Meta released Llama 3.2, a family of vision-capable large language models and lightweight text models for edge devices. The new lineup includes 11 billion and 90 billion parameter multimodal models that can reason about images, outperforming Claude 3 Haiku and GPT4-mini on visual understanding tasks. Meta also launched 1 billion and 3 billion parameter models optimized for on-device use with 128K token context lengths, with the 3B model outperforming Gemma 2 2.6B and Phi 3.5-mini on tasks like instruction following and summarization. The company also introduced Llama Stack distributions to simplify deployment across various environments, and updated safety tools, including Llama Guard 3 for vision tasks. (Meta AI)
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| ✨ New course! Enroll in Introducing Multimodal Llama 3.2 | 4 | 389 | October 11, 2024 | |
| Meta shrinks Llama models for faster on-device AI | 1 | 68 | October 28, 2024 | |
| Runaway LLaMA: How Meta's LLaMA NLP model leaked | 1 | 76 | May 20, 2023 | |
| Llama 3.2 finetuning and evaluations? | 6 | 73 | October 18, 2024 | |
| Llama3.2 from Huggingface in Google Colab | 6 | 169 | November 7, 2024 |