Posts tagged Benchmarks collect Will Spurlock's writing on this topic. The three newest excerpts: Mistral Small 3 24B Apache-2: The Open SLM That Beats Llama 3.3 70B on SpeedOn January 30, 2025, Mistral AI dropped what I consider one of the most strategi Cerebras just launched hosted inference on CS-3: up to 1,800 tok/s on Llama 3.1 8B and 450 tok/s on 70B per its Aug 27 release, with Artificial Analysis quoting 1,800+ and 446+ output tok/s respectively. xAI ships Grok-2 and Grok-2 mini in beta—the same stack that hit Chatbot Arena as sus-column-r. Here is what the LMSYS Elo story actually says next to the spreadsheet benchmarks.
5 posts as of 2025-01-30, counted from content/blog frontmatter
Discover insights and strategies for leveraging technology in business
Posts tagged Benchmarks collect Will Spurlock's writing on this topic. The three newest excerpts: Mistral Small 3 24B Apache-2: The Open SLM That Beats Llama 3.3 70B on SpeedOn January 30, 2025, Mistral AI dropped what I consider one of the most strategi Cerebras just launched hosted inference on CS-3: up to 1,800 tok/s on Llama 3.1 8B and 450 tok/s on 70B per its Aug 27 release, with Artificial Analysis quoting 1,800+ and 446+ output tok/s respectively. xAI ships Grok-2 and Grok-2 mini in beta—the same stack that hit Chatbot Arena as sus-column-r. Here is what the LMSYS Elo story actually says next to the spreadsheet benchmarks.
5 posts as of 2025-01-30, counted from content/blog frontmatter.
Start with "Mistral Small 3 24B Apache-2: The Open SLM That Beats Llama 3.3 70B on Speed". Mistral Small 3 24B Apache-2: The Open SLM That Beats Llama 3.3 70B on SpeedOn January 30, 2025, Mistral AI dropped what I consider one of the most strategi
