Our new fine-tuning technique at @everartai has improved so much in terms of output quality and generalization that I am afraid to release it publicly lol
LLMS
-
Surprising AI Capabilities: PDFs, Images, Code Execution
By
–
I totally agree with you. I guess, for me, it's more about the extent of the things they can do. Like, people may not be aware that they can "talk" to a PDF, replicate image styles, or run code natively, etc.
To this day, I still get surprised by the things it can do, so it's -
LangChain v0.1 Graph QnA: Learning LLM Integration
By
–
Graph QnA with LangChain v0.1 @tb_tomaz has some of the best content out there on how to use LLMs to interact with graphs Over the weekend, he updated all his content to use LangChain v0.1 Perfect opportunity to jump in and start learning https://
github.com/tomasonjo/blog
s/tree/master/llm
… -
Enterprise LLMs: Abacus.AI’s Open-Source Retrieval APIs Solution
By
–
Enterprise adoption of LLMs has exploded yet building custom LLMs on your own knowledge base remains challenging. At http://
Abacus.AI, we have our own open-source Retrieval APIs based on our enterprise AI platform. But here’s what makes us unique: •Choose your own LLM -
Building Question Answering Apps for Audio Data
By
–
Good quick overview of how to build a question/answering app over audio data!
-
CodeQAI: Local Code Search with Vector Embeddings and LLMs
By
–
👨💻codeqai
— LangChain (@LangChain) 15 janvier 2024
100% local code search and chat powered by vector embeddings and LLMs
Uses a lot of awesome tech to accomplish this!
🌲Tree sitter (for splitting code files)
🧱Sentence Transformers and Instructor embeddings
🌎FAISS
🦙Llama Cpp and @ollama https://t.co/AMxujuMSRG pic.twitter.com/WKKuK3LgALcodeqai 100% local code search and chat powered by vector embeddings and LLMs Uses a lot of awesome tech to accomplish this! Tree sitter (for splitting code files)
Sentence Transformers and Instructor embeddings
FAISS
Llama Cpp and @ollama https://
github.com/fynnfluegge/co
deqai
… -
5 ChatGPT hacks for easier learning and memorization
By
–
98% think ChatGPT can be used to skip learning. But no one mentions that ChatGPT makes learning easier. Here are 5 ChatGPT hacks to memorize everything:
-
GPT-4 Text Handling and Information Hiding Techniques
By
–
Part of the magic is that GPT-4 can read text written this way but in any other context you would need a decoder on the other side. And if you allow a decoder there are lots of ways to hide information in non-printing chars like ZWSP etc.
-
LLMs’ Value in Simplifying Technical Tasks like Matplotlib Usage
By
–
If LLMs achieve nothing but letting you use matplotlib without docs it will all have been worth it.