I don’t think that’s necessarily true! text is actually relatively few bytes, while each 32-bit float is 4 bytes, so at least at short lengths things could be lossless
CODE
-
Text Embeddings Lossy Collisions Detection Problem
By
–
yeah! but I don’t know if text embeddings are necessarily lossy? and if so how to locate collisions? (but yes; this is the entire point)
-
Building Multi-Agent Apps with LangGraph Gemini Pro Streamlit
By
–
LangGraph + Gemini Pro + Custom Tool + Streamlit = Multi-Agent Application Development Awesome tutorial (YouTube and blog) from Gao Dalie on how to build a multi-agent application with LangGraph and Gemini-Pro YouTube: https://
youtube.com/watch?v=FWBnNc
Zv3kw
… Blog: https://
levelup.gitconnected.com/langgraph-gemi
ni-pro-custom-tool-streamlit-multi-agent-application-development-79c1473086b8
… -
Cohere Rerank: Semantic Search for Legacy Retrieval Systems
By
–
You can use Rerank to infuse legacy retrieval systems with semantic search capabilities that increase the relevance of retrieved results.
— Cohere (@cohere) 10 février 2024
Check out Cohere's Senior Product Manager, Elliott Choi, explaining how a sample Rerank API call works. pic.twitter.com/jxxcqbNusHYou can use Rerank to infuse legacy retrieval systems with semantic search capabilities that increase the relevance of retrieved results. Check out Cohere's Senior Product Manager, Elliott Choi, explaining how a sample Rerank API call works.
-
LangChain 101: Document Loading, Splitting and RAG with LCEL
By
–
LangChain 101: Talking to Documents: Load, Split, and simple RAG with LCEL Loading documents and splitting them are a key part of RAG In his most recent update to the "LangChain 101" series Ivan Reznikov covers this in detail https://
pub.towardsai.net/langchain-101-
part-3a-talking-to-documents-load-split-and-simple-rag-with-lcel-26b005ccb30a
… -
Understanding Tensors: Fundamental Data Structure in Machine Learning
By
–
Tensor is a fundamental data structure in Machine Learning. I will clearly explain it today! Imagine you have pictures of cats. Each picture is made up of tiny pixels, and each pixel has some color values. Now, a tensor is a container that stores this data. TensorFlow,
-
Text Embedding Collision Discovery: Purple Repetition Experiment
By
–
I think I found the first evidence of a collision in text embeddings! i.e. for an embedding function f, i found a text pair x, y where x ≠ y but emb(x) = emb(y) ok granted it isn't thatttt interesting but what I did was repeat the word purple from 1 to 8191 times. 'purple ' is
-
GPU Throughput Increase with Multiple Processes Using MPS
By
–
With workloads that don’t fully load the GPU you can often get a substantial throughput increase by running multiple processes on each GPU, but you need to manually enable MPS to get concurrency: https://
docs.nvidia.com/deploy/mps/ind
ex.html
… -
Programs as coarse-grained projections of physical computational states
By
–
The program is not orthogonal to the physical state of the computer, it is a coarse grained projection of the physical state into a causal model