Greetings fellow Bubble community!
We are thrilled to announce the launch of our new Trainer Plugin for GPT-3. This plugin is designed to help you train your own language models using OpenAI’s cutting-edge technology, directly from within your Bubble application. With our Trainer Plugin, you can now fine-tune GPT-3 to your specific needs, creating models that generate texts that are similar to what a human would write.
Demo:
https://gptuning.bubbleapps.io/
Our Trainer Plugin is user-friendly and easy to use, with customizable settings for temperature, maximum number of tokens, and other parameters. With the ability to list and retrieve models and files, make requests, upload and delete files, fine-tune your models, and more, you have everything you need to bring your creativity to life.
By using the Trainer Plugin, you can save time and effort, as you no longer have to go through the complex process of fine-tuning GPT-3 manually. With just a few clicks, you can create and use custom language models that meet your specific requirements, and the possibilities are endless!
So why wait? Give it a try today and discover the incredible potential of GPT-3. We are confident that our Trainer Plugin will help you achieve your creative and business goals faster and more efficiently.
API Calls:
Data:
- List models
- Retrive models
- List files
- Retrive file content
- List fine-tunes
- Retrive fine-tune
- List fine-tune events
Actions:
- Make request
- Upload file
- Delete file
- Retrive file
- Create fine-tune
- Delete fine-tune model
P.S. At the moment, we are waiting for approval from Bubble to publish the plugin page link. Stay tuned and don’t miss the chance to experience the amazing features of the GPT-3 Trainer and simplify the Fine Tune process.
5 Likes
Great news! Our GPT3 Trainer plugin is now available on the Bubble io marketplace! This powerful tool allows you to fine-tune and customize your GPT-3 models for your specific needs. With the GPT3 Trainer, you can easily train your models with your own datasets, and improve their accuracy and relevance for your application.
Whether you’re building a chatbot, a virtual assistant, or any other AI-powered application, the GPT3 Trainer will help you get the most out of GPT-3’s impressive capabilities. So why wait? Try the GPT3 Trainer today and take your AI development to the next level!
1 Like
@leomdsanchez Looks promising. Do you have any documentation on how we can use this effectively? I’m ChatGPT through it’s API on handling prompts using text-davinci. Fine tuning is something I’m currently trying to learn.
3 Likes
Sure, I can help. I have prepared a step-by-step tutorial on how to perform the Fine Tune process on an OpenAI model using the Bubble io plugin. This tutorial includes how to prepare the data, upload the file, create the Fine Tune process, check the status of the process, and make requests using the Fine-tuned model. You can find the tutorial at this link: GPT-3 Trainer
1 Like
@leomdsanchez could I pls ask for an ELI5 explanation of how this works 
This is my current understanding: ChatGPT cannot access any given database at the moment without a proper OpenAI plugin (i.e. those featured here). To get around this, you’ve created a fine-tuning tool that can feed ChatGPT anyone’s Bubble data AND produce output based on that.
Am I off or 80/20 there??
1 Like
Thank you for your question. The explanation you provided is partially correct. I created a fine-tuning tool for ChatGPT that allows it to be trained with any dataset and generate more accurate and customized responses based on that data. While ChatGPT doesn’t have direct access to any specific database, it can be fed with data from any source through my fine-tuning tool. I hope this clears up any confusion you may have had.
1 Like
Hi There,
Looks like a really great plugin, I’m trying to understand how this would work though, like how is the front-end look, example, I have chatgpt now installed how would I use this plugin to train my chatgpt ?
Hi, does this work with 3.5?
GPT3.5/4 (ChatGPT) do not support fine tuning. Remember fine tuning is only useful for a very small set of use cases.
2 Likes
@leomdsanchez are you still working on this plugin with an update?
