We go through periods of liberalization, and then it's opposite, as the ills of each regime become salient. For example, in the US, Oregon abandoned its legalization of hard-drugs, and I expect to see a national push to restrict and regulate cannabis more heavily, for example, regulating THC content more stringently, etc.
With so many things, finding the right balance takes trial and error, and what the right balance is may change as other variables change as well.
RAM and what GPU you have are big determinants of how fast it will run, and how smart a model you can run. A large amount of RAM and GPU memory is required for larger models without significant slowdown because its much faster if it can keep the entire model in memory. Small models range from 3-8 gigabytes, but a 70B parameter model will be 30-50 gigabytes.
Q4_K_S. While not as good as top commercial models like chatgpt, they are still quite capable and I like that there are also uncensored/abliterated models like Dolphin.
If you just want to chat, download https://lmstudio.ai/, then download their recommended LLM files, and you're good to go. Really that simple these days.
> Subject to the Agreement, Company grants you a limited license to reproduce portions of Company Properties for the sole purpose of using the Services for your personal, non-commercial purposes.
As long as the email account is secure, and the throw-away one-time passwords are good, you have the frequent-rotation passwords security advocates dream about. Indeed, hand them a secure password they have to use (and forget).
I just made it an option in the "I don't have a password" form that instead of setting a password it just logs them in. So they don't even see / have a password to remember.
When using code assist, I've occasionally found some perplexing changes to my code I didn't remember making (and wouldn't have made). Can be pretty frustrating.
Code quality is the hardest thing to measure. Seems like they were measuring commits, pull-requests, builds, and build success rate. This sort of gets at that, but is probably inadequate.
The few attempts I've made at using genAI to make large-scale changes to code have been failures, and left me in the dark about the changes that were made in ways that were not helpful. I needed suggestions to be in much smaller chunks. paragraph sized. Right now I limit myself to using the genAI line completion suggestions in Pycharm. It very often guesses my intentions and so actually is helpful, particularly when laboriously typing out lots of long literals, e.g. keys in a dictionary.
It isn't always the piece of code that is hard to read. It's about how it fits into the 200,000+ line application you're working with, what it was trying to solve, whether it solved the right problem, and whether there will be unexpected interactions.
I think in RL engineering circles offline, off-policy, generative experience replay is. In neuroscience, the idea is being actively explored, but it isn't a consensus position.
reply