This is an implementation of deepseek-ai/deepseek-llm-67b-base as a Cog model.
Follow the model pushing guide to push your own model to Replicate.
To run a prediction:
cog predict -i prompt="An attention function can be described as mapping a query and a set of key-value pairs to an output, where the query, keys, values, and output are all vectors. The output is"