Wow, this whole thread is awesome Data ecosystem on HF
OPEN SOURCE
-
DeepSeek Coder V2 Matches Best Closed-Source Code Models
By
–
New chart with Code LLMs Based on the BigCodeBench Leaderboard, it shows the impressive performance of DeepSeek Coder V2 (236B, 21B active), which is on par with the best closed-source models. Personally, I've switched to Sonnet and I'm very happy with its code performance.
-
Yi LLM Reaches 7.4K Stars on GitHub
By
–
Yi now has 7.4K stars and 454 forks on @github ! We love seeing all of the amazing things being built with our LLMs. Explore the Yi models and share your work with us:
-
Delta Lake UniForm GA: Unified Data Access Across Tools
By
–
Delta Lake UniForm is now GA! UniForm combines Delta Lake’s superior price-performance with access to every tool in your stack. Write a single copy of your data and make it available to any engine that supports the primary open table formats. https://
dbricks.co/3XSy9q3 -
Paints Undo: AI Image Editing Tool Project Exploration
By
–
Si queréis cotillear un poquito más en el modelo, esta es la web del proyecto. https://
lllyasviel.github.io/pages/paints_u
ndo/
… -
OSS Advocate Testifies at U.S. Generative AI Hearings
By
–
Yeah, there was an OSS advocate who said exactly that in the U.S. hearings about generative technology.
-
Open-Source App Tracks Maps Animal Movement Using AI
By
–
This Open-Source App Tracks and Maps Animal Movement!
— MathWorks (@MathWorks) 9 juillet 2024
🔗 https://t.co/6folFY9nK6 pic.twitter.com/hGgRsigpJHThis Open-Source App Tracks and Maps Animal Movement! https://
spr.ly/60159Y4BS -
Q2_K Quantization Achieves FP32 Base Model Performance
By
–
Q2_K quantized model meets its FP32 base
-
Keras and JAX: Combining Simplicity with Performance for Deep Learning
By
–
Keras = simplicity & productivity
JAX = performance & scalability The combination of the two gives you an unfair advantage for developing deep learning solutions. -
Claude Sonnet 3.5 enables BabyAGI implementation without frameworks
By
–
Apparently Claude Sonnet 3.5 can code you your own implementation of BabyAGI Starting here and iterating on this with Claude should be a good start/exercise for anyone looking to build an autonomous agent without relying on external frameworks.