eduardo

macro photo of a fly things we share our food with...

Found this bright yellow spotted beetle in my backyard. Lens: Canon RF 100mm f/2.8 #macrophotography

At least 2,000 people are watching this "Castell" human tower event in Spain. #cv

- how many people are there in this image? - at least 1282 people. shoutout to @ediardo for building this awesome supervision demo supervision repo: github.com/roboflow/super…

football AI code is finally open-source - player detection and tracking - team clustering - camera calibration I still need to work on README; don't judge me on that code: github.com/roboflow/sports

@c_valenzuelab ¡Felicidades al equipo!

Nothing great happens overnight. Runway is turning 6 years old in October, and every year we have made continuous progress toward our long-term goal of creating a new art form. There is still a long way to go.

How does the brain interpret what we see? Nobel Prize laureates Torsten Wiesel and David Hubel clarified how this process works during the 1960s. In 1981, they were awarded the medicine prize: bit.ly/2XSL3Ce

Getting unstuck is an intense, yet delightful moment. Yesterday I had no idea wtf was going on. Today, I still don't know wtf is happening, but it's a completely new challenge. :')

Grabando a Spiderman grabando a Spiderman admirando el atardecer en el faro de Mazatlán.

@SpicyCyberLab @LMStudioAI Desde la superficie, es evidente que son la misma familia de LLMs porque las respuestas comienzan de manera muy similar. No he encontrado alguna diferencia importante entre distintos quants. Me gustaría evaluarlos sistemáticamente para entender mejor las capacidades de cada uno.

@ediardo @LMStudioAI Cual es el resultado de las pruebas?

Correr LLMs en localhost es alucinante!🤩 Experimentando con 5 LLMs (Llama 3 8B) usando diversos niveles de quantización en @LMStudioAI para explorar respuestas de modelos fácilmente.

If I had 10 tries, I couldn't have timed this hyperlapse more perfectly than this. Because of the clouds obscuring the corona, this is probably the best shot in all of Mazatlan of the event. Mazatlan eclipse.

@LMStudioAI Great update. Looking forward to prompting multiple LLMs locally.

Today we're excited to introduce Multi-Model Sessions! — Load & prompt multiple local LLMs simultaneously — Use the API to build networks, pipelines of LLMs that can talk with each other — Use the new JSON Mode to constrain each model's output format LM Studio 0.2.17 is available for Mac, Windows, and Linux. Get it here: lmstudio.ai

Corriendo Llama 2 70B Q4 en una Macbook Pro M3 Max con 64 GB RAM. Algunas generaciones van a > 7 tokens/segundo. Básicamente, esta es una laptop que podría programarse así misma a travez de código.

@iamcurib Muy bien. Ahora, para mayor tropicalización, usar wey.

Tropicalizando a mi compañero argentino. Ya dice güey! 😂

.@AIatMeta's codellama 70B is here! It achieves 67.8% on HumanEval which reaches the initial GPT-4 performance 🤯🤯 To run it: ollama run codellama:70b Learn more: ollama.ai/library/codell…

🌉🌉 Ferry Sea lions AI meetup Night walk Ferry

@rauchg Es fácil caer en la trampa del "solo falta un feature más para..."

If you’re building a better product with fewer features, be patient. “Multiple studies have shown that before using a product, people judge its quality based on the number of features it has. It's only after using the product that they realize the usability issues too many features create” lukew.com/ff/entry.asp?2…

Mi cara después del primer commute en el Ferry entre Oakland y SF. 🌉🌃⛴️

The shimmering lights of San Francisco's skyline from the shores of Alameda.