There are multiple variant Llama 2 models on Replicate (7B, 13B, 70B, chat vs base, etc), each with their own strengths. Check out the post to see which one is right for your needs. https://
replicate.com/blog/run-llama
-2-with-an-api#choosing-which-model-to-use
…
CODE
-
Choosing the Right Llama 2 Model Variant for Your Needs
By
–
-
Python developers can now install Replicate via pip
By
–
For you Python lovers, there's `pip install replicate`:
-
Installing Replicate with npm for Node.js
By
–
If you're using Node.js, then `npm install replicate` and you're off to the races:
-
Llama 2: First Open-Source Model Matching OpenAI’s Caliber
By
–
Llama 2 is the first open-source language model of the same caliber as OpenAI's models. You can run it on Replicate using JavaScript, Python, cURL, Go, Swift, or any other language. Here's a guide to get you started running Llama 2 with an API.
-
Next.js Template for SDXL App Launch
By
–
We made a @nextjs template for getting an SDXL app up and running in seconds. https://
github.com/replicate/gett
ing-started-nextjs
… -
Tree-of-Thought Reasoning Method Implemented in LangChain
By
–
Tree-of-Thought A new reasoning method, originally by @ShunyuYao12
, implemented in langchain_experimental by Vadim Gubergrits Paper: https://
arxiv.org/abs/2305.10601 Docs: https://
python.langchain.com/docs/modules/c
hains/additional/tot
… LangSmith Trace: https://
smith.langchain.com/public/cf896fa
9-0ec8-42e7-83a0-d1f9107b3aaf/r
… -
Programmatic Custom Model Creation Now Available in Python SDK
By
–
We are introducing programmatic custom model creation with our Python SDK. This empowers developers with a more flexible and efficient way to create custom models.
-
GenAI Models Creating Opinionated Markup Outputs
By
–
would be fun if a genai model could create opinionated markups like this
-
Tie-breaking fix improves nearest neighbor classification accuracy to 71.91%
By
–
Small update: In the orig code example in the paper, tie breaking was based on the lowest class label index, i.e., it was biased towards class 0. When I fixed the tie-breaking to select the closest neighbor, I could get a 70.05% -> 71.91% improvement (
https://
github.com/rasbt/nn_plus_
gzip
…). -
Shared Source Terminology: Microsoft Legacy and Perception
By
–
Don't use my words against me! (I do think "shared source" is fine. The downside to it is that its origin (under a less loved Microsoft) gave it a pejorative connotation. The denotation is fine. The connotation is soured. But maybe could be improved?