Je pourrais regarder ce graphe toute la journée. La vitesse des progrès est littéralement vertigineuse. En 2023, GPT-4 atteignait à peine le niveau des compétiteurs moyens sur Codeforces. (compétition mondiale de codage) En 2025, le modèle o3 dépasse déjà les meilleurs humains
LLMS
-
Step by step guide to fine tune Qwen 3
By
–
Step by step guide to fine tune Qwen 3. Thanks for sharing Avi!
-
Are all 20-year-olds using ChatGPT for life decisions?
By
–
So is every 20 yr old using ChatGPT for life choices
-
Zuckerberg and Amodei Predict AI Will Write Most Code Soon
By
–
– Mark Zuckerberg : d'ici 12 à 18 mois, la majorité du code sera écrite par l'IA. – Dario Amodei : dans seulement 3 à 6 mois, l'IA écrira 90 % du code, et dans un an, presque la totalité. Il reste difficile de comprendre pourquoi ils font tous deux des affirmations aussi
-
LangChain Ambassador Shares Mandarin Translation of Agentic Frameworks
By
–
Check out LangChain Ambassador @zhanghaili0610
's Mandarin translation of Harrison's blog post on how to think about agentic frameworks -
Running Llama4 on SambaNova Cloud feels like Top Gun
By
–
No, this isn’t a Top Gun: Maverick poster. It’s just what running #Llama4 on SambaNova Cloud feels like.
-
AI Agent Auto-Selects Best LLM for Each Task
By
–
And folks, ICYMI, you also need to see this!
— Charly Wargnier (@DataChaz) 3 mai 2025
Nate Herkelman also built an AI agent in @n8n_io that auto-selects the best LLM for each task… and logs why! 🔥https://t.co/IfaS13R7I2And folks, ICYMI, you also need to see this! Nate Herkelman also built an AI agent in @n8n_io that auto-selects the best LLM for each task… and logs why!
-
Gemini and LangChain Integration Guide with Multimodal Capabilities
By
–
Gemini x LangChain Guide Developer guide for integrating Gemini with LangChain, packed with multimodal capabilities, tool calling, and structured output implementations. Includes latest model support and practical code examples. Check out the guide here
-
Gemini Long Context Pre-Training: Scaling to Infinite Context
By
–
A deep conversation with @SavinovNikolay, the Gemini long context pre-training co-lead…
— Logan Kilpatrick (@OfficialLoganK) 3 mai 2025
We go from the basics to what is needed to scale to infinite context to long context best practices for devs:pic.twitter.com/UMk5Svw3qYA deep conversation with @SavinovNikolay
, the Gemini long context pre-training co-lead… We go from the basics to what is needed to scale to infinite context to long context best practices for devs: