'll find an image titled `output.png` in your directory, reflecting the emotional sentiment of your input text as a radiant gradient.
With this Replicate cog-example, we aim to bridge the gap between machine learning and user-friendly application, providing you with a fun and captivating way to visualize textual emotions as color gradients.