| Commit message (Expand) | Author | Age |
* | Add support for modeling fast memory close to the processor/gpu | A. Unique TensorFlower | 2018-10-09 |
* | Consolidate device parameter arguments into a shared DeviceInfo struct | A. Unique TensorFlower | 2018-10-08 |
* | Directly import tensor.proto.h (the transitive import will be removed from te... | Eugene Brevdo | 2018-08-23 |
* | Small fix to MaybeGetMinimumShape() in op_level_cost_estimator. | Doe Hyun Yoon | 2018-08-15 |
* | Add two counters in Costs Struct for number of ops processed/predicted in tot... | Peter Ma | 2018-08-10 |
* | Added QuantizedMatMul to OpLevelCostEstimator. | A. Unique TensorFlower | 2018-07-24 |
* | Support filter format for FusedConv2DBiasActivation. | Yao Zhang | 2018-06-20 |
* | Estimate Squeeze cost in the same way as Reshape. | Max Galkin | 2018-06-06 |
* | Expose MaybeGetMinimumShape for use in cost estimators other than OpLevelCost... | A. Unique TensorFlower | 2018-05-11 |
* | Fix a dropped line in the DepthwiseConv2dNative model | A. Unique TensorFlower | 2018-05-09 |
* | Add cost model of depthwiseConv2dNative. Tensorflow computes depthwise separa... | A. Unique TensorFlower | 2018-05-08 |
* | GuaranteeConst is a NoOp for the op_level_cost_estiamtor. | Max Galkin | 2018-05-05 |
* | Estimate IdentityN as Identity. | Max Galkin | 2018-04-17 |
* | Add analytical cost model for FusedConv2DBiasActivation. | Rob Sloan | 2018-04-06 |
* | TPU Cost Estimator has been modified to also account for the memory cost in t... | A. Unique TensorFlower | 2018-04-06 |
* | Re-enable Gather and Slice estimators with output size check. | Max Galkin | 2018-04-02 |
* | Fix a bug in AvgPoolGrad op cost in extracting input x's shape. AvgPoolGrad | A. Unique TensorFlower | 2018-04-02 |
* | Add op cost model for MaxPool, AvgPool, FusedBatchNorm, their grad ops, and | A. Unique TensorFlower | 2018-03-28 |
* | Disable new Gather/Slice estimators for now to fix the crashes during some TF... | Max Galkin | 2018-03-27 |
* | Shorter definitions for elementwise_ops in op_level_cost_estimator. | Max Galkin | 2018-03-21 |
* | Further improve accuracy of op_level_cost_estimator (Gather, GatherV2, Slice). | Max Galkin | 2018-03-21 |
* | Improved accuracy of op_level_cost_estimator (QuantizeV2, Dequantize, Gather). | Max Galkin | 2018-03-20 |
* | Add cost estimator tests for the BiasAdd, ReLU, and Conv2D operations. | A. Unique TensorFlower | 2018-02-22 |
* | Implement the logic to parse TensorProto (the tensor value for input or filte... | A. Unique TensorFlower | 2018-02-22 |
* | Extract the filter and input shape for Conv2DBackpropFilter/Conv2DBackpropInput | Benoit Steiner | 2018-02-13 |
* | Add more vlogging to op level estimator. | Yao Zhang | 2018-02-08 |
* | Add logging to diagnose device properties parsing problem in Grappler. | Max Galkin | 2018-02-05 |
* | Fixed the description of the fake GPU device to avoid a division by 0 | Benoit Steiner | 2018-02-01 |
* | Add op level memory usage estimation to the op_level_cost_estimator | Benoit Steiner | 2018-01-26 |
* | Estimate Placeholder as a no-op. | Max Galkin | 2017-12-05 |
* | Prefix inaccurate costs with "~" in VirtualScheduler verbose log. | Max Galkin | 2017-12-01 |
* | Some extra checks and warnings in grappler to provide more helpful error mess... | Max Galkin | 2017-11-23 |
* | Pass VirtualCluster to tf_optimizer to enable cross-optimization of models. | Yao Zhang | 2017-11-21 |
* | Improved encoding on shapes in grappler. | Benoit Steiner | 2017-11-06 |
* | Register a dummy estimation function for _Send ops. This ensures that we don'... | Benoit Steiner | 2017-10-12 |
* | Minor code cleanup in grappler cost estimation. | Max Galkin | 2017-10-10 |
* | Define OpContext and use it for OpLevelCostEstimator. | A. Unique TensorFlower | 2017-09-26 |
* | Merge changes from github. | Shanqing Cai | 2017-09-25 |
* | Remove extra checks from Grappler MatMul estimator. | A. Unique TensorFlower | 2017-08-04 |
* | Add option to use compute_memory_overlap; if true, use max of memory_cost and... | A. Unique TensorFlower | 2017-07-21 |
* | Set the op cost of RefIdentity, StopGradient, and PreventGradient to zero in ... | A. Unique TensorFlower | 2017-07-21 |
* | Improve the accuracy of the cost estimates for the size, shape, and rank ops. | Benoit Steiner | 2017-06-29 |
* | Prepare to remove a bunch of proto.h includes from tensorflow/core headers | Geoffrey Irving | 2017-06-29 |
* | Add a standalone cost analysis tool. Improved logging. | Yao Zhang | 2017-06-13 |
* | Simplified the virtual placer code | Benoit Steiner | 2017-06-13 |
* | Don't mark costs estimates of scalar tensors as potentially inaccurate. | Benoit Steiner | 2017-06-12 |
* | Initialize found_unknown_shapes variable. | A. Unique TensorFlower | 2017-06-12 |
* | Set inaccurate field when it should be false. | Yao Zhang | 2017-06-09 |
* | Estimate cost for element wise ops and a minimum compute cost for dummy execu... | Yao Zhang | 2017-06-09 |
* | Profile memory usage in VirtualScheduler and report peak memory usage. | A. Unique TensorFlower | 2017-06-06 |