
The Technology Innovation Institute (TII) in Abu Dhabi has unveiled the Falcon Mamba 7B, a groundbreaking artificial intelligence model. This model is notable for being the world's first attention-free AI model, utilizing a state space architecture. It has been trained on 5.5 trillion tokens and features 7 billion parameters. The Falcon Mamba 7B sets a new benchmark in AI research, outperforming several established models such as Llama 3 8B, Llama 3.1 8B, Gemma 7B, and Mistral 7B in various benchmarks. It is open-source, permissively licensed, and can process unlimited sequence lengths, fitting on a single 24GB GPU. The model is also recognized for its constant time to generate new tokens regardless of context size, and its impressive 8,192 context length. The Falcon Mamba 7B is available on the Open LLM Leaderboard, scoring 15 points on average.


📢 4-bit Llama 3.1 405B, 70B, 8B Now Available! 📢 @AIatMeta's Llama 3.1 models are now quantized to 4 bits by @neuralmagic's research team and available with ~100% recovery. These enable 4X cheaper deployments (405B goes from 2 8x80GB nodes to 1 4x80GB). Continued in next... https://t.co/RkqUzyeDfY
Falcon Mamba 7B’s powerful new AI architecture offers alternative to transformer models #FalconMamba #AI #TechAI #LearningAI #GenerativeAI #DeepbrainAI #LLM #LM #ML #ArtificialIntelligence https://t.co/IzpKHl4J45
Falcon Mamba 7B’s powerful new AI architecture offers alternative to transformer models: In several benchmark, Falcon Mamba 7B convincingly outperformed Llama 3 8B, Llama 3.1 8B, Gemma 7B and Mistral 7B. https://t.co/SnR8jqqvfM #AI #Business