docs: Update ML library versions and documentation#651
docs: Update ML library versions and documentation#651PrathamTI wants to merge 1 commit intoTexasInstruments:masterfrom
Conversation
This commit updates the ML library versions and documentation - Update ARM Compute Library from 24.12 to 52.7.0 - Update Arm NN from 24.11 to 26.01 - Update NNStreamer from 2.4.2 to 2.6.0 - Update ONNX Runtime from 1.20.1 to 1.23.2 - Update TensorFlow Lite from 2.18.0 to 2.20.0 - Refresh all test outputs and benchmark results - Add ML components to AM62DX documentation TOC - Update component table with latest library information Signed-off-by: Pratham Deshmukh <p-deshmukh@ti.com>
PrathamTI
requested review from
StaticRocket,
VeeruPrudhvi,
cshilwant,
gehariprasath,
jeevantelukula,
praneethbajjuri and
uditkumarti
as code owners
github-actions
bot
assigned
devarsht,
gehariprasath,
praneethbajjuri,
r-vignesh and
uditkumarti
cshilwant
left a comment
cshilwant
left a comment
There was a problem hiding this comment.
Did you build test this patch locally & checked the HTML outputs? - https://github.com/TexasInstruments/processor-sdk-doc?tab=readme-ov-file#issue-make
There was a problem hiding this comment.
ONNX-RT should be a separate block under Runtime Engines.
Currently, it's stacked up with ACL & ArmNN which may send out a false message that they are tightly coupled.
There was a problem hiding this comment.
Ok I'll fix the diagram
There was a problem hiding this comment.
The image size doubled.
Older one - 40.3 KB
New one - 71.7 KB
Any idea why?
Also, any reasons to switch from png to jpeg?
There was a problem hiding this comment.
No reason as such. While generating the image, image format might have changed. I'll fix this too.
| * Available on AM57x device only. | ||
| * Supports CNN at the moment, and imports Caffe, ONNX, and Tensorflow models. | ||
| * Open source framework based on GStreamer for neural network pipelines. | ||
| * Runs on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices). |
There was a problem hiding this comment.
AM5x SDKs won't include the meta-ti-ml layer
| * Runs on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices). | |
| * Runs on all Cortex-A ARM cores (AM3x, AM4x, AM6x Sitara devices). |
| | Arm Compute Library | 52.7.0 | NA (Backend Library) | Yes | YES | | ||
| +--------------------------+-----------+-------------------------+--------------------+--------------------+ | ||
| | NNStreamer | 2.6.0 | NA (Pipeline Framework) | Yes | YES | | ||
| +--------------------------+-----------+-------------------------+--------------------+--------------------+ |
There was a problem hiding this comment.
In this table, it's Yes at few places & YES in some.
Keep it consistent...switch to Yes to make vale happy
Yeah I tested it locally first. I'll update the PR according to your comments |
7 participants
This commit updates the ML library versions and documentation