Qwen3.0 MOE? New Reasoning Model?

Where to host deepseek R1 671B model?

8xB200 - Fully Idle for the Next Few Weeks - What Should I Run on It?

Any Russians here? From what sites do you order tea directly from China?

А что звучит вкусно

How does DeepSeek compare to other LLMs with creative writing and non "deep-thinking" tasks?

deepseek is a side project

The first performant open-source byte-level model without tokenization has been released. EvaByte is a 6.5B param model that also has multibyte prediction for faster inference (vs similar sized tokenized models)

Deepseek R1 GRPO code open sourced 🤯

kluster.ai now hosts deepseek R1

Google Pours Another $1 Billion Into OpenAI Competitor Anthropic

Bytedance Model: Doubao 1.5pro - Doubao Team

Bug or misunderstanding settings?

q1-3B-PRIME, a tiny reasoning model trained with RL on top of SmallThinker-3B

Free Deepseek Web Chat using R1?

Russian Sea Icelandic Herring in Oil

o1 performance at ~1/50th the cost! And Open Weights!

Deepseek just uploaded 6 distilled verions of R1 + R1 "full" now available on their website.

Deepseek R1 / R1 Zero

Earl gray tea with bergamot flavoring? (instead of oil)

MiniMax-Text-01 - A powerful new MoE language model with 456B total parameters (45.9 billion activated)

Can green tea be drunk with milk?

NVidia's official statement on the Biden Administration's Ai Diffusion Rule

New Model from https://novasky-ai.github.io/ Sky-T1-32B-Preview, open-source reasoning model that matches o1-preview on popular reasoning and coding benchmarks — trained under $450!

⚡️ Геймеры будут входить в игры по БИОМЕТРИИ или номеру телефона — законопроект уже поступил в Госдуму, его рассмотрят в январе. Это именно тот безумный закон, которым депутаты пугали в декабре. Геймеров будут не только шмонать на входе в игры, но и предупреждать об их «деструктивном содержании».