diff --git a/src/images/ONNX-Dark.svelte b/src/images/ONNX-Dark.svelte new file mode 100644 index 0000000000000..370718d8a09ff --- /dev/null +++ b/src/images/ONNX-Dark.svelte @@ -0,0 +1,33 @@ + + + diff --git a/src/images/ONNX-Light.svelte b/src/images/ONNX-Light.svelte new file mode 100644 index 0000000000000..cf55b4ed8b8df --- /dev/null +++ b/src/images/ONNX-Light.svelte @@ -0,0 +1,33 @@ + + + diff --git a/src/images/undraw/image_largemodeltraining.svelte b/src/images/undraw/image_largemodeltraining.svelte index 03180553de67f..0113740420c55 100644 --- a/src/images/undraw/image_largemodeltraining.svelte +++ b/src/images/undraw/image_largemodeltraining.svelte @@ -4,6 +4,7 @@ import { base } from '$app/paths'; - import OnnxIcon from '../images/onnx-icon.svelte'; + import OnnxLight from '../images/ONNX-Light.svelte';

4

- +

4

diff --git a/src/routes/components/code-blocks.svelte b/src/routes/components/code-blocks.svelte index fac3298d6dfdf..d7285c31954e3 100644 --- a/src/routes/components/code-blocks.svelte +++ b/src/routes/components/code-blocks.svelte @@ -53,24 +53,26 @@

Use ONNX Runtime with your favorite language, or dive right into the tutorials.

- + Quickstart TutorialsTutorials Install ONNX RuntimeInstall ONNX Runtime Hardware accelerationHardware acceleration +
ONNXRuntime
- +
diff --git a/src/routes/components/hero.svelte b/src/routes/components/hero.svelte index 772c3ff328e65..ed02a89795804 100644 --- a/src/routes/components/hero.svelte +++ b/src/routes/components/hero.svelte @@ -6,7 +6,8 @@ import { Highlight } from 'svelte-highlight'; import { bash } from 'svelte-highlight/languages'; import FaRegClipboard from 'svelte-icons/fa/FaRegClipboard.svelte'; - import OnnxIcon from '../../images/onnx-icon.svelte'; + import OnnxLight from '../../images/ONNX-Light.svelte'; + import OnnxDark from '../../images/ONNX-Dark.svelte'; import { fade, fly, blur } from 'svelte/transition'; import { quintOut } from 'svelte/easing'; @@ -83,7 +84,7 @@

In a rush? Get started easily:

-
+
@@ -94,7 +95,7 @@ >
-
+
@@ -110,13 +111,13 @@ More interested in training? More info here.

-->

- Don't see your favourite platform? See the many others we support →Don't see your favorite platform? See the many others we support →

-
diff --git a/src/routes/components/landing-hero.svelte b/src/routes/components/landing-hero.svelte index 562bf7e882b72..28d403259d007 100644 --- a/src/routes/components/landing-hero.svelte +++ b/src/routes/components/landing-hero.svelte @@ -1,11 +1,9 @@
@@ -20,7 +18,7 @@
{#if imgsrc == 'onnxruntimelogo'}
- +
{:else} {imgalt} diff --git a/src/routes/components/performance.svelte b/src/routes/components/performance.svelte index e87f0f98e8732..216b0518866f1 100644 --- a/src/routes/components/performance.svelte +++ b/src/routes/components/performance.svelte @@ -8,13 +8,12 @@

Performance



- ONNX Runtime provides high performance for running deep learning models on a range of - hardwares. Based on usage scenario requirements, latency, throughput, memory utilization, - and model/application size are common dimensions for how performance is measured. + ONNX Runtime runs performantly on a range of hardware, excelling in performance dimensions + including latency, throughput, memory utilization, and model and application size.

- While ORT out-of-box aims to provide good performance for the most common usage patterns, there - are model optimization techniques and runtime configurations that can be utilized to improve - performance for specific use cases and models.. + In addition to high performance out-of-box for the most common usage patterns, additional + model optimization techniques and runtime configurations can further improve performance for specific + use cases and models.

diff --git a/src/routes/components/training-and-inference.svelte b/src/routes/components/training-and-inference.svelte index 8ccda35be736b..78023cb7a54b9 100644 --- a/src/routes/components/training-and-inference.svelte +++ b/src/routes/components/training-and-inference.svelte @@ -13,32 +13,32 @@

Learn more about ONNX Runtime Training →
-
-
-
-

Large Model Training

+
+
+
+

Large Model Training

- ORTModule accelerates training of large transformer PyTorch models. The training time - and cost are reduced with a single line code change. It is built on top of highly - successful and proven technologies of ONNX Runtime and ONNX format. + ORTModule accelerates training of large transformer PyTorch models, reducing training + time and cost with a single line code change.

-
-
+
-
-

On-Device Training

+
+

On-Device Training

On-Device Training refers to the process of training a model on an edge device, such as - mobile phones, embedded devices, gaming consoles, web browsers, etc. This is in contrast - to training a model on a server or a cloud. + mobile phones, embedded devices, gaming consoles, web browsers, etc. This is useful for + when performance, connectivity, or privacy is a consideration and server-based training + is not an option.

- @@ -53,30 +53,29 @@

Learn more about ONNX Runtime Inferencing →
-
-
-
-

ONNX Runtime Web

+
+
+
+

ONNX Runtime Web

ONNX Runtime Web allows JavaScript developers to run and deploy machine learning models in browsers.

-
-
+
-
-

ONNX Runtime Mobile

+
+

ONNX Runtime Mobile

- ONNX Runtime Mobile allows you to run model inferencing on mobile devices (iOS and - Android). + ONNX Runtime Mobile allows you to run model inferencing on mobile devices.

- diff --git a/src/routes/generative-ai/+page.svelte b/src/routes/generative-ai/+page.svelte index 63b0e87bd5a00..cd5f2b5f244a6 100644 --- a/src/routes/generative-ai/+page.svelte +++ b/src/routes/generative-ai/+page.svelte @@ -2,7 +2,7 @@ import LandingHero from '../components/landing-hero.svelte'; const title = 'Generative AI + ONNX Runtime'; const description = - 'Integrate the power of generative AI in your apps and services with ONNX Runtime. Broad platform support and deep optimizations enable you to benefit from state of the art models for image synthesis, text generation, and more.'; + 'Integrate the power of generative AI in your apps and services with ONNX Runtime. Broad platform support and deep optimizations empower usage of state-of-the-art models for image synthesis, text generation, and more.'; const imgsrc = 'onnxruntimelogo'; const imgalt = 'ONNX Runtime Logo'; import stablediffusion1 from '../../images/StableDiffusion1.webp'; @@ -12,15 +12,12 @@

Stable Diffusion + ONNX Runtime

-

- Using ONNX Runtime with this hugely popular text-to-image latent diffusion model for image - generation can be hugely beneficial. -

+

Use ONNX Runtime to accelerate this popular image generation model.

Benefits

-

Run Stable Diffusion outside of a Python environment

+

Run Stable Diffusion outside of a Python environment

-

Speed up inference of Stable Diffusion on Nvidia and AMD GPUs

+

+ Speed up inference of Stable Diffusion on NVIDIA and AMD GPUs +

Large Language Models + ONNX Runtime

- ONNX Runtime also supports many increasingly popular large language model (LLM) families. - Several of these model families are available in the Hugging Face Model Hub with hundreds, or - even thousands, of models that are easily convertible to ONNX using the Optimum API: + ONNX Runtime supports many popular large language model (LLM) families in the Hugging Face Model + Hub. These, along with thousands of other models, are easily convertible to ONNX using the + Optimum API.

LLaMA → diff --git a/src/routes/inference/+page.svelte b/src/routes/inference/+page.svelte index dc8e6da06363a..04d7b671f18a8 100644 --- a/src/routes/inference/+page.svelte +++ b/src/routes/inference/+page.svelte @@ -2,7 +2,8 @@ import LandingHero from '../components/landing-hero.svelte'; import ImageInference2 from '../../images/undraw/image_inference2.svelte'; import ImageInference1 from '../../images/undraw/image_inference1.svg'; - import OnnxIcon from '../../images/onnx-icon.svelte'; + import OnnxLight from '../../images/ONNX-Light.svelte'; + const title = 'ONNX Runtime for Inferencing'; const description = 'ONNX Runtime Inference powers machine learning models in key Microsoft products and services across Office, Azure, Bing, as well as thousands of community projects.'; @@ -24,7 +25,7 @@ >
- +
diff --git a/src/routes/pytorch/+page.svelte b/src/routes/pytorch/+page.svelte index edc2ac70cf73b..d1366ab24b4e8 100644 --- a/src/routes/pytorch/+page.svelte +++ b/src/routes/pytorch/+page.svelte @@ -63,7 +63,7 @@

optimized deployment + > optimized deployment →
@@ -111,7 +111,7 @@ platforms.

- Performance enhancements → @@ -132,7 +132,7 @@ days.

- Testimonials → diff --git a/src/routes/testimonials/testimonial-card.svelte b/src/routes/testimonials/testimonial-card.svelte index aa4647859662a..63af84b304bbf 100644 --- a/src/routes/testimonials/testimonial-card.svelte +++ b/src/routes/testimonials/testimonial-card.svelte @@ -34,7 +34,7 @@

{title}

{description}


-

-{author}

+

-{author}

diff --git a/src/routes/training/+page.svelte b/src/routes/training/+page.svelte index f1b0a7fca0d95..5ddcc2279052f 100644 --- a/src/routes/training/+page.svelte +++ b/src/routes/training/+page.svelte @@ -56,7 +56,7 @@ class="text-blue-500" href="https://learn.microsoft.com/en-us/azure/machine-learning/resource-azure-container-for-pytorch?view=azureml-api-2" >Azure Container for PyTorch (ACPT) and seamlessly integrates with existing training pipelines for PyTorch models​. + > and seamlessly integrates with existing training pipelines for PyTorch models.

@@ -64,8 +64,8 @@

Flexible & extensible hardware support

- The same model and API works with NVIDIA and AMD GPUs, and the extensible "execution - provider" architecture allow you to plug-in custom operators, optimizer and hardware + The same model and API works with NVIDIA and AMD GPUs, and the extensible execution + provider architecture allow you to plug-in custom operators, optimizer and hardware accelerators.

@@ -80,7 +80,7 @@

Composable with popular acceleration systems

- Compose with Deepspeed, Fairscale, Megatron, and more for even faster and more efficient + Compose with DeepSpeed, FairScale, Megatron, and more for even faster and more efficient training.

@@ -194,7 +194,7 @@
-

Improving data privacy & security

+

Improves data privacy & security

especially when working with sensitive data that cannot be shared with a server or a cloud

diff --git a/src/routes/windows/+page.svelte b/src/routes/windows/+page.svelte index 7466c6a6c5569..7c0cf855eb74a 100644 --- a/src/routes/windows/+page.svelte +++ b/src/routes/windows/+page.svelte @@ -1,6 +1,6 @@