*   `tf.compat.v1.nn.fused_batch_norm` backprop to `offset` when
            `is_training=False`
        *   `tf.image.adjust_contrast` forward
        *   `tf.nn.depthwise_conv2d` backprop to `filter` when not using cuDNN
            convolution
        *   `tf.image.resize` with `method=ResizeMethod.NEAREST` backprop
        *   `tf.math.bincount` - TODO: confirm exception added

tScrollbar(),t(e._element).trigger(h.HIDDEN)})},p._removeBackdrop=function(){this._backdrop&&(t(this._backdrop).remove(),this._backdrop=null)},p._showBackdrop=function(e){var n=this,i=t(this._element).hasClass(d)?d:"";if(this._isShown&&this._config.backdrop){var s=P.supportsTransitionEnd()&&i;if(this._backdrop=document.createElement("div"),this._backdrop.className=u,i&&t(this._backdrop).addClass(i),t(this._backdrop).appendTo(document.body),t(this._element).on(h.CLICK_DISMISS,function(t){n._ignor...

  let description = [{
    Automatic space to depth transform is done by adding space to depth transform op after host input
    and applying space to depth transform for the first convolution and its backprop filter on TPU.

    For example, original program:

    ```mlir
    module {
      func @while_body {
        %input = "tf.IteratorGetNext"(...) {device = "/CPU:0"}: -> tensor<2x224x224x3xf32>

ents(),t._resetScrollbar(),g(t._element).trigger(le.HIDDEN)})},t._removeBackdrop=function(){this._backdrop&&(g(this._backdrop).remove(),this._backdrop=null)},t._showBackdrop=function(t){var e=this,n=g(this._element).hasClass(de)?de:"";if(this._isShown&&this._config.backdrop){if(this._backdrop=document.createElement("div"),this._backdrop.className=ue,n&&this._backdrop.classList.add(n),g(this._backdrop).appendTo(document.body),g(this._element).on(le.CLICK_DISMISS,function(t){e._ignoreBackdropClick...

ents(),t._resetScrollbar(),g(t._element).trigger(le.HIDDEN)})},t._removeBackdrop=function(){this._backdrop&&(g(this._backdrop).remove(),this._backdrop=null)},t._showBackdrop=function(t){var e=this,n=g(this._element).hasClass(de)?de:"";if(this._isShown&&this._config.backdrop){if(this._backdrop=document.createElement("div"),this._backdrop.className=ue,n&&this._backdrop.classList.add(n),g(this._backdrop).appendTo(document.body),g(this._element).on(le.CLICK_DISMISS,function(t){e._ignoreBackdropClick...

Automatic space to depth transform is done by adding space to depth transform op after host input
and applying space to depth transform for the first convolution and its backprop filter on TPU.

For example, original program:

```mlir
module {
  func @while_body {
    %input = "tf.IteratorGetNext"(...) {device = "/CPU:0"}: -> tensor<2x224x224x3xf32>

  %arg2: tensor<64xf32>
) -> tensor<1x28x28x64xf32> {

  // CHECK: "tf.FusedBatchNormGradV3"
  // CHECK-SAME: (%[[X_TRANSPOSE:[0-9]*]], %[[Y_TRANSPOSE:[0-9]*]],
  // CHECK-SAME: data_format = "NCHW"
  %x_backprop, %scale_backprop, %offset_backprop, %reserve_1, %reserve_2
    = "tf.FusedBatchNormGradV3"(%arg0, %arg1, %arg2, %arg2, %arg2, %arg2)
       {
         data_format = "NHWC",
         epsilon = 1.001 : f32,

  %arg2: tensor<64xf32>
) -> tensor<1x28x28x64xf32> {

  // CHECK: "tf.FusedBatchNormGradV3"
  // CHECK-SAME: (%[[X_TRANSPOSE:[0-9]*]], %[[Y_TRANSPOSE:[0-9]*]],
  // CHECK-SAME: data_format = "NCHW"
  %x_backprop, %scale_backprop, %offset_backprop, %reserve_1, %reserve_2
    = "tf.FusedBatchNormGradV3"(%arg0, %arg1, %arg2, %arg2, %arg2, %arg2)
       {
         data_format = "NHWC",
         epsilon = 1.001 : f32,

  // CHECK: %[[RES_PERM:.*]] = "tf.Const"()
  // CHECK-SAME: <{value = dense<[0, 2, 3, 1]> : tensor<4xi64>}>

  // CHECK: %[[RES_TPOSE:[0-9]*]] = "tf.Transpose"
  // CHECK-SAME: (%x_backprop, %[[RES_PERM]])
  // CHECK: return %[[RES_TPOSE]]

  %x_backprop, %scale_backprop, %offset_backprop, %reserve_1, %reserve_2
    = "tf.FusedBatchNormGradV3"(%arg0, %arg1, %arg2, %arg2, %arg2, %arg2)
       {
         data_format = "NHWC",

## Pre-pre requisite

- A working knowledge of MinIO codebase and its various components.
- A working knowledge of AWS S3 API behaviors and corner cases.

## Pre-requisite for backporting any fixes

Fixes that are allowed a backport must satisfy any of the following criteria's: