Merge branch 'update_deploying_docs' into 'master'

docs: Update deploying-models.

See merge request ai/esp-dl!35

README.md

@@ -36,7 +36,7 @@ ESP-DL provides some model APIs in the [Model Zoo](./include/model_zoo), such as
 
 ## Deploy Your Own Models
 
-We recommend using TVM to deploy your models, for details please refer to [ESP-DL/tutorial/tvm_example](tutorial/tvm_example).
+For details please refer to [Manual Model Quantization and Deployment Guide](https://docs.espressif.com/projects/esp-dl/en/latest/esp32s3/tutorials/deploying-models.html).
 
 
 ## Feedback

README_cn.md

@@ -38,7 +38,7 @@ ESP-DL 在 [模型库](./include/model_zoo) 中提供了一些模型的 API，
 
 ## 部署你的模型
 
-我们推荐使用 TVM 来部署你的模型，具体可参考 [ESP-DL/tutorial/tvm_example](tutorial/tvm_example)。
+具体可参考 [手动部署模型](https://docs.espressif.com/projects/esp-dl/zh_CN/latest/esp32s3/tutorials/deploying-models.html)。
 
 
 ## 反馈

docs/en/tutorials/deploying-models-through-tvm.rst

@@ -5,6 +5,7 @@ Auto-Generating Model Deployment Project using TVM
 :link_to_translation:`zh_CN:[中文]`
 
 This case introduces the complete process of deploying a model with TVM.
+**This project is based on the TVM v0.14.0 branch and is in an experimental state, with no plans for further iteration or maintenance at the moment. Currently, only the conv2d operator from ESP-DL is integrated, and other operators may cause exceptions.**
 
 Preparation
 -----------

docs/en/tutorials/deploying-models.rst

@@ -21,6 +21,13 @@ Step 1: Convert Your Model
 
 In order to be deployed, the trained floating-point model must be converted to an integer model, the format compatible with ESP-DL. Given that ESP-DL uses a different quantization scheme and element arrangements compared with other platforms, please convert your model with our :doc:`Quantization Toolkit Overview <../../tools/quantization-toolkit/quantization-toolkit-overview>`.
 
+Currently, the operators supported by ESP-DL are as follows:
+
+Conv2d, Gemm, Relu, AvgPool2d, MaxPool2d, Add, Mul, Sub, Softmax, Tanh,
+Sigmoid, Concat, Expand, Flatten, Max, Min, Pad, Reshape, Squeeze, Transpose,
+Slice, Shape, Resize
+
+
 Step 1.1: Convert to ONNX Format
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 

docs/zh_CN/tutorials/deploying-models-through-tvm.rst

@@ -5,6 +5,7 @@
 :link_to_translation:`en:[English]`
 
 本案例介绍了使用 TVM 部署模型的完整流程。
+**该项目基于 TVM v0.14.0 分支，处于试验状态，暂时没有进一步迭代维护的计划。目前仅对接了 ESP-DL 的 conv2d 算子，其它算子可能会引发异常。**
 
 准备
 ----

docs/zh_CN/tutorials/deploying-models.rst

@@ -21,6 +21,13 @@
 
 为了部署模型，必须将训练好的浮点模型转换为 ESP-DL 适配的整型模型格式。由于本库使用的量化方式和参数排列方式与一些平台不同，请使用我们提供的工具 :doc:`量化工具包 <../../tools/quantization-toolkit/quantization-toolkit-overview>` 来完成转换。
 
+目前 ESP-DL 支持的算子如下：
+
+Conv2d, Gemm, Relu, AvgPool2d, MaxPool2d, Add, Mul, Sub, Softmax, Tanh,
+Sigmoid, Concat, Expand, Flatten, Max, Min, Pad, Reshape, Squeeze, Transpose,
+Slice, Shape, Resize
+
+
 步骤 1.1：转换为 ONNX 格式模型
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~