131–140 of 817 posts
Earlier quoted context omitted.
Something that follows my actual requests, without trying to lecture me about feminism and other U.S. Democrats topics.
What is an example of a request that is causing these issues?
Earlier quoted context omitted.
Try out Bard, it's coding is much improved in the last 2 weeks. I've unfortunately switched over for the time being.
“Bard isn’t currently supported in your country. Stay tuned!”
https://console.cloud.google.com/vertex-ai/generative/langua...
Let me guess: the "woke fine-tuning" deteriorates quality of the model. Seriously, this lecturing style of GPT is totally useless. You ask it something vaguely unwoke, it lectures you about your style, you explain what it didn't understand about your request, it appologizes and does vaguely what you originally requested. This sort of deture is totally useless and a waste of my time. To stay with the terminology, if I…
Genuinely asking, what's an "unwoke" prompt?
LLMs are (naturally) very sensitive to their prompts, so if the prompt includes something about being inclusive - at ChatGPT's obviously does - the LLM output will sometimes find ways to work that in in odd ways.
I don't mind as much as the commenter seems to - I have no beef with inclusivity - but I have rolled my eyes at times at what ChatGPT comes up with.
I've played with local LLMs too, including some 'uncensored' ones, but the gap in ability between them and something like GPT is still vast. I look forward to progress on this front though.
There's no doubt that it's gotten a lot worse on coding, I've been using this benchmark on each new version of GPT-4 "Write a tiptap extension that toggles classes" and so far it's gotten it right every time, but not any more, now it hallucinates a simplified solution that don't even use the tiptap api any more. It's also 200% more verbose in explaining it's reasoning, even if that reasoning makes no sense whatsoever…
Do you have API access? If so, have you tried your tiptap question on the gpt-4-0314 model? That is supposedly the original version released to the public on March 14.
Yes. Before the update, when its avatar was still black, it solved pretty complex coding problems effortlessly and gave very nuanced, thoughtful answers to non-programming questions. Now it struggles with just changing two lines in a 10-line block of CSS and printing this modified 10-line block again. Some lines are missing, others are completely different for no reason. I'm sure scaling the model is hard, but they l…
There's no doubt that it's gotten a lot worse on coding, I've been using this benchmark on each new version of GPT-4 "Write a tiptap extension that toggles classes" and so far it's gotten it right every time, but not any more, now it hallucinates a simplified solution that don't even use the tiptap api any more. It's also 200% more verbose in explaining it's reasoning, even if that reasoning makes no sense whatsoever…
Do you have by any chance tested the same question on the playground? I've noticed a quality decrease iny telegram bot as well that directly uses the API, and it drives me crazy because model versioning was supposedly implemented specifically to avoid response change without notice
"You are ChatGPT, a large language model trained by OpenAI, based on the GPT-4 architecture. Knowledge cutoff: 2021-09 Current date: 2023-05-31"
And custom roles with custom content via API.