I test the new #ChatGPTVoice: gadget or real revolution? → https://youtu.be/ENcsaIvQhAg
LLMS
-
Meta Leads Open Models While OpenAI Pursues Closed Approach
By
–
Who could have predicted years ago that OpenAI builds closed models and Meta, of all companies, releases industry-leading open models.
-
Programming Examples: Improving Pre-training Data Quality at Scale
By
–
Hacker News for AI research papers Programming Every Example: Lifting Pre-training Data Quality like Experts at Scale is right now
-
Fine-tune Llama 3.2 Models Locally and on Cloud
By
–
Here's how you can easily fine-tune latest llama 3.2 (1b and 3b) locally and on cloud:
-
Meta LLM Lawsuit Reaches Summary Judgment on Fair Use
By
–
The LLM lawsuit against Meta will reach 'summary judgement' on the Fair Use question soon. When the CEO says something like this at such an important phase, it's not a coincidence. This is the outcome that Meta has been lobbying, litigating, pressuring for — with all its power.
-
Zuckerberg’s Orion, OpenAI Changes, Tesla Robotaxi Launch
By
–
Prêt pour le FlashTweet ?
Mark Zuckerberg met Orion sur orbite
Mira Murati quitte OpenAI
Molmo, le LMM qui bat ChatGPT
Tesla Robotaxi : la super production d'Elon Musk Meta AI passe la barre des 500 M d’utilisateurs
C’est parti pour le #FlashTweet du 26/09/2024 -
Fine-tuning and Model Merging: Practical Tips and Tricks
By
–
Here's my @aiDotEngineer talk in SF about fine-tuning and model merging It's a quick overview of the field with practical tips and tricks to make the best merges and finetunes. Hope you'll like it! 🙂
-
AI Cambrian Explosion: VLMs, Multimodal Models, and Speech Interfaces
By
–
The Cambrian Explosion in AI this week: numerous VLMs beating SOTA, positive transfer from multimodal to text-only, chain-of-thought takes off, many new speech interfaces, …. more papers than one can read, and more AI people moving than one can keep track of! Whoever thinks
-
Document Store Vector Embeddings Synchronization for Retrieval
By
–
A great way of using document store to upsert vector embeddings, keeping everything in sync, and have updated information during retrieval querying 🔎 https://t.co/OgerYftAlj
— FlowiseAI (@FlowiseAI) 26 septembre 2024A great way of using document store to upsert vector embeddings, keeping everything in sync, and have updated information during retrieval querying
-
NVLM Big Models Comparison Results
By
–
Hi Jim, can you add the big models too for a full comparison. I love these results from NVLM: