Type something to search...
to navigateto selectESC to close

Parameters

Mini MiniCPM-o 2.6: The 8B Parameter Multimodal LLM Beating GPT-4o

Mini MiniCPM-o 2.6: The 8B Parameter Multimodal LLM Beating GPT-4o

In a groundbreaking development, Mini CPM-o has taken the world of multimodal large language models (LLMs) by storm. With its 8-billion parameter architecture, it not only outperforms GPT-4o on

Read More
DeepSeek: The Chinese Rival to ChatGPT Aiming for AI Market Leadership

DeepSeek: The Chinese Rival to ChatGPT Aiming for AI Market Leadership

DeepSeek V3, China’s bold AI model, challenges GPT-4 with 671B parameters, cost-efficient training, and innovation under U.S. sanctions.Ali Shaker- The Chinese startup DeepSeek has

Read More
Dramatically Reduce Inference Costs with DeepSeek-V3: A New Era in Open-Source LLMs

Dramatically Reduce Inference Costs with DeepSeek-V3: A New Era in Open-Source LLMs

Introduction DeepSeek-V3 has emerged as the new heavy weight for open-source enthusiasts and enterprise users alike. Developed by a Chinese AI research company with a commitment to an

Read More
DeepSeek V3: The best Open-source LLM | by Mehul Gupta | Data Science in your pocket | Dec, 2024 | Medium

DeepSeek V3: The best Open-source LLM | by Mehul Gupta | Data Science in your pocket | Dec, 2024 | Medium

Better than Claude 3.5 Sonnet, GPT-4o, Llama3.1 405B The year is about to end and just now, China’s DeepSeek has released its open-sourced model DeepSeek-v3, which has outperformed al

Read More
Meta’s Llama 3.3: The Evolution of Open-Source Large Language Models

Meta’s Llama 3.3: The Evolution of Open-Source Large Language Models

Meta’s recent release of Llama 3.3 represents a milestone in the development of large language models (LLMs). It introduces improvements in scale, efficiency, and safety, while remaining open

Read More
Alibaba QwQ: Better than OpenAI-o1 for reasoning?

Alibaba QwQ: Better than OpenAI-o1 for reasoning?

32b open-sourced model beats o1 mini and competes with o1-preview A few days back, Alibaba released Marco-o1, a 7b reasoning model. Now, they have released another, improved version cal

Read More
The Most Ambitious AI Crypto Project Ever is Here

The Most Ambitious AI Crypto Project Ever is Here

AI & Blockchains: A Match Made in Heaven, or a Scam? One of the founding fathers of modern AI wants to use the blockchain to train the world's largest open-source Large Language Model (LLM),

Read More
Meet Qwen2.5-Coder-32B-Instruct -Coder -open source better than gpt4o.

Meet Qwen2.5-Coder-32B-Instruct -Coder -open source better than gpt4o.

Meet Qwen2.5-Coder-32B-Coder, Your New AI Coding Buddy Have you ever wished that coding was a little easier, faster, and maybe even more fun? So, prepare to meet your new AI coding friend, Qw

Read More
SmolLM2: Very Good Alternatives to Qwen2.5 and Llama 3.2

SmolLM2: Very Good Alternatives to Qwen2.5 and Llama 3.2

And it's fully open! Hugging Face has doubled down on their SmolLM initiative. They released SmolLM2: 1.7B, 360M, and 135M models trained on 11T tokens (against 1T for SmolLM). They release

Read More
Google Releases Gemma — A Lightweight And Open Source Model

Google Releases Gemma — A Lightweight And Open Source Model

In just a week, the world has witnessed the most groundbreaking AI advancements from two tech giants. OpenAI introduced its jaw-dropping AI video generator, [Sora](https://readmedium.com/3d1638

Read More