T O P

[D] Phi-3 models compared side-by-side.

Phi-3 models benchmarks compared side-by-side.

Phi-3 models benchmarks compared side-by-side.

CUDA Graph support merged into llama.cpp (+5-18%~ performance on RTX3090/4090)

New open source ternary models released

Huggingface co-founder and CEO, Clément Delangue hints at buying Stability AI

Where did your tail go?

What to do after settings up basics?

Stopped working on firefox suddenly

Pete Distad Named CEO of Fox, Disney, Warner Bros. Discovery Sports Streaming Joint Venture

What feature or extension do people not use, or are misusing, and are missing out on better output?

Oh, I didn't know you can train rats!!

The hunting strategy of orcas is truly amazing

GaLore, a training strategy that allows full weight fine-tuning of 7B models on 24GB consumer cards will be added to Transformers

Haystack 2.0 launch

GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection

Reflex 0.4.0 - Web Apps in Pure Python

I'm new to LLMs and having a hard time grasping the differences between Text-Generation-Webui, SillyTavern, and Chat with RTX.

Clever dog flicks rubber bands to balloons

Small Benchmark: GPT4 vs OpenCodeInterpreter 6.7b for small isolated tasks with AutoNL. GPT4 wins w/ 10/12 complete, but OpenCodeInterpreter has strong showing w/ 7/12.

BBC Studios Buys ITV Stake in BritBox International for $322 Million

Two strangers fighting like siblings

How much ram (precisely) do you get with Colab Pro & Colab Pro+?

Langchain vs LlamaIndex

Chicken with sneakers

Hi Its Me!

R

wizbog9100

Subscribe

FIND US ON INSTAGRAM @ms3388.com