Hmm what if you did token->thought1->token->thought2->token->thought3->token->thought4->token->thought1->token->thought2… ?
LLMS
-
Token Budget Impact on Model Performance and Downstream Tasks
By
–
Have you tried seeing what impact different amounts of extra tokens has? Have you looked at the impact on downstream tasks too?
-
Fine-tuning pretrained models with new tokens
By
–
That's interesting – I haven't seen that done before. So did you just add a few tokens to a pretrained model, and then fine-tune them somehow? Or left them as randomly initialised?
-
Researcher Access Program for deprecated models and GPT-4
By
–
There's a Researcher Access Program for deprecated models, including at least code-davinci-002. They also offer base GPT-4:
-
Mathpix & ChatGPT: Screenshots to LaTeX to Code Workflow
By
–
…or, as a few replies suggested, use Mathpix to turn screenshots into LaTeX and then ChatGPT to turn LaTeX into code:
-
Are Thought Tokens Different Pause Tokens in LLMs?
By
–
Are thought1, thought2, thought3… actually different tokens? Or all copies of the same pause token?
-
Voyager AI Agents with Foundation Models Research
By
–
Thank you @CadeMetz for covering my team's work on #AI agents with foundation models: Voyager. @yukez @guanzhi_wang @DrJimFan @YunfanJiang @nvidia @caltech
-
Poe Subscriptions Launch in Mexico and Brazil
By
–
Poe subscriptions are now available in Mexico and Brazil! Subscribing gives you access to powerful models like GPT-4 and Claude-2 and any user-created bots built on them, higher message limits, and faster response times. You can subscribe on iOS, Android, MacOS, or the web.
-
Deep Learning Fundamentals: Optimizing Performance from Scratch
By
–
Making Deep Learning go Brrrr From First Principles https://
bit.ly/46dO3fC
#AI #MachineLearning #DeepLearning #LLMs #DataScience -
GPT-4 Vision: Exploring Counting and Multimodal AI Capabilities
By
–
After a week of access to GPT4 with Vision I finally figured it out. It can count squares on the pavement, this is very useful I wonder what else it can count?