> DeepSeek-R1 has been making waves recently by rivaling OpenAI's O1 reasoning model while being fully open-source.
Do we finally have a model with access to the training architecture and training data set, or are we still calling non-reproducible binary blobs without source form open-source?
I assume when people say "open source model" they mean "open weights model". The "open source" term doesn't really make sense here, since machine learning models are not compilations of source code. (Though DeepSeek has published several papers with details on their training process. It's more than just open weights.)
Do we finally have a model with access to the training architecture and training data set, or are we still calling non-reproducible binary blobs without source form open-source?