Unsloth and NVIDIA Team Up to Slash LLM Training Times
Unsloth's open-source framework dramatically accelerates LLM fine-tuning on NVIDIA GPUs, cutting training time and memor…
Latest articles in LLM News
Unsloth's open-source framework dramatically accelerates LLM fine-tuning on NVIDIA GPUs, cutting training time and memor…
LG AI Research releases EXAONE 4.0 as open-source, bringing multilingual AI capabilities to the global developer communi…
NTT Data unveils Tsuzumi 2.0, a lightweight Japanese LLM that outperforms GPT-4 on local language benchmarks while runni…
Open-source startup Unsloth partners with Nvidia to deliver 25% faster LLM fine-tuning on consumer-grade GPUs, democrati…
Samsung's new Gauss 3 on-device LLM brings native AI processing to Galaxy S26, eliminating cloud dependency for core AI …
Anthropic launches Dreaming for Claude Managed Agents, enabling AI to reflect, clean memory, and self-improve during idl…
Google introduces Multi-Token Prediction drafters for its Gemma 4 AI models, achieving up to 3x faster inference without…
Google's Gemini 2.5 Ultra achieves state-of-the-art results across major multimodal reasoning benchmarks, outperforming …
OpenAI unveils GPT-5 Turbo featuring native real-time video understanding, marking a major leap in multimodal AI capabil…
Hugging Face releases a massive 1 trillion token multilingual training dataset under an open license, democratizing larg…
Anthropic's Claude 4 sets new records on MATH and GPQA benchmarks, surpassing GPT-4o and Gemini Ultra in advanced reason…
Vietnam's VinAI Research releases PhoGPT-2, an open-source large language model optimized for Vietnamese and Southeast A…