michabbb<p><a href="https://social.vivaldi.net/tags/Meta" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Meta</span></a> releases <a href="https://social.vivaldi.net/tags/Llama3" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Llama3</span></a>, (Llama 3.2) the next iteration of open-source <a href="https://social.vivaldi.net/tags/LLMs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLMs</span></a>:</p><p>• 🖼️ <a href="https://social.vivaldi.net/tags/Multimodal" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Multimodal</span></a> models: 11B & 90B sizes with vision capabilities for tasks like visual reasoning & document QA<br>• 💻 Small on-device models: 1B & 3B text-only versions for efficient deployment<br>• 🌐 <a href="https://social.vivaldi.net/tags/Multilingual" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Multilingual</span></a> support: 8 languages for text-only prompting<br>• 📏 128k token context length for all models<br>• 🛡️ Updated <a href="https://social.vivaldi.net/tags/LlamaGuard" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LlamaGuard</span></a>: New 1B version for content moderation</p><p>Key features:<br>• 🧠 Strong performance on benchmarks like MMMU, VQAv2, DocVQA<br>• 🔧 <a href="https://social.vivaldi.net/tags/Huggingface" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Huggingface</span></a> <a href="https://social.vivaldi.net/tags/Transformers" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Transformers</span></a> & TGI integration<br>• ☁️ Deployment options: Inference Endpoints, <a href="https://social.vivaldi.net/tags/GoogleCloud" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GoogleCloud</span></a>, <a href="https://social.vivaldi.net/tags/AmazonSageMaker" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AmazonSageMaker</span></a>, <a href="https://social.vivaldi.net/tags/DELL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DELL</span></a> Enterprise Hub<br>• 🔬 Fine-tuning support with TRL and PEFT</p><p><a href="https://social.vivaldi.net/tags/opensource" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>opensource</span></a> <a href="https://social.vivaldi.net/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://social.vivaldi.net/tags/machinelearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>machinelearning</span></a> <a href="https://social.vivaldi.net/tags/NLP" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>NLP</span></a> <a href="https://social.vivaldi.net/tags/computervision" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>computervision</span></a></p><p><a href="https://huggingface.co/blog/llama32" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">huggingface.co/blog/llama32</span><span class="invisible"></span></a></p>