2/ As many enterprises are finding out, real AI use cases often require specialist models fine-tuned on domain-specific data, and tunable “base” models like Llama 2 are the foundation for this.
GENERATIVE AI
-
Llama-v2 Model Excels at Question Answering Tasks
By
–
Biggest use case we see is question answering over your data
— Harrison Chase (@hwchase17) 19 juillet 2023
The new Llama-v2 model is pretty good at that! As shown by @RLanceMartin
Check out the trace: https://t.co/tZY4wrkzYB https://t.co/BKrpdTyktzBiggest use case we see is question answering over your data The new Llama-v2 model is pretty good at that! As shown by @RLanceMartin Check out the trace: https://
smith.langchain.com/public/6128653
f-1857-488b-b02d-4f4b19b555ee/r
… -
Open Source LLM Tools: Text Generation Inference and Chat UI
By
–
Powered by: Text-generation-inference, production-ready tool for LLM serving – https://
github.com/huggingface/te
xt-generation-inference
…
Chat UI, open source LLM UI, – https://
github.com/huggingface/ch
at-ui
…
Llama 2 Models – https://
huggingface.co/meta-llama -
Llama 2 70B Chat Now Available Free on Hugging Chat
By
–
Llama 2: Now on Hugging Chat 🤗🦙
— Hugging Face (@huggingface) 19 juillet 2023
Try out the 70B Chat model for free with super fast inference, web search, and powered by open-source tools!
👉 https://t.co/TQ0ZaVZcdi pic.twitter.com/AnD2eknCMtLlama 2: Now on Hugging Chat Try out the 70B Chat model for free with super fast inference, web search, and powered by open-source tools! http://
hf.co/chat -
Databricks acquires MosaicML to advance generative AI
By
–
Welcome to Databricks, @MosaicML
! #GenerativeAI will drive the next enterprise #data app wave — and we’re excited to work together to accelerate developments and make generative AI accessible to all organizations. Hear what's next from our founders! -
AI as Tools: Choosing the Right Solution for Enterprise Problems
By
–
Lastly, it’s important to note that #AI is a bag of tools for a bag of problems, knowing what tool to use to solve what problem makes all the difference! here is a framework to think about #GenerativeAI for enterprises: h/t Alex Ratner
-
ChatGPT Performance Degradation Possibly Intentional Fine-Tuning
By
–
At least one area “Answering sensitive questions'' performance degradation may be due to purposely fine-tuning! The reality is that nobody can replicate or validate these results (you can’t unless you are at OpenAI and have access to the March version of ChatGPT)
-
ChatGPT Performance Decline Paper Sparks Debate Among Experts
By
–
This “#ChatGPT is losing it..” paper got all the views and highlights https://
lnkd.in/gSx95EUB. Seems to be mostly from the doomers, and the experts pushing their agendas (because they build LLMs). We still don't know what is happening here. -
LLMs as Tools: Building Enterprise-Grade AI Systems
By
–
LLMs are one of the many tools you have in #AI to solve real problems. These models are powerful next word predictions and are optimized for human-like conversational experiences, but in an enterprise context, you have to actually build enterprise grade systems
-
Generative AI Data Causes Model Autophagy Disorder MAD
By
–
Generative AI generated data can make #GenerativeAI models go MAD (Model Autophagy Disorder) – super cool paper to review: https://
lnkd.in/gN8QuxW7. It underscores the importance of Data and domain expertise in the age of #AI.