From 6f9c0a2fe7b175be2c55e867355377a7d7298dc8 Mon Sep 17 00:00:00 2001 From: Hector Li Date: Mon, 18 Mar 2024 15:58:17 -0700 Subject: [PATCH] Add option to enable FP16 precision for QNN EP (#19971) add option to enable FP16 precision for QNN EP --- docs/execution-providers/QNN-ExecutionProvider.md | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/docs/execution-providers/QNN-ExecutionProvider.md b/docs/execution-providers/QNN-ExecutionProvider.md index 61d0aae21c..26a963a705 100644 --- a/docs/execution-providers/QNN-ExecutionProvider.md +++ b/docs/execution-providers/QNN-ExecutionProvider.md @@ -114,6 +114,11 @@ Alternatively to setting profiling_level at compile time, profiling can be enabl |---|---| |Device ID (string)|The ID of the device to use when setting `htp_arch`. Defaults to "0" (for single device).| +|`"enable_htp_fp16_precision"`|Description [Example](https://github.com/microsoft/onnxruntime-inference-examples/tree/main/c_cxx/QNN_EP/mobilenetv2_classification)| +|---|---| +|'0'|default.| +|'1'|Enable the float32 model to be inferenced with fp16 precision.| + ## Supported ONNX operators |Operator|Notes|