*   Added `ignore_class` argument in the loss
        `SparseCategoricalCrossentropy` and metrics `IoU` and `MeanIoU`, to
        specify a class index to be ignored during loss/metric computation (e.g.
        a background/void class).
    *   Added
        [`tf.keras.models.experimental.SharpnessAwareMinimization`](https://www.tensorflow.org/api_docs/python/tf/keras/models/experimental/SharpnessAwareMinimization).

7}, crypto.SHA3_256: {2, 16, 840, 1, 101, 3, 4, 2, 8}, crypto.SHA3_384: {2, 16, 840, 1, 101, 3, 4, 2, 9}, crypto.SHA3_512: {2, 16, 840, 1, 101, 3, 4, 2, 10}, } for h, oid := range prefixes { want, err := asn1.Marshal(struct { HashAlgorithm pkix.AlgorithmIdentifier Hash []byte }{ HashAlgorithm: pkix.AlgorithmIdentifier{ Algorithm: oid, Parameters: asn1.NullRawValue, }, Hash: make([]byte, h.Size()), }) if err != nil { t.Fatal(err) } want = want[:len(want)-h.Size()] got := hashPrefixes[h.String()] if...