Sumit Agarwal
Sumit Agarwal
**Description**: Today, DML EP is special from other EPs because: - During Partitioning, DML EP fuses it's subgraph on its own and does not use [IExecutionProvider::Compile()](https://github.com/microsoft/onnxruntime/blob/master/include/onnxruntime/core/framework/execution_provider.h#L249) method, unlike other EPs....
### Describe the issue CPU EP implementation of LayerNormalization Contrib Operator does not broadcast the Scale and Bias Tensor and it ends up [reading value outside of memory bound](https://github.com/microsoft/onnxruntime/blob/main/onnxruntime/contrib_ops/cpu/layer_norm.cc#L119-L125). Unfortunately...
**Description**: Describe your changes. **Motivation and Context** - Why is this change required? What problem does it solve? - If it fixes an open issue, please link to the issue...
### Description This PR adds a feature to serialize all DML EP partitions into DML currency individually for a given a model. This feature can be dynamically turned on by...
I am looking for a tool that can export GGML graph directly to a MLIR may with any existing dialect or a custom dialect.