Enjoyed the article and I was ready to try the promoted featherless.ai. I signed up and spent 15 minutes trying load or chat with Llama 3 models. All attempts f

Enjoyed the article and I was ready to try the promoted featherless.ai. I signed up and spent 15 minutes trying load or chat with Llama 3 models. All attempts failed. Naturally I would ask, if it's so cheap to run GPU's, why I would need to sign up to try a model?