Is it open if it doesn't include the training data? Genuine question - I am not familiar enough with the terms and technology to know. But my understanding is the weights is just a more or less static collection of data that has been (to paraphrase Ted Chiang) lossily compressed from the actual raw training data.
Without the training data to thoroughly evaluate what is in there, the only way you can figure it out is through experimentation - e.g. running it up in a chatbot and asking it questions.
Is this roughly correct or am I misunderstanding what you can do with the weights?
Without the training data to thoroughly evaluate what is in there, the only way you can figure it out is through experimentation - e.g. running it up in a chatbot and asking it questions.
Is this roughly correct or am I misunderstanding what you can do with the weights?