> By 2028, AI Clusters will reach the power consumption of entire countries Reminder : “Scaling laws” are empirical laws saying that if you keep multiplying your compute by x10, your models will mechanically keep getting better and better. To give you an idea, GPT-3 can barely
LLMS
-
OpenAI’s $1B safety fund and latest AI breakthroughs
By
–
Top stories in AI today: -OpenAI co-founder raises $1B for AI safety
-The fastest AI model goes multimodal
-Turn any text into speech in seconds
-AI gets smarter by re-reading questions
-6 new AI tools & 4 new AI jobs Read more: http://
therundown.ai/p/safe-superin
telligence-1-billion
… -
Google’s problem: too many steps to use Gemini
By
–
Here's Google's problem. To use Gemini, you need to create a GCP account, create a service account, go to Google AI Studio, activate the product, then generate your key, which you use using AI Studio but the product is actually called Generative API.
-
Comparison of memory features in ChatGPT and Gemini
By
–
It is hard not to draw parallels here ChatGPT is rolling out its Memory feature in EU and KR right after Gemini Saved Info release (which works similarly) Interestingly, ChatGPT refuses to remember that I am Groot h/t @btibor91
-
4 interesting ChatGPT prompt frameworks to save
By
–
4 interesting ChatGPT prompt frameworks (Save for later)
-
Yi-Coder Open-Sourced: Compact LLM for Coding Performance
By
–
🚀 Yi-Coder is open-sourced!
— Yi-01.AI (@01AI_Yi) 5 septembre 2024
The 'Small but Mighty' LLM offers SOTA coding performance under 10B parameters. Excel in code editing, completion, debugging, and math reasoning.
✅ 2 sizes: 9B & 1.5B (Chat & Base)
✅ 128K context length
✅ Support 52 programming languages
Explore… pic.twitter.com/QVscshn6PE🚀 Yi-Coder is open-sourced! The 'Small but Mighty' LLM offers SOTA coding performance under 10B parameters. Excel in code editing, completion, debugging, and math reasoning. ✅ 2 sizes: 9B & 1.5B (Chat & Base) ✅ 128K context length ✅ Support 52 programming languages Explore it now👇 huggingface.co/collections/0… #YiCoder #01AI #LLMs #OpenSource
-
Unsloth and Axolotl: Best Fine-Tuning Options Comparison
By
–
Then Unsloth and Axolotl are your best options imo. There's no difference in terms of PEFT/FFT. I think Spectrum is only supported by Axolotl at the moment if you're interested.
-
Unsloth vs Axolotl: Choosing the Right Fine-Tuning Framework
By
–
In general, I'd recommend Unsloth for single-GPU settings and Axolotl for multi-GPU settings. For research, you might want to have more fine-grained control though. TRL works there, but reimplementing it in PyTorch is also a good option for full control.
-
Advanced RAG and AI Future Sessions Registration
By
–
Register for our upcoming sessions! Tuesday, September 10:
3:20 – 3:40 PM: Advanced Retrieval-Augmented Generation and Tool Use with Cohere: https://
reg.rf.oracle.com/flow/oracle/oc
w24/catalog/page/catalog/session/1722635719286001qu0t
… Wednesday, September 11:
The Future of AI and Data:
9:45 – 10:30 AM: https://
reg.rf.oracle.com/flow/oracle/oc
w24/catalog/page/catalog/session/1718295997989001ez6I
… 3:30 – 4:15 PM:
