| Commit message (Expand) | Author | Age |
* | Add an axis parameter to tf.gather. Fixes GitHub issue #11223. | RJ Ryan | 2017-07-11 |
* | Add a PadV2 op with support for specifying a pad value. | RJ Ryan | 2017-07-10 |
* | Adds QuantizeAndDequantizeV3 op, which allows using a tensor for the number o... | A. Unique TensorFlower | 2017-06-29 |
* | Fix bugs in ScatterNd and add ScatterNdNonAliasingAdd. | Eugene Brevdo | 2017-06-27 |
* | Merge changes from github. | A. Unique TensorFlower | 2017-05-18 |
* | This is to address a long standing issue (probably from day 1 of TensorFlow) ... | Yuan Yu | 2017-03-30 |
* | Add OnesLike op and kernel. | Suharsh Sivakumar | 2017-03-23 |
* | Add NoGradients for QuantizeAndDequantizeV2Op's input_min and input_max argum... | Suharsh Sivakumar | 2017-03-06 |
* | Merge changes from github. | Vijay Vasudevan | 2017-02-17 |
* | Make input_min and input_max Inputs of QuantizeAndDequantize rather than attr... | Suharsh Sivakumar | 2017-02-15 |
* | Better error message from trying to differentiate PreventGradient. | Alexandre Passos | 2017-02-07 |
* | Switch tf.concat_v2 references in third_party/tensorflow to tf.concat. | A. Unique TensorFlower | 2017-01-10 |
* | Add support for gradient calculation for concat_v2 with negative axis to fix ... | A. Unique TensorFlower | 2017-01-06 |
* | Adds the following new ops: | A. Unique TensorFlower | 2017-01-03 |
* | Allows int64 indices in gather_nd gradient. | A. Unique TensorFlower | 2016-12-22 |
* | Replace array_ops.unpack with array_ops.unstack in third_party/tensorflow. | A. Unique TensorFlower | 2016-12-16 |
* | Switch array_ops.pack/unpack to array_ops.stack/unstack. Also switch a few re... | A. Unique TensorFlower | 2016-12-15 |
* | Switch leftover reference to tf.concat to tf.concat_v2 in array_grad.py. | A. Unique TensorFlower | 2016-12-13 |
* | Change implementation of concat grad to use new split. Results in up to a 5x | A. Unique TensorFlower | 2016-12-12 |
* | Fix legacy Reverse gradient | Andrew Selle | 2016-12-08 |
* | Switch tf.concat references to tf.concat_v2 in array_grad.py. | A. Unique TensorFlower | 2016-12-08 |
* | Remove remaining uses of tf.reverse() and remove tf.reverse() | Andrew Selle | 2016-12-07 |
* | Rename more argument names to be consistent with NumPy. | Andrew Selle | 2016-11-30 |
* | Add concat_v2 to array_ops.py. Also, add gradient support for concat_v2 op. | A. Unique TensorFlower | 2016-11-21 |
* | Merge changes from github. | A. Unique TensorFlower | 2016-11-17 |
* | Expose tf.reverse_v2 which takes indices rather than a bool tensor. | Andrew Selle | 2016-11-15 |
* | Add a new op split_v that can handle variable size splits. | A. Unique TensorFlower | 2016-11-10 |
* | Adding CPU kernels for tf.scatter_nd(), tf.scatter_nd_update(), tf.scatter_nd... | A. Unique TensorFlower | 2016-11-02 |
* | Move SparseTensor into its own framework.py, add get_shape(). | Eugene Brevdo | 2016-11-01 |
* | Automated rollback of change 137564676 | A. Unique TensorFlower | 2016-10-28 |
* | Adding CPU kernels for tf.scatter_nd(), tf.scatter_nd_update(), tf.scatter_nd... | A. Unique TensorFlower | 2016-10-28 |
* | Update trace, matrix_set_diag, matrix_diag_part and their gradients to work f... | A. Unique TensorFlower | 2016-09-29 |
* | Add SpaceToBatchND and BatchToSpaceND ops for > 2 spatial dimensions. | A. Unique TensorFlower | 2016-09-21 |
* | 1. Rename batch_matrix_* ops to matrix_*. Specifically, this applies to | A. Unique TensorFlower | 2016-09-13 |
* | Rename NoGradient -> NotDifferentiable, to make it clear | Vijay Vasudevan | 2016-09-09 |
* | Remove NoGradient for ops which could have gradients defined but don't. | Mrinal Kalakrishnan | 2016-09-02 |
* | Fix gather gradient for empty slices | Geoffrey Irving | 2016-08-18 |
* | Updated the check_numerics function to also validate the gradient corresponding | Benoit Steiner | 2016-08-18 |
* | Automated rollback of change 130554783 | Geoffrey Irving | 2016-08-17 |
* | Add a BUILD rule for extract_image_patches_grad_test, and make the code py3 | Gunhan Gulsoy | 2016-08-17 |
* | Fix gather for nonempty indices, empty slices | Geoffrey Irving | 2016-08-17 |
* | Merge changes from github. | Benoit Steiner | 2016-08-12 |
* | Take advantage of the shape information whenever it is known at graph creation | Benoit Steiner | 2016-08-11 |
* | Broken up changes toward enabling strided slice | Andrew Selle | 2016-07-28 |
* | Added a gradient for the check_numeric operation. This makes it possible to | Benoit Steiner | 2016-07-26 |
* | Automated rollback of change 128378997 | Andrew Selle | 2016-07-25 |
* | Automated rollback of change 127492429 | Andrew Selle | 2016-07-25 |
* | Add BatchMatrixWithDiag op & gradient | Eugene Brevdo | 2016-07-13 |
* | Rename ellipse to ellipsis to avoid confusion | Andrew Selle | 2016-07-13 |
* | Implement gradient for StridedSlice op. | Andrew Selle | 2016-06-24 |