tjake

Joined 18 years ago

Karma

915

Submissions

103

Member since

Mar 12

coder, dreamer...

http://twitter.com/tjake

View on Hacker News

Recent submissions

Embedding Models for Information Retrieval in 2025 datastax.com

12 tjake 1 year ago 1 comments

ColBERT-Live: ColBERT search on vector indexes with live updates datastax.com

1 tjake 2 years ago 0 comments

Apache Cassandra 5 Release cassandra.apache.org

4 tjake 2 years ago 1 comments

Comment · 2 years ago

YouTube interview with creators: https://www.youtube.com/watch?v=gJXZZJZc5gI

Open Source FPGA Expansion Card for the Apple II wiki.reactivemicro.com

99 tjake 2 years ago 17 comments

JLama: The First Pure Java Model Inference Engine Implemented with Vector API infoq.com

2 tjake 2 years ago 0 comments

Comment · 2 years ago

You can demo this here: https://jvectordemo.com:8443/

GH Project: https://github.com/jbellis/jvector

Comment · 2 years ago

it's 35M 1024 vectors Plus the text

Vector indexing all of Wikipedia on a laptop foojay.io

513 tjake 2 years ago 140 comments

Comment · 3 years ago

The default for `convert.py` is F32. This is just SIMD CPU comparison.

Jlama uses the vector api in java20 but also better thread scheduling with work stealing and zero allocation.

Comment · 3 years ago

GH: https://github.com/tjake/Jlama

Jlama (Java) outperforms llama.cpp in F32 Llama 7B Model twitter.com

7 tjake 3 years ago 9 comments

Show HN: Jlama – A fast Java inference engine for GPT and Llama models github.com

7 tjake 3 years ago 1 comments

Comment · 3 years ago

Missed AstraDB (Cassandra based). It ticks all the green boxes (HIPPA, SOC II, Cloud, in DB etc..)

https://docs.datastax.com/en/astra-serverless/docs/vector-se...

Vector Search for Apache Cassandra and AstraDB datastax.com

2 tjake 3 years ago 0 comments