Efficient embedding table storage and lookup

1. A computer-implemented method for performing efficient embedding table storage and lookup in machine-learning models, comprising:

obtaining, by one or more processors, an embedding table associated with a machine-learning model, the embedding table comprising a plurality of individually compressed embeddings allowing each respective embedding to be decompressed independent of any other embedding in the embedding table;

receiving, by the one or more processors, an input to use for locating an embedding in the embedding table;

determining, by the one or more processors, a lookup value based on the input for searching indexes of the embedding table to locate the embedding;

locating, by the one or more processors, the embedding based on searching the indexes of the embedding table for the determined lookup value; and

decompressing, by the one or more processors, the embedding independent of any other embedding in the embedding table in view of the locating.

2. The computer-implemented method of claim 1 , further comprising:

processing, by the one or more processors, the decompressed embedding in association with running the machine-learning model.

3. The computer-implemented method of claim 1 , wherein the determining of the lookup value is based on a hashing operation that maps the input to one of the indexes in the embedding table.

4. The computer-implemented method of claim 1 , wherein respective embeddings in the embedding table are associated with an item from a plurality of items included in a vocabulary.

5. The computer-implemented method of claim 1 , wherein the decompressing is performed independent from a machine-learning platform.

6. The computer-implemented method of claim 2 , wherein the decompressing is performed using compression unavailable from a machine-learning platform.

7. A computing system for performing efficient embedding table storage and lookup in machine-learning models:

one or more processors; and

one or more non-transitory computer-readable media that collectively store:

a machine-learned model configured to receive model input and process the model input to generate model output, wherein the machine-learned model comprises an embedding table comprising individually compressed embeddings allowing each respective embedding to be decompressed independent of any other embedding in the embedding table, and wherein the machine-learned model is configured to perform operations comprising:

obtaining the model input for processing in view of the embedding table;

determining a lookup value based on the model input for searching indexes of the embedding table to locate an embedding;

locating the embedding based on searching the indexes of the embedding table for the determined lookup value; and

decompressing, by the one or more processors, the embedding independent of any other embedding in the embedding table in view of the locating.

8. The computing system of claim 7 , further comprising:

processing the decompressed embedding with other respective decompressed embeddings located from the embedding table.

9. The computing system of claim 7 , wherein the operations further comprise:

generating the model output based on processing the decompressed embedding.

10. The computing system of claim 7 , wherein the determining of the lookup value is based on a hashing operation that maps the model input to one of the indexes in the embedding table.

11. The computing system of claim 7 , wherein respective embeddings in the embedding table are associated with an item from a plurality of items included in a vocabulary.

12. The computing system of claim 7 , wherein the decompressing is performed independent from a machine-learning platform.

13. The computing system of claim 7 , wherein the decompressing is performed using compression unavailable from a machine-learning platform.