Update docs

microsoft · Dec 10, 2024 · 46e7b10 · 46e7b10
1 parent 27754e1
commit 46e7b10
Show file tree

Hide file tree

Showing 5 changed files with 16 additions and 9 deletions.
diff --git a/docs/source/how-to/configure-workflows/pass/convert-onnx.md b/docs/source/how-to/configure-workflows/pass/convert-onnx.md
@@ -61,19 +61,26 @@ b. More fine-grained control of the conversion conditions is also possible:
 
 See [Float16 Conversion](https://onnxruntime.ai/docs/performance/model-optimizations/float16.html#float16-conversion) for more detailed description of the available configuration parameters.
 
-## Inputs/Outputs Float16 to Float32 Conversion
+## Inputs/Outputs DataType Conversion
 
-Certain environments such as Onnxruntime WebGPU prefers Float32 logits. The `OnnxIOFloat16ToFloat32` pass converts the inputs and outputs to use Float32 instead of Float16.
+In certain environments, such as Onnxruntime WebGPU, Float32 logits are preferred. The `OnnxIODataTypeConverte` pass enables conversion of model inputs and outputs to a specified data type. This is particularly useful for converting between data types such as Float16 and Float32, or any other supported ONNX data types.
 
 ### Example Configuration
 
-a. The most basic configuration, which is suitable for many models, leaves all configuration options set to their default values:
+The simplest configuration converts all inputs and outputs from Float16 (source_dtype = 10) to Float32 (target_dtype = 1), which is suitable for many models:
+
 ```json
 {
-    "type": "OnnxIOFloat16ToFloat32"
+    "type": "OnnxIODataTypeConverter",
+    "source_dtype": 10,
+    "target_dtype": 1
 }
 ```
 
+### Datatype Mapping
+
+The `source_dtype` and `target_dtype` are integers corresponding to ONNX data types. You can find the complete mapping in the ONNX protobuf definition [here](https://github.com/onnx/onnx/blob/96a0ca4374d2198944ff882bd273e64222b59cb9/onnx/onnx.proto3#L503-L551).
+
 ## Mixed Precision Conversion
 Converting model to mixed precision.
 

diff --git a/docs/source/reference/pass.rst b/docs/source/reference/pass.rst
@@ -43,9 +43,9 @@ OnnxFloatToFloat16
 
 .. _onnx_io_float16_to_float32:
 
-OnnxIOFloat16ToFloat32
+OnnxIODataTypeConverter
 ----------------------
-.. autoconfigclass:: olive.passes.OnnxIOFloat16ToFloat32
+.. autoconfigclass:: olive.passes.OnnxIODataTypeConverter
 
 .. _ort_mixed_precision:
 

diff --git a/examples/phi2/phi2.py b/examples/phi2/phi2.py
@@ -156,7 +156,7 @@ def main(raw_args=None):
             template_json["systems"]["local_system"]["accelerators"] = [
                 {"device": "GPU", "execution_providers": ["JsExecutionProvider"]}
             ]
-            fl_type = {"type": "OnnxIOFloat16ToFloat32"}
+            fl_type = {"type": "OnnxIODataTypeConverte"}
             template_json["passes"]["fp32_logits"] = fl_type
         new_json_file = "phi2_web.json"
         with open(new_json_file, "w") as f:

diff --git a/examples/phi3/phi3_template.json b/examples/phi3/phi3_template.json
@@ -95,7 +95,7 @@
         "merge_adapter_weights": { "type": "MergeAdapterWeights" },
         "awq": { "type": "AutoAWQQuantizer" },
         "builder": { "type": "ModelBuilder", "precision": "<place_holder>" },
-        "fp32_logits": { "type": "OnnxIOFloat16ToFloat32" },
+        "fp32_logits": { "type": "OnnxIODataTypeConverte" },
         "tune_session_params": {
             "type": "OrtSessionParamsTuning",
             "data_config": "gqa_transformer_prompt_dummy_data",

diff --git a/olive/cli/auto_opt.py b/olive/cli/auto_opt.py
@@ -440,7 +440,7 @@ def _get_passes_config(self, config: Dict[str, Any], olive_config: OlivePackageC
             ),
             ("peephole_optimizer", {"type": "OnnxPeepholeOptimizer"}),
             # change io types to fp32
-            ("fp16_to_fp32", {"type": "OnnxIOFloat16ToFloat32"}),
+            ("fp16_to_fp32", {"type": "OnnxIODataTypeConverte"}),
             # qnn preparation passes
             ("to_fixed_shape", {"type": "DynamicToFixedShape", "dim_param": None, "dim_value": None}),
             ("qnn_preprocess", {"type": "QNNPreprocess"}),