jbentley1

7 ms·

1. 61.▲
   by jbentley1 2y ago

   It's 24B parameters
2. 62.▲
   by jbentley1 2y ago

   The technical parts are less common and specialized, like understanding the hyperparameters and all that, but I don't think that is the main problem. Most people don't understand how to build a good dataset or how to evaluate thei
3. 63.▲
   by jbentley1 2y ago

   Tiny language models can do a lot if they are fine tuned for a specific task, but IMO a few things are holding them back: 1. Getting the speed gains is hard unless you are able to pay for dedicated GPUs. Some services offer LoRA as serverle
4. 64.▲
   by jbentley1 2y ago

   Things can both please share holders and create value for users of that thing.
5. 65.▲
   by jbentley1 2y ago

   Multi-armed bandits make a big assumption that effectiveness is static over time. What can happen is that if they tip traffic slightly towards option B at a time when effectiveness is higher (maybe a sale just started) B will start to overw