In this video, I explore the newly released DeepSeek v3, an impressive open weight model that matches proprietary models in benchmarks and independent tests. I'll cover its technical details, my own reasoning and coding tests, and highlight both the strengths and some minor issues of this powerful AI model.

LINKS:
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]
[ Ссылка ]


💻 RAG Beyond Basics Course:
[ Ссылка ]

Let's Connect:
🦾 Discord: [ Ссылка ]
☕ Buy me a Coffee: [ Ссылка ]
|🔴 Patreon: [ Ссылка ]
💼Consulting: [ Ссылка ]
📧 Business Contact: engineerprompt@gmail.com
Become Member: [ Ссылка ]

💻 Pre-configured localGPT VM: [ Ссылка ] (use Code: PromptEngineering for 50% off).

Signup for Newsletter, localgpt:
[ Ссылка ]



00:00 I DeepSeek v3
00:29 Technical Details of DeepSeek v3
01:14 Training Innovations and Cost Efficiency
02:14 Performance Benchmarks and Comparisons
07:00 Testing DeepSeek v3: Coding Capabilities
13:11 Testing DeepSeek v3: Reasoning Capabilities
23:53 Addressing Limitations and Ethical Concerns


All Interesting Videos:
Everything LangChain: [ Ссылка ]

Everything LLM: [ Ссылка ]

Everything Midjourney: [ Ссылка ]

AI Image Generation: [ Ссылка ]

свернуть