Google made massive improvements to Bard, with Gemini Pro now ranking higher than regular GPT-4. Keep in mind, Gemini Pro is Google's SECOND best model. We're still waiting on Gemini Ultra. Expect OpenAI to drop a new model anytime now to stay in the lead.
LLMS
-
Building End-to-End Web RAG Chatbot with LangChain
By
–
Building a web RAG chatbot Our newest YouTube video on building with LangChain features how to build and end-to-end app (notebook to hosting) for a web RAG chatbot Video: https://
youtu.be/dA1cHGACXCo To celebrate the update of @exaailabs (prev. Metaphor), we've teamed up to show -
No, ChatGPT hasn’t become bad; the problem might be users
By
–
No, ChatGPT hasn't become BAD! I come back to the video of
@Micode and @UnderscoreTalk to give you my opinion →
https://
youtu.be/RFElI18_fac And what if the problem wasn't GPT-4 but the users? #ChatGPT #GPT4 -
LangSmith JavaScript Evals for LLM Application Evaluation
By
–
LangSmith Evals in JavaScript LangSmith now supports a native `runOverDataset` method for evaluating your JavaScript LLM applications! You can configure custom or off-the-shelf evaluators to compare + score the the outputs of your app against ideal reference outputs.
-
LLM Inference Margins Race to Bottom No Sustainable Moat
By
–
Great LLM Inference benchmark — all the right variables are in there.
— Soumith Chintala (@soumithchintala) 26 janvier 2024
Now, the race to the bottom begins, because in LLM Inference there isn't a sustainable long-term technology or software moat. IMHO the margins are going to resemble grocery stores and laundromats…
Some of… https://t.co/pR5xVIpik3Great LLM Inference benchmark — all the right variables are in there.
Now, the race to the bottom begins, because in LLM Inference there isn't a sustainable long-term technology or software moat. IMHO the margins are going to resemble grocery stores and laundromats…
Some of -
Embeddings Training Specificity and Cross-Model Compatibility
By
–
well they train this in; if you try this with other embeddings, it won’t work, for more or less the reasons the guy is saying in the middle
-
LangChain Releases LangGraph Updates and New astream_events Method
By
–
New release notes! Catch up on everything LangChain related over the past two weeks, including: LangGraph
LangChain v0.1
LangSmith updates
New astream_events method
Revamp of key use case pages! Blog: https://
blog.langchain.dev/week-of-1-22-2
4-langchain-release-notes/
… -
SOLAR-10.7B-Instruct Model Available via Together Compute
By
–
If you'd like to quickly check out SOLAR-10.7B-Instruct-v1.0, try using @togethercompute playaround and endpoint. They're really good! https://
api.together.xyz/playground/cha
t/upstage/SOLAR-10.7B-Instruct-v1.0
… -
Mental Health App Built as LLM State Machine
By
–
Mental Health Therapy as an LLM State Machine In this new guest blog, @Sonia_Health shares more about how they are developing their app Excited to highlight this for two reasons: App with really positive societal benefit
Built as a state machine – which matches our -
LangChain Hacking Hours Event – Meet the Team
By
–
Friday Fun Time! Come join us for hacking hours later today! Meet the team, meet others building in the space, and build some cool stuff with LangChain, LangGraph, LangSmith, LangServe, Lang?? RSVP: