Google and Blogger.com compliant news article
Introducing Llama 2: A Versatile Generative Text Model
Enhanced Capabilities and Open Source Availability
Google AI has announced the release of Llama 2, an advanced generative text model that offers a wide range of capabilities for researchers and developers alike.
Llama 2 is available in three model sizes (7B, 13B, and 70B) and comes both pretrained and fine-tuned, providing a flexible solution for various tasks. All three sizes are available for download from HuggingFace, the popular platform for machine learning models.
This release includes model weights and starting code for both pre-trained and fine-tuned Llama language models. Additionally, Google AI has provided an open-source notebook demonstrating how to fine-tune the Llama 2 model using QLoRa TRL and a Korean text classification dataset.
The versatility of Llama 2 allows for a range of applications, including text generation, machine translation, question answering, and dialogue systems. Researchers and developers are encouraged to explore the possibilities of this powerful language model.
Llama 2 is available for free for both research and commercial use. Google AI's commitment to open source ensures that this advanced technology is accessible to all.
For those eager to experiment with Llama 2, HuggingFace provides an easy-to-use interface for text generation inference on one's own infrastructure. Alternatively, HuggingFace also offers hosted inference services.
Comments