{ continue } k, v, _ := strings.Cut(line, " = ") switch k { case "GCD": GCD = v case "A": A = v case "B": B = v case "LCM": LCM = v t.Run(fmt.Sprintf("line %d", lineNum), func(t *testing.T) { if A == "0" || B == "0" { t.Skip("skipping test with zero input") } if LCM == "1" { t.Skip("skipping test with LCM=1") } p, _ := bigmod.NewModulus(addOne(decodeHex(t, A))) a, _ := bigmod.NewNat().SetBytes(decodeHex(t, A), p) q, _ := bigmod.NewModulus(addOne(decodeHex(t, B))) b, _ := bigmod.NewNat().SetBytes(decodeHex(t,...

{ continue } k, v, _ := strings.Cut(line, " = ") switch k { case "GCD": GCD = v case "A": A = v case "B": B = v case "LCM": LCM = v t.Run(fmt.Sprintf("line %d", lineNum), func(t *testing.T) { if A == "0" || B == "0" { t.Skip("skipping test with zero input") } if LCM == "1" { t.Skip("skipping test with LCM=1") } p, _ := bigmod.NewModulus(addOne(decodeHex(t, A))) a, _ := bigmod.NewNat().SetBytes(decodeHex(t, A), p) q, _ := bigmod.NewModulus(addOne(decodeHex(t, B))) b, _ := bigmod.NewNat().SetBytes(decodeHex(t,...

      advanced use cases can also write their own custom
      `tf.train.experimental.ShardingCallback`s.

* `tf.train.CheckpointOptions`
    * Added `experimental_skip_slot_variables` (a boolean option) to skip
    restoring of optimizer slot variables in a checkpoint.

*   `tf.saved_model.SaveOptions`

    * `SaveOptions` now takes a new argument called