…"in that llm.c will never be as fast nor as simple as pytorch"
OPEN SOURCE
-
Open-source tool converts chat prompts for Mixtral base model
By
–
Introducing `instruct-to-base-prompt-converter` ✍️
— Matt Shumer (@mattshumer_) 12 avril 2024
Prompting base models like Mixtral 8x22B is insanely hard.
Now, AI can do it for you!
Uses Claude 3 Opus to automatically convert a chat prompt to a prompt that works with the Mixtral-8x22B base model!
And it's open-source! pic.twitter.com/rKAgkXxidnIntroducing `instruct-to-base-prompt-converter` Prompting base models like Mixtral 8x22B is insanely hard. Now, AI can do it for you! Uses Claude 3 Opus to automatically convert a chat prompt to a prompt that works with the Mixtral-8x22B base model! And it's open-source!
-
LLM.c Versus PyTorch: Practical Limitations of Low-Level Implementation
By
–
my take on all this llm.c stuff is that it’s very impressive, and karpathy is certainly brilliant, but it was sort of a futile exercise in that llm.c will never be as fast nor as simple as pytorch this whole movement to “write everything in the lowest language possible” is a tad
-
Fine-tuning Open Source LLMs to Outperform GPT-4
By
–
There's a common belief that #opensource LLMs aren't as performant as #GPT4. Well, oftentimes they're not… unless you #finetune! Join our upcoming virtual workshop to learn how you can fine-tune OSS #LLMs to outperform GPT-4: https://
pbase.ai/3xAPXuV -
Redis license change: 703 contributors consent question
By
–
https://
github.com/redis/redis lists 703 contributors, how many of them are on board with the license change I wonder? -
Community Fine-Tuned Model Now Available on Hugging Face
By
–
Not yet officially, there's at least one community instruction fine-tuned model on Hugging Face already
-
GPT4T Reclaims SOTA with Open Source Lightweight Evals
By
–
GPT4T retakes the SOTA LLM spot (probably) but also i like this way of announcing things: open source your evals, make it reproduceable and comparable. and **LIGHTWEIGHT**: eleuther’s harness and oai/evals always felt far too heavy of a lift and not flexible. this thing
-
Open-source program-form implementation of ARC 1 training tasks released
By
–
Just open-sourced by Michael Hodel: a program-form implementation of each of the 400 training tasks of ARC 1, capable of generating more task instances. I expect this will be very useful to everyone working on ARC. GitHub: https://
github.com/michaelhodel/r
e-arc
…
ArXiv: https://
arxiv.org/abs/2404.07353 -
Apple’s 512GB RAM Investment Strategy Strengthens MLX Framework
By
–
If Apple have 512GB of shared RAM/VRAM in the pipeline then their recent investments in the open source MLX framework make even more sense
