Love this comprehensive comparison from @_anshulr on Copilot vs Tabnine vs Ghostwriter vs @codeiumdev https://codeium.com/blog/code-assistant-comparison-copilot-tabnine-ghostwriter-codeium 3 axes of comparison: Features/Price, Latency, Quality.
@swyx
-
Training Models for Specific Music Styles Using RLHF-Inspired Approaches
By
–
oh i mean split-to-analyse-progress not split-into-different-products, that would be horrible haha i think the insight for pop styles maybe similar to RLHF PPO. instead of generating a bunch of music and seeing what sticks, train models that reflects a style you want & set loose
-
Music Generation Needs Specialization Like DAW Components
By
–
“music generation” is too broad. need to split various jobs of audio out like a DAW would (someone could prob draw an unbundling chart from that) background music generation is essentially now solved. then voice synth, then melodies. feel like pop SHOULD be tractable bc formulaic
-
Few Shot Learning Illustrated Guide and Examples
By
–
Few shot learning, illustrated pic.twitter.com/q0E4Ev9zTx
— swyx 🐣 (@swyx) 28 janvier 2023Few shot learning, illustrated
-
Sundar pressured by ChatGPT from multiple directions
By
–
and the always excellent @reckless on the @verge breaking down how Sundar is being pressured from both above and below by ChatGPT:
-
Prompt Engineering as Systems Engineering: LLM Workflow Abstraction
By
–
“Prompt engineering looks more like a systems engineering problem, not a machine learning one.” i like the concept of an “LLM friendly workflow engine”. concats all my recent professional interests into one sorely needed abstraction
-
Eager to Explore and Experiment with DeepFloydIF
By
–
i want to play with @DeepFloydIF so badly, let us at it!!!
-
GANs, Diffusion, and Transformers: Three Competitive AI Approaches
By
–
accurate to say that all 3 approaches (GANs, diffusion, transformers) are all somehow competitive with each other now? what a strange timeline!
-
OpenAI’s Indifference to Prompt Disclosure and Security
By
–
“OpenAI doesn’t seem to care, nor should they. The prompt only tells you one small piece of how to build ChatGPT.” looks like @goodside
, originator of prompt injection, agrees! https://
news.ycombinator.com/item?id=344823
18
… -
Request for LLM Metrics Spreadsheet Comparison Table
By
–
AI Twitter, request for help: 1-2 weeks ago I saw someone tweet out a spreadsheet/table of all known LLMs with details like parameter size, FLOPs, and other quantitative metrics. Basically an updated version of this 2021 "AI and Memory Wall" post below. I can't find it. Help?