During this process we actually want to open up everything—the training data, the pretraining data, the source code, the details of the parameters, and so on… We are also releasing all these intermediate check points that we have obtained throughout training.
We need to put in place a very clear methodology to evaluate how these models are working… the only way to be able to do that is if we have full access to the data, so that we can go back and really understand how the model is behaving.
[
add
]
[
|
|
...
]