#Latxa txatbota probatarako erabilgarri jarri dugu! latxa.hitz.eus

Jaso ditugun eskaerei erantzunez zuen eskura jarri dugu Latxaren bertsio ahaltsuena, chatGPT-tik gertu dabilena, baina euskara txukunagoa sortuz.

A fun read on Deep RL from Alex Irpan.
www.alexirpan.com/2018/02/14/r...

In summary, the infra layer is eating the model layer, the model layer is eating the application layer, the application layer might soon eat the people layer. AI food chain.

« appending "Wait" multiple times to the model's generation » is our current most likely path to AGI :)

See the fresh arxiv.org/abs/2501.19393 by Niklas Muennighoff et al.

Cohere's Maya: Open-source (code, dataset, and model) Multimodal Multilingual LLM

- Paper: Maya: An Instruction Finetuned Multilingual Multimodal Model ( arxiv.org/abs/2412.07112 )
- Model and Dataset: huggingface.co/maya-multimo...
- Repo: github.com/nahidalam/maya

Have you ever wondered how to train an autoregressive generative transformer on text and raw pixels, without a pretrained visual tokenizer (e.g. VQ-VAE)?

We have been pondering this during summer and developed a new model: JetFormer 🌊🤖

arxiv.org/abs/2411.19722

A thread 👇

1/

I wrote a long note on LLaMA 3 based speech models.

get it while it’s hot #ai

go.bsky.app/Pxcnfu6