Upload README.md with huggingface_hub

README.md

@@ -47,12 +47,14 @@ architecture, hyperparameters, training epoch and validation loss, readable
 without loading the checkpoint.
 
 `esmc_human_cache.pt` (95 MB) sits at the repository root and is required by the
-three `_esmc` checkpoints. It is the frozen embedding table those models were
-trained against: ESM-C 600M run over the canonical human proteome, mean-pooled
-to one 1152-dimensional vector per UniProt accession. It is shipped rather than
-regenerated so the embeddings are byte-identical to training, and because
-recomputing it means downloading a 600M-parameter model and running inference
-over 20,000 sequences. The `learned` checkpoints do not need it.
+three `_esmc` checkpoints, which use it as a frozen lookup table instead of
+learning their feature embeddings. It holds one 1152-dimensional vector per
+UniProt accession: ESM-C 600M run over the canonical human proteome, mean-pooled
+per sequence. It can be recomputed, but that means downloading a
+600M-parameter model and running inference over 20,000 sequences. Shipping it
+keeps the embeddings bit-identical to what the models saw during training, and
+lets anyone run inference with the `_esmc` checkpoints straight away. The
+`learned` checkpoints do not need it.
 
 ## Use