We evaluated Code Llama against existing solutions on both HumanEval & MBPP.
– It performed better than open-source, code-specific LLMs & Llama 2.
– Code Llama 34B scored the highest vs other SOTA open solutions on MBPP — on par w/ ChatGPT. More info https://
bit.ly/45JiPwJ
OPEN SOURCE
-
Code Llama Outperforms Open-Source Solutions on HumanEval MBPP
By
–
-
LIMA dataset recommendation for instruction finetuning with high-quality examples
By
–
Agreed! And there’s LIMA if you are looking for 1k high-quality examples for instruction finetuning:
-
Python ecosystem shifting to Rust over C++ for performance
By
–
Or more like a competitor to Rust since the Python ecosystem is moving to that instead of C++ (eg see polars and some others)? And for GPU & AI stuff a competitor to Triton/CUDA?
-
Code Llama Now Available in Hugging Face Playground
By
–
You can try Code Llama now in the Code Llama playground @huggingface space — it's also available in the Hugging Face ecosystem, starting with transformers version 4.33.
-
Groq Achieves 100 Tokens/Second Milestone with Llama2 70B
By
–
@GroqInc recently became the world's first to achieve the 100 t/s/u responsiveness milestone for text generation on @MetaAI
's #Llama2 70B model. Before the celebratory swag made it to Toronto, the team already blew past their record. Stay tuned for news about our latest milestone -
Meta Launches SeamlessM4T, Open Source Multilingual Model
By
–
Meta unveils SeamlessM4T, an open source multilingual and multimodal foundational model https://actuia.com/actualite/meta-devoile-seamlessm4t-un-modele-de-base-multilingue-et-multimodal-open-source/
… #AI #artificialintelligence #opensource -
Potential Competitor to Triton for Python Users
By
–
Yeah, we’ll see. It’s probably not going to be used by the average Python users, but I can see it becoming a competitor for Triton though.
-
New startup founded by former TensorFlow.js creators
By
–
A new startup founded by former members of team that created TensorFlow.js at Google Brain 🧠 https://t.co/xVsGIdBBzG
— hardmaru (@hardmaru) 25 août 2023A new startup founded by former members of team that created TensorFlow.js at Google Brain
-
Simplifying PyTorch by removing C++ dependencies
By
–
What I am really hoping for is that it will help simplify code bases, though. Maybe we can get rid of the C++ parts of PyTorch this way to make development easier.
-
PyTorch Python overhead and realistic performance gains discussion
By
–
2/2
PyTorch's Python overhead is usually quoted to be at ~10% (compared to PyTorch's C++ API). I think for PyTorch users (or CUDA-dependent packages), the speed advantage is probably more going to be in the single digit range, not 35,000x.
Still super exciting though!