Llama got its worthy opponent!

In few tech revolutions, GenAI and NLP Language models are enhacing exponentially. At starting stage, GPT,BERT like models got viral and analysed and reviewed by various Data Science Experts.Definitely,These models aren't built with ABCD's . These are trained over billions of unique NLP dataset with thousands of parameters which improves efficiency of them.Meta's Llama ranked a major place in the improvement of language models.

After GPT-3 , Transformers architecture become state of art architecture that had greater accuracy and a descent choice for language models. Google AI unveiled PaLM, a 540 billion parameter transformer-based model. This immense size placed it among the largest language models at the time.

Meta holds it gear up and developed Llama which is a autoregressive large language model.Meta AI released the first version of Llama AI, with four model sizes (7B, 13B, 33B, and 65B parameters). Initially, it was available only for research purposes with a non-commercial license.

Gradually Llama beats other language models and become the state-of-the-art large language model. It achieves good accuracy and compared to other models it feels faster and low violation with safety. Newly launched Llama-2 also performs well and its open source!....

Recently, Google launches its open source Large language model which far enough to compete Llama-2,which is also a open source model.According to google,Gemma is more efficient than Llama-2. Gemma is a family of lightweight, state-of-the-art open models built from the same research and technology used to create the Gemini models.

Google focused the previous fall about the density of the model and developed Gemma as a light weight model that can compete in state-of-the-art models.We shall wait for further enhancments in AI.If you wish to take a glance on these models,both are available in huggingface .You can also visit in kaggle at models section like here .