#TinyZero
UC Berkeley Scientists Replicate DeepSeek AI for Just $30 interestingengineering.com/innovation/u...
DeepSeek replicated for peanuts: US scientists achieve feat in $30
TinyZero achieves impressive results with minimal resources, raising questions about the cost of AI development.
interestingengineering.com
February 1, 2025 at 3:58 AM
La boîte de Pandore a été ouverte par deepseek 😅 Jiayi pan vient de sortir Tinyzero une méthode capable de re-entraîner des modèles 3B ou 0.5B à la manière de Deepseek zéro. Ça donne un modèle sympa pour moins de 30$ 😯 #ia #deepseek #mistral #llm

github.com/Jiayi-Pan/Ti...
GitHub - Jiayi-Pan/TinyZero: Clean, accessible reproduction of DeepSeek R1-Zero
Clean, accessible reproduction of DeepSeek R1-Zero - Jiayi-Pan/TinyZero
github.com
January 31, 2025 at 6:58 AM
"You can experience the Ahah moment yourself for < $30"

github.com/Jiayi-Pan/Ti...
GitHub - Jiayi-Pan/TinyZero: Clean, accessible reproduction of DeepSeek R1-Zero
Clean, accessible reproduction of DeepSeek R1-Zero - Jiayi-Pan/TinyZero
github.com
January 30, 2025 at 4:44 PM
TinyZero AI: Berkeley's $30 Answer to DeepSeek AI Affordable AI Research
## TinyZero AI: Berkeley's $30 Alternative to DeepSeek AI **Game-Changer in AI Research: TinyZero** State-of-the-art AI research has generally required enormous resources. These resources are generally in the possession of large tech corporations. However, UC Berkeley researchers have just unveiled TinyZero AI. TinyZero AI is a revolutionary low-cost AI model. It mimics the reasoning of DeepSeek R1-Zero for only $30. This breakthrough shatters the illusion that upper-level AI requires a Silicon Valley-sized budget. It proves that innovation can occur in open-source environments. ## TinyZero AI - $30 Alternative to DeepSeek ### What is TinyZero AI? **Democratizing AI Research at Low Cost** Jiayi Pan and researchers at UC Berkeley created TinyZero. TinyZero is an AI model that leverages a simple language model and reinforcement learning (RL) to reason effectively. > "You can try the 'Aha' for yourself for less $30." - Jiayi Pan, UC Berkeley Researcher ### How TinyZero Learned to Reason: The Countdown Game The research team trained TinyZero using the Countdown game. The Countdown game is a mathematical challenge where AI must reach a target number using addition, subtraction, multiplication, and division. ### TinyZero vs. DeepSeek AI: Cost Breakdown Review the AI Model/Service cost comparison: AI Model/Service | Cost ---|--- OpenAI’s API | $15 per million tokens DeepSeek-R1 | $0.55 per million tokens TinyZero (Total Training Cost) | $30 (One-Time) ### Is TinyZero AI Open-Source? Yes. TinyZero is open-source on GitHub. This allows anyone to: * Download and modify the source code * Run experiments on reinforcement learning * Explore alternative training methods ## Final Thoughts: The Future of Affordable AI **Key Takeaways:** * TinyZero is evidence that cutting-edge AI research doesn't require expensive resources. * It self-learns & self-validates solutions, same as DeepSeek AI. * Open-source availability ensures future innovation.
www.technetbooks.com
March 17, 2025 at 3:16 PM
TinyZero: Democratizing AI Research with $30 Reasoning Model UC Berkeley Breakthrough
Groundbreaking AI research often requires vast resources, typically found within major technology corporations. However, UC Berkeley researchers have presented TinyZero, an intriguing development. This model effectively emulates the core reasoning of DeepSeek R1-Zero at an astonishingly low cost of approximately US$30. This showcases that significant AI advancements do not always necessitate Silicon Valley-sized budgets. ## Democratization of AI Research **TinyZero AI Research Concept** Jiayi Pan and his collaborators are spearheading efforts to make advanced AI reasoning models more accessible. Their concept, TinyZero, achieves this without extensive infrastructure investments. It is trained using a straightforward language model and a functional reinforcement learning system, RL. This elegant approach avoids substantial computing power and expensive cloud services, democratizing AI experimentation. > "You can try the 'Aha' for yourself for <$30," - Jiayi Pan, UC Berkeley Researcher. Pan expressed enthusiasm on social media, noting TinyZero as the first publicly available replication of advanced reasoning models. Crucially, TinyZero learns to generate answers and critically evaluate and refine them, a hallmark of advanced reasoning. ## How TinyZero Learned to Reason: The Countdown Game Researchers trained TinyZero using the Countdown game. In this game, the AI must reach a target number using addition, subtraction, multiplication, and division. Initially, TinyZero's approach was random. However, with training, it began to learn. TinyZero developed the ability to self-check answers, identify better solutions, and adjust strategies. It was genuinely learning to reason. Model sizes from 500 million to 7 billion parameters were tested. Smaller models resorted to guessing. Larger models, starting at 1.5 billion parameters, demonstrated self-verification and refinement, leading to significant accuracy improvements. ## Astonishing Cost Difference TinyZero's cost-effectiveness is remarkable. Consider this comparison: AI Model/Service | Cost ---|--- OpenAI’s API | $15 per million tokens DeepSeek-R1 | $0.55 per million tokens TinyZero (Total Training Cost) | $30 (One-Time) The contrast is significant. Accessing powerful AI models via APIs is expensive. TinyZero's one-time $30 training cost unlocks opportunities for individuals, small research groups, and educational institutions to experiment with advanced AI reasoning. ## Open Source and the Future of Affordable AI TinyZero is open-source on GitHub. This allows open access for review, modification, and further development. Currently validated in the Countdown game, its broader implications are substantial. Jiayi Pan intends TinyZero to catalyze wider accessibility to reinforcement learning research, fostering broader community innovation. While TinyZero is still in early stages and its general reasoning capabilities across diverse areas are yet to be fully explored, its core message is clear. AI development need not be prohibitively expensive. Projects like TinyZero pave the way for a future where open-source, affordable AI drives a new era of innovation and discovery.
www.technetbooks.com
February 1, 2025 at 8:30 PM
Side note: we saw some nice adoption from DeepSeek-R1 reprod repos, which is humbling, if not thrilling!

github.com/Jiayi-Pan/Ti...
GitHub - Jiayi-Pan/TinyZero: Clean, minimal, accessible reproduction of DeepSeek R1-Zero
Clean, minimal, accessible reproduction of DeepSeek R1-Zero - Jiayi-Pan/TinyZero
github.com
February 5, 2025 at 5:02 PM
June 26, 2025 at 6:34 PM
TinyZero: reproduction of DeepSeek R1-Zero for countdown and multiplication tasks using reinforcement learning
June 26, 2025 at 6:34 PM
Jiayi-Pan/TinyZero
Check out Jiayi-Pan/TinyZero on GitHub
github.com
January 26, 2025 at 11:45 PM
eh that's a fork of github.com/Jiayi-Pan/Ti..., has a bunch of python repos with half baked code in it to call out.
GitHub - Jiayi-Pan/TinyZero: Clean, minimal, accessible reproduction of DeepSeek R1-Zero
Clean, minimal, accessible reproduction of DeepSeek R1-Zero - Jiayi-Pan/TinyZero
github.com
February 4, 2025 at 9:17 PM
Easily train AlphaZero-like agents on any environment you want github.com/s-casci/tiny...
GitHub - s-casci/tinyzero: Easily train AlphaZero-like agents on any environment you want!
Easily train AlphaZero-like agents on any environment you want! - GitHub - s-casci/tinyzero: Easily train AlphaZero-like agents on any environment you want!
github.com
January 4, 2024 at 11:58 AM
Marko Elez quit his Nazi boss bcuz he was caught erasing his own racist comments 🤔 that’s weird, Elons Grok AIs competition supposedly was #DeepSeek but Marko Elez, not mentioned in the article,built a dupe w/ this guy. A dupe that is not being banned. 🤔 huh.

www.notebookcheck.net/DeepSeek-AI-...
DeepSeek AI clone called TinyZero has been developed for just $30
It can be made even cheaper. PhD students from Berkeley have created an AI that supposedly is comparable to DeepSeek's R1-Zero model. Allegedly, the entire AI has been developed for a grand total of j...
www.notebookcheck.net
February 7, 2025 at 10:52 AM
lololololol
a bajillion dollar industry, and now folks are like, "oh I can do that ezpz".
how's #stargate going, hm?'
#ai #deepseek
DeepSeek replicated for peanuts: US scientists achieve feat in $30
TinyZero achieves impressive results with minimal resources, raising questions about the cost of AI development.
interestingengineering.com
February 2, 2025 at 9:03 PM