Not great w/o adjusting the weights. This result took weeks of intense data tuning + full-scale training.
@mattshumer_
-
Testing Advanced AI Model Capabilities with Writing Prompts
By
–
If you want to get a sense of what this beast of a model can do, comment your writing prompts. I'll generate and respond to as many as I can.
-
HyperWrite Transforms Into Interactive Choose-Your-Own-Adventure Experience
By
–
Done! It kind of turns HyperWrite into a choose-your-own adventure sort of thing.
-
HyperWrite experimental asshole mode unlikely to launch
By
–
experimental "asshole mode" for HyperWrite something tells me… we're not gonna launch this one
-
Price Decrease Expected Within Few Years Ahead
By
–
More expensive currently. I expect that to shift within a few years.
-
Scalable Self-Play with LLMs: Current Research and New Approaches
By
–
What's the closest researchers have gotten to scalable self-play w/ LLMs? Had an idea I think is unique and easily implementable, curious if there is anything I should take a look at.
-
Fine-tuning Shifting to Many-shot Prompting in Coming Years
By
–
I’ve been saying this for a while now. Over the next couple of years, *most* fine-tuning work will become many-shot prompting.
-
LLaMA 3 Context Window Extended to 1 Million Tokens
By
–
Thanks @KalleyHuang and @theinformation for writing about my work extending LLaMA 3’s context window. And shoutout to @winglian @nisten and @markatgradient who took this even further and scaled LLaMA to 1M tokens. Article in next tweet:
-
AI Agent Version Gains Rapid Follower Growth
By
–
wtf, @aimattshumer (the AI version of me) is growing followers fast going to sleep, excited to see where it’s at in the morning