Jacinto 7 TIDL Release Notes
Version: 02.00.00
Contents
- Introduction
- Licensing
- What's New
- Documentation
- Upgrade and Compatibility Information
- Device Support
- Validation Information
- Fixed Issues
- Known Issues
- Technical Support
- Package Versioning
Introduction
The package consists of the Texas Instruments J7 TIDL Software
Licensing
The licensing information of this library, as well as any third-party components included which are made available under a number of other open-source licenses are enumerated as part of the manifest. Note : Below Links would work only after installing the package
Documentation
| 1 | User's Guide | [HTML] |
| 3 | Software Manifest | [HTML] |
What's New
Here are a few of the new features supported in this release:
- Added ONNX Runtime support
- Mixed Precision Quantization (parameters/activations in 16bit) for user selected layers
Upgrade and Compatibility Information
NA
Device Support
| Target SoC | Target Plaform | Build environment (OS) | Target environment (OS) |
| J721E EVM | x86_64 Linux PC | x86_64 (Linux) | NA |
| J721E EVM | x86_64 Windows PC | x86_64 Windows | NA |
Validation Information
This release was built and validated using the following tools:
Dependencies (included in Processor SDK RTOS Automotive):
- ivision - 01_00
- MMALIB - 02_00_00_02
- PDK - CORE SDK 07.03.00
- C7x CGT - ti-cgt-c7000_1.4.2.LTS
Dependencies (NOT included in Processor SDK RTOS Automotive):
- Microsoft Visual Studio Verison 14.2
- GCC version 5.4.0 (In Ubuntu 18.04)
- OpenCV: 3.1.0 (for Test bench Only)
- Protobuf: 3.11.3 (for Import tool only)
- Flatbuffers : 1.12.0 (for Import tool only)
Fixed Issues
| ID | Description | Module | Affected Versions | Affected Platforms |
|---|---|---|---|---|
| TIDL-1211 | Detection output import is failing with 7 prior boxes | TOOL | 01.03.00 | J721E |
| TIDL-1258 | Tensorflow Argmax operator not supported | TOOL | 01.03.00 | J721E |
| TIDL-1271 | CSL files to be removed from TIDL | PACKAGE | 01.04.00 | J721E |
| TIDL-1302 | TFlite dilaton factor is not read correctly | TOOL | 01.04.00 | J721E |
| TIDL-1323 | Leaky Relu Layer output scale is coming as large value | COMPUTE | 01.04.00 | J721E |
| TIDL-1324 | Sigmoid layer output mismatch on EVM | COMPUTE | 01.04.00 | J721E |
| TIDL-1326 | Sum of softmax output don't equal to 1 | COMPUTE | 01.04.00 | J721E |
| TIDL-1327 | Deconvolution layer output mismatching between EVM and PC emulation | COMPUTE | 01.04.00 | J721E |
| TIDL-1328 | Pad amount is not calculated correctly for pooling layer | COMPUTE | 01.04.00 | J721E |
| TIDL-1330 | Onnx opset 11 pad layer parameters not read correctly | TOOL | 01.04.00 | J721E |
| TIDL-1331 | Logistic function sigmoid bug in tidl_j7_01_04_01_01 | COMPUTE | 01.04.00 | J721E |
| TIDL-1332 | In 16-bit Leaky relu the output is not correct for negative input values | COMPUTE | 01.04.00 | J721E |
| TIDL-1334 | Custom Layer : Build and Integration documentation missing | DOCUMENTATION | 01.04.00 | J721E |
| TIDL-1337 | TDA4: MaxPool layer outputs all zeros during execution on EVM | COMPUTE | 01.04.00 | J721E |
| TIDL-1339 | Nearest neighbor resize not supported in tflite runtime | TOOL | 01.04.00 | J721E |
| TIDL-1340 | Issue with EfficientDet-Lite4 performance simulation | TOOL | 01.04.00 | J721E |
| TIDL-1341 | Tflite runtime : Add support for tflite tensor type int8 in rt_invoke | TOOL | 01.04.00 | J721E |
Known Issues
| ID | Description | Module | Reported in Release | Affected Platforms | Occurrence | Workaround in this release |
|---|---|---|---|---|---|---|
| TIDL-1333 | ONNX Opset 11 with "sizes" as input tensor is not working | TOOL | 01.04.00 | J721E | Rare | Instead of "sizes" use "scales" as input tensor |
| TIDL-1338 | TDA4VM: GobalAvgPool layer is returning error for customer model | COMPUTE | 01.04.00 | J721E | Rare | NA |
| TIDL-864 | Concat trace dump mismatch between REF & CN/CI | COMPUTE | 01.02.00 | J721E | Rare | NA |
| TIDL-1197 | Model execution order for a branch is not optimal | COMPUTE | 01.02.00 | J721E | Rare | No fucntional issue, Only Performance drop would be obsrved |
| TIDL-1269 | TF maxpool avgpool functionality mismatch for k=3,s=2 | COMPUTE | 01.03.00 | J721E | Rare | NA |
| TIDL-1208 | ERR_MMEORY_INSERTION in customer model | TOOL | 01.02.00 | J721E | Rare | NA |
| TIDL-705 | Max Pooling Output is not correct with signed input | COMPUTE | 01.00.01 | J721E | Rare | NA |
Technical Support
For technical support, please post your questions on TI E2E Forum for Automotive ADAS SoCs.
For additional assistance, contact local TI Field Application Engineer
Package Versioning
Each package version is composed of 4 period-delimited numbers - represented here by the letters M, m, p and b [M.m.p.b]. The table below provides a descriptive reference regarding package version numbering.
| Digit | Meaning | Description |
|---|---|---|
| 1 (M=Major) | Major revision | Incremented when the new version is substantially different from the previous For example, a new module added or an existing module's algorithm significantly altered. |
| 2 (m=minor) | Minor revision | Incremented when the new version has changed but not in a major way. For example, some minor changes in the API or feature set. |
| 3 (p=patch) | Patch number | Incremented for all other source code changes. This include any packaging support code. |
| 4 (b=build) | Build number | Incremented for each release delivery to CM. Reset for any change to M, m or p |
Copyright 2018, Texas Instruments Incorporated