Last mile quality & robustness (the high-hanging fruits) is going to be the test bed for enterprise LLM players in 2024.

e/ia - Intelligence Amplification - Does not seek to build superintelligent God entity that replaces humans. - Builds “bicycle for the mind” tools that empower and extend the information processing capabilities of humans. - Of all humans, not a top percentile. - Faithful to

🤖 It was a great conversation with Ben Ben Lorica 罗瑞卡 to exchange thoughts on GenAI copilot & agents for cybersecurity and enterprise workflow in general! ⛰At @NexusflowX , we are very excited to build the Gen AI foundations for enterprise workflow agents, owned, controllable and

Through many customer conversations, we discovered the top adoption criterion of GenAI agents to be the reliability to accomplish tasks with minimal hallucinations. 📢Checkout how @NexusflowX leverage a new agent building paradigm, with extractive reasoning as the first-class

Very exciting long context length multi modality capability from Google. The full repo code analysis part is pretty amazing!

📢Exciting release of Starling-7B-beta chat model and Starling-34B-RM reward model powered by Nexusflow latest technology. I am continuously amazed by how fast and powerful the small striking team behind Starling is!

Great work from Banghua Zhu and team Top ranking reward model on reward bench from Ai2 and new starling beta for chat

🔥 Starling-7B Beta from @NexusflowX climbing fast on Chatbot Arena, outperforming or rivaling larger models like Gemini Pro, Mixtral 8 * 7B and ranked as #1 7B chat model. 🤔 While DBRX from Databricks presents a new strong open model, the story of Starling-7B Beta shows how

🚀🌟🚀Excited to announce Samba-CoE v0.2, which outperforms DBRX by Databricks Mosaic Research and Databricks, Mixtral-8x7B from Mistral AI, and Grok-1 by Grok at a breakneck speed of 330 tokens/s. These breakthrough speeds were achieved without sacrificing precision and only on 8 sockets,

Running at 430 tokens/second using full precision and 8 sockets, #Llama3 from AI at Meta is now available on SambaNova Platform: fast.snova.ai 🚀Get full 16-bit precision 🚀Spend on only 8 chips, not 576 chips for 430 tokens/second! Trim chips, not precision! Test

📢 New function calling and GenAI agent course launched in collaboration with Andrew Ng and DeepLearning.AI. Come and try out the tutorial built by Venkat and @NexusflowX team! Hope you enjoy it.

📢 Excited to release Athene-Llama3-70B chat LLM, delivering new record on Arena Hard from @lmsys Chatbot Arena! 🔥For the first time, open-weight models really breathe down the neck of Claude-3.5 and GPT-4o on Arena Hard. 🛠️Athene-70B comes from @NexusflowX targeted

Human specialists excel through education & experience. This applies to LLMs: post-training of free/open models can produce top performers on various benchmarks. Athene-70B, a fine-tuned version of Llama-3-70B, is ascending to the top of Arena-Hard. #OpenAlwaysWins

investors when you say agentic workflows instead of function calls in a while loop

Chatbot Arena update! @NexusflowX's Athene-70B, an open-weight model fine-tuned from Llama-3-70B, is now ranked #8 on the leaderboard with a significant 30+ ELO boost from Llama-3. We see balanced multilingual capability and strong performance in hard prompts/coding. Congrats

Athene-V2 Advancing Beyond the Limits of Scaling with Targeted Post-training

Did you see those shiny new Athene models from our friends at @NexusflowX ? Good news, as usual you can run them right now in moved to: @lmstudio ! :D Find them using the CLI tool 'lms get athene-v2-chat' or on huggingface as usual! huggingface.co/lmstudio-commu… huggingface.co/lmstudio-commu…

Congrats Nexusflow on the latest Athene-V2-72B release, matching top models across hard benchmarks! Now it comes the real test—Athene is live in Arena for human evaluation. Come ask tough prompts at lmarena. ai!

Exciting update from Chatbot Arena! Athene-V2-Chat-72B by Nexusflow debuts as the best open model, matching proprietary models like GPT-4o/Sonnet in technical domains (e.g., math, coding, hard prompts)! Category ranking: - Math: #3 - Coding: #7 - Hard Prompt: #6 - Overall #10

🚀 We’re hiring at NVIDIA! Our team is pushing the frontier of LLM / DLM post-training and system optimization. We are looking for exceptional people with large-scale LLM + systems experience to join us (full time only). 🔹 Focus areas include: •Post-training of large models