From 5cbd3ce71197fa8c868d5ca87e6d79e04cc4267f Mon Sep 17 00:00:00 2001 From: Greg DeCarlo Date: Wed, 22 Nov 2023 16:14:36 -0500 Subject: [PATCH] Add custom GPTs --- README.md | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) diff --git a/README.md b/README.md index 7ea9c2c..1b4bdd9 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,10 @@ # gpt-spa (created by Greg DeCarlo - @mrinreality1 on X) A customizable GPT in a single page, using OpenAI models text-embedding-ada-002, tts-1, whisper-1, dall-e-3, and gpt-4-vision-preview. +### Create custom GPTs! +![image](https://github.com/scalable-dynamics/gpt-spa/assets/47045161/da15d179-7d37-4093-a014-03b630ffa89c) + +### Details - This example does not use the OpenAI Assistants API, but it does allow you to add files to a local vector db and perform retrieval (RAG) before interacting with the GPT - Image files can also be uploaded, which will be sent to the GPT 4 Vision model for analysis. Selecting 'Use DALL-E' allows the user to generate an image based on the current conversation (including document retrieval and images) - Selecting 'Use Speech' will speak the output using the new OpenAI Text-to-Speech model @@ -17,15 +21,15 @@ A customizable GPT in a single page, using OpenAI models text-embedding-ada-002, 3. Drag/Drop image files and ask your GPT about it 4. Use DALL-E and ask your GPT to generate an image (prompt will be generated using the context) 5. Use Speech and select a voice to hear your GPT outloud +6. [new!] Add multiple GPTs and choose between them +7. [new!] GPT builder interface for creating a GPT ### Upcoming Features -6. Talk to your GPT using Whisper Speech-to-Text transcription -7. Search Bing and include the results in the retrieval (RAG) -8. Have your GPT ask multiple-choice questions and suggest responses -9. Use Microsoft Graph API to send emails, schedule appointments and find people -10. Store memories and set reminders -11. Create charts, diagrams and tabular data using your GPT -12. Connect to external data sources which allow cross-origin resources (CORS) -13. Add multiple GPTs and choose between them -14. GPT builder interface for creating a GPT +8. Talk to your GPT using Whisper Speech-to-Text transcription +9. Search Bing and include the results in the retrieval (RAG) +10. Have your GPT ask multiple-choice questions and suggest responses +11. Use Microsoft Graph API to send emails, schedule appointments and find people +12. Store memories and set reminders +13. Create charts, diagrams and tabular data using your GPT +14. Connect to external data sources which allow cross-origin resources (CORS) 15. Calling functions to switch between models or change modality