Falcon-180B Takes Open Source LLMs Closer to GPT-4
Just a few months ago, The Technology Innovation Institute (TII) in the United Arab Emirates (UAE) shook the world of foundation models with the release of Falcon LLM, setting a new benchmark for open-source AI Falcon-180B, the latest masterpiece from TII, is here to rewrite the rules once again! This colossal model boasts a jaw-dropping 180B parameters, making it a true game-changer in the world of AI. Trained on a staggering 3.5 trillion tokens, utilizing 4096 GPUs and a whopping 7M GPU hours, Falcon 180B stands as a testament to human ingenuity and the limitless potential of open-source AI It is touted as the "Llama 2" killer due to its higher performance as a pretrained-only model. As of September 2023, Falcon 180B ranked as the highest-performing pretrained LLM on the Hugging Face Open LLM Leaderboard . The model is big. Inference requires 640GB of memory — a mere eight A100 80GB GPUs — when quantized to half-precision (FP16). Alternatively, we can quantize down to int4,