microsoft · leestott · Jul 9, 2024 · Jul 8, 2024 · Jul 8, 2024 · Jul 8, 2024
diff --git a/README.md b/README.md
@@ -1,4 +1,4 @@
-# Welcome to Microsoft Phi-3 Cookbook
+# Phi-3 Cookbook: Hands-On Examples with Microsoft's Phi-3 Models
 
 [![Open in GitHub Codespaces](https://github.com/codespaces/badge.svg)](https://codespaces.new/microsoft/phi-3cookbook)
 [![Open in Dev Containers](https://img.shields.io/static/v1?style=for-the-badge&label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/microsoft/phi-3cookbook)
@@ -14,75 +14,11 @@
 
 [![](https://dcbadge.vercel.app/api/server/ByRwuEEgH4)](https://discord.com/invite/ByRwuEEgH4?WT.mc_id=aiml-137032-kinfeylo)
 
-
-Phi-3, is a family of open AI models developed by Microsoft. Phi-3 models are the most capable and cost-effective small language models (SLMs) available, outperforming models of the same size and next size up across a variety of language, reasoning, coding, and math benchmarks. Here is a manual on how to use the Microsoft Phi-3 family.
+Phi-3, is a family of open AI models developed by Microsoft. Phi-3 models are the most capable and cost-effective small language models (SLMs) available, outperforming models of the same size and next size up across a variety of language, reasoning, coding, and math benchmarks. The Phi-3 Family includes mini, small, medium and vision versions, trained based on different parameter amounts to serve various application scenarios. For more detailed information about Microsoft's Phi-3 family, please visit the [Welcome to the Phi-3 Family](/md/01.Introduce/Phi3Family.md) page.
 
 ![Phi3Family](/imgs/00/Phi3getstarted.png)
 
-## Microsoft's Phi-3 family
-
-Phi-3 models significantly outperform language models of the same and larger sizes on key benchmarks (see benchmark numbers below, higher is better). Phi-3-mini does better than models twice its size, and Phi-3-small and Phi-3-medium outperform much larger models, including GPT-3.5T.
-
-All reported numbers are produced with the same pipeline to ensure that the numbers are comparable. As a result, these numbers may differ from other published numbers due to slight differences in the evaluation methodology. More details on benchmarks are provided in our technical paper.
-
-### Phi-3-mini
-
-Phi-3-mini, a 3.8B language model is available on [Microsoft Azure AI Studio](https://aka.ms/phi3-azure-ai), [Hugging Face](https://huggingface.co/collections/microsoft/phi-3-6626e15e9585a200d2d761e3), and [Ollama](https://ollama.com/library/phi3).
-
-![phi3modelminibenchmark](/imgs/00/phi3minibenchmark.png)
-
-![phi3modelminibenchmark128k](/imgs/00/phi3minibenchmark128.png)
-
-### Phi-3-small
-
-Phi-3-small with only 7B parameters beats GPT-3.5T across a variety of language, reasoning, coding, and math benchmarks.
-
-![phi3modelsmall](/imgs/00/phi3smallbenchmark.png)
-
-![phi3modelsmall128k](/imgs/00/phi3smallbenchmark128.png)
-
-
-### Phi-3-medium
-
-Phi-3-medium with 14B parameters continues the trend and outperforms Gemini 1.0 Pro.
-
-![phi3modelmedium](/imgs/00/phi3mediumbenchmark.png)
-
-![phi3modelmedium128k](/imgs/00/phi3mediumbenchmark128.png)
-
-### Phi-3-vision
-
-Phi-3-vision with just 4.2B parameters continues that trend and outperforms larger models such as Claude-3 Haiku and Gemini 1.0 Pro V across general visual reasoning tasks, OCR, table and chart understanding tasks.
-
-![phi3modelvision](/imgs/00/phi3visionbenchmark.png)
-
-> **Note**
->
-> Phi-3 models do not perform as well on factual knowledge benchmarks (such as TriviaQA) as the smaller model size results in less capacity to retain facts.
-
-### Phi-Silica
-
-We are introducing Phi Silica which is built from the Phi series of models and is designed specifically for the NPUs in Copilot+ PCs. Windows is the first platform to have a state-of-the-art small language model (SLM) custom built for the NPU and shipping inbox. Phi Silica API along with OCR, Studio Effects, Live Captions, and Recall User Activity APIs will be available in Windows Copilot Library in June. More APIs like Vector Embedding, RAG API, and Text Summarization will be coming later.
-
-## Phi-3 on Azure AI Studio
-
-You can learn how to use Microsoft Phi-3 and how to build E2E solutions in your different hardware devices. To experience Phi-3 for yourself, start by playing with the model and customizing Phi-3 for your scenarios using the [Azure AI Studio, Azure AI Model Catalog](https://aka.ms/phi3-azure-ai)
-
-**Playground**
-Each model has a dedicated playground to test the model [Azure AI Playground](https://aka.ms/try-phi3).
-
-## Phi-3 on Hugging Face
-
-You can also find the model on the [Hugging Face](https://huggingface.co/microsoft)
-
-**Playground**
- [Hugging Chat playground](https://huggingface.co/chat/models/microsoft/Phi-3-mini-4k-instruct)
-
-## Contents
-
-This cookbook includes:
-
-## **Microsoft Phi-3 Cookbook**
+## Table of Contents
 
 - [Introduction]()
   - [Setting up your environment](./md/01.Introduce/EnvironmentSetup.md)(✅)
@@ -159,13 +95,30 @@ This cookbook includes:
   - [Run C# Phi-3 samples in a CodeSpace](./md/07.Labs/CsharpOllamaCodeSpaces/CsharpOllamaCodeSpaces.md)(✅)
   - [Using Phi-3 with Promptflow and Azure AI Search](./code/07.Lab/RAG_with_PromptFlow_and_AISearch/README.md)(✅)
 
-## Multi-language support
+## Using Phi-3 Models
+
+### Phi-3 on Azure AI Studio
+
+You can learn how to use Microsoft Phi-3 and how to build E2E solutions in your different hardware devices. To experience Phi-3 for yourself, start by playing with the model and customizing Phi-3 for your scenarios using the [Azure AI Studio, Azure AI Model Catalog](https://aka.ms/phi3-azure-ai)
+
+**Playground**
+Each model has a dedicated playground to test the model [Azure AI Playground](https://aka.ms/try-phi3).
+
+### Phi-3 on Hugging Face
+
+You can also find the model on the [Hugging Face](https://huggingface.co/microsoft)
+
+**Playground**
+ [Hugging Chat playground](https://huggingface.co/chat/models/microsoft/Phi-3-mini-4k-instruct)
+
+## Multi-Language Support
 
-- [閱讀正體中文](./translations/zh-tw/README.md) (Translator - **Microsoft MVP & Microsoft Regional Director** [@doggy8088](https://github.com/doggy8088))
-- [阅读简体中文](./translations/zh-cn/README.md) (Translator - **Microsoft MVP** [@shijiong](https://github.com/shijiong), **Microsoft Student Ambassador** [@JamboChen](https://github.com/JamboChen)）
+| Language            | Code | Translation                                   | Translators                                                        |
+|---------------------|------|-----------------------------------------------|--------------------------------------------------------------------|
+| Traditional Chinese | zh-tw| [閱讀正體中文](./translations/zh-tw/README.md) | [@doggy8088](https://github.com/doggy8088) (MVP & RD)              |
+| Simplified Chinese  | zh-cn| [阅读简体中文](./translations/zh-cn/README.md) | [@shijiong](https://github.com/shijiong) (MVP), [@JamboChen](https://github.com/JamboChen) (Student Ambassador) |
 
 ## Trademarks
 
 This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow [Microsoft's Trademark & Brand Guidelines](https://www.microsoft.com/legal/intellectualproperty/trademarks/usage/general).
-Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship.
-Any use of third-party trademarks or logos are subject to those third-party's policies.
+Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.
diff --git a/imgs/00/phi3mediumbenchmark.png → imgs/01/phi3mediumbenchmark.png b/imgs/00/phi3mediumbenchmark.png → imgs/01/phi3mediumbenchmark.png
diff --git a/imgs/00/phi3mediumbenchmark128.png → imgs/01/phi3mediumbenchmark128.png b/imgs/00/phi3mediumbenchmark128.png → imgs/01/phi3mediumbenchmark128.png
diff --git a/imgs/00/phi3minibenchmark.png → imgs/01/phi3minibenchmark.png b/imgs/00/phi3minibenchmark.png → imgs/01/phi3minibenchmark.png
diff --git a/imgs/00/phi3minibenchmark128.png → imgs/01/phi3minibenchmark128.png b/imgs/00/phi3minibenchmark128.png → imgs/01/phi3minibenchmark128.png
diff --git a/imgs/00/phi3smallbenchmark.png → imgs/01/phi3smallbenchmark.png b/imgs/00/phi3smallbenchmark.png → imgs/01/phi3smallbenchmark.png
diff --git a/imgs/00/phi3smallbenchmark128.png → imgs/01/phi3smallbenchmark128.png b/imgs/00/phi3smallbenchmark128.png → imgs/01/phi3smallbenchmark128.png
diff --git a/imgs/00/phi3visionbenchmark.png → imgs/01/phi3visionbenchmark.png b/imgs/00/phi3visionbenchmark.png → imgs/01/phi3visionbenchmark.png
diff --git a/md/01.Introduce/Phi3Family.md b/md/01.Introduce/Phi3Family.md
@@ -1,8 +1,8 @@
-# **Phi-3 Family**
+# Microsoft's Phi-3 family
 
-The Phi-3 models are the most capable and cost-effective small language models (SLMs) available, outperforming models of the same size and the next size up across a variety of language, reasoning, coding, and math benchmarks. This release expands the selection of high-quality models for customers, offering more practical choices for composing and building generative AI applications.
+The Phi-3 models are the most capable and cost-effective Small Language Models(SLMs) available, outperforming models of the same size and the next size up across a variety of language, reasoning, coding, and math benchmarks. This release expands the selection of high-quality models for customers, offering more practical choices for composing and building generative AI applications.
 
-The Phi-3 Family includes mini, small, medium and vision versions, trained based on different parameter amounts to serve various application scenarios each model is instruction-tuned and developed in accordance with Microsoft's Responsible AI, safety and security standards to ensure it's ready to use off-the-shelf.
+The Phi-3 Family includes mini, small, medium and vision versions, trained based on different parameter amounts to serve various application scenarios. Each model is instruction-tuned and developed in accordance with Microsoft's Responsible AI, safety and security standards to ensure it's ready to use off-the-shelf. Phi-3-mini outperforms models twice its size, and Phi-3-small and Phi-3-medium outperform much larger models, including GPT-3.5T.
 
 ## Example of Phi-3 Tasks
 | | |
@@ -15,38 +15,56 @@ The Phi-3 Family includes mini, small, medium and vision versions, trained based
 |Self Orchestration (Assistant)|No|
 |Dedicated Embedding Models|No|
 
-## **Phi-3-Mini**
+## Phi-3-mini
 
-Phi-3-mini is a 3.8B parameter language model, available in two context lengths [128K](https://aka.ms/phi3-mini-128k-azure-ai) and [4K.](https://aka.ms/phi3-mini-4k-azure-ai)  
+Phi-3-mini, a 3.8B parameter language model, is available on [Microsoft Azure AI Studio](https://ai.azure.com/explore/models?selectedCollection=phi), [Hugging Face](https://huggingface.co/collections/microsoft/phi-3-6626e15e9585a200d2d761e3), and [Ollama](https://ollama.com/library/phi3). It offers two context lengths: [128K](https://ai.azure.com/explore/models/Phi-3-mini-128k-instruct/version/9/registry/azureml) and [4K](https://ai.azure.com/explore/models/Phi-3-mini-4k-instruct/version/9/registry/azureml).
 
-Phi-3-Mini is a Transformer-based language model with 3.8 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources consisting of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. Additionally, Phi-3-Mini has been chat fine-tuned after pre-training through supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-Mini has demonstrated significant improvements in several capabilities, particularly in alignment, robustness, and safety. The model is part of the Phi-3 family and comes in the Mini version with two variants, 4K and 128K, which represent the context length (in tokens) that it can support. 
+Phi-3-mini is a Transformer-based language model with 3.8 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources consisting of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. Additionally, Phi-3-mini has been chat fine-tuned after pre-training through supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-mini has demonstrated significant improvements in several capabilities, particularly in alignment, robustness, and safety. The model is part of the Phi-3 family and comes in the mini version with two variants, 4K and 128K, which represent the context length (in tokens) that it can support. 
 
-## **Phi-3-Small**
+![phi3modelminibenchmark](../../imgs/01/phi3minibenchmark.png)
 
-Phi-3-small is a 7B parameter language model, available in two context lengths [128K](https://aka.ms/phi3-small-128k-azure-ai) and [8K.](https://aka.ms/phi3-small-8k-azure-ai)
+![phi3modelminibenchmark128k](../../imgs/01/phi3minibenchmark128.png)
 
-Phi-3-Small is a Transformer-based language model with 7 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources that consist of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. In addition, Phi-3-Small has been chat fine-tuned after pre-training via supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-Small has shown significant improvements in several capabilities, particularly in alignment, robustness, and safety. Phi-3-Small is also more intensively trained on multilingual datasets compared to Phi-3-Mini. The model family offers two variants, 8K and 128K, which represent the context length (in tokens) that it can support.
+## Phi-3-small
 
-## **Phi-3-Medium**
+Phi-3-small, a 7B parameter language model, available in two context lengths [128K](https://ai.azure.com/explore/models/Phi-3-small-128k-instruct/version/2/registry/azureml) and [8K.](https://ai.azure.com/explore/models/Phi-3-small-8k-instruct/version/2/registry/azureml) outperforms GPT-3.5T across a variety of language, reasoning, coding, and math benchmarks.
 
-Phi-3-medium is a 14B parameter language model, available in two context lengths [128K](https://aka.ms/phi3-medium-128k-azure-ai) and [4K.](https://aka.ms/phi3-medium-4k-azure-ai)
+Phi-3-small is a Transformer-based language model with 7 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources that consist of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. In addition, Phi-3-small has been chat fine-tuned after pre-training via supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-small has shown significant improvements in several capabilities, particularly in alignment, robustness, and safety. Phi-3-small is also more intensively trained on multilingual datasets compared to Phi-3-Mini. The model family offers two variants, 8K and 128K, which represent the context length (in tokens) that it can support.
 
-Phi-3-Medium is a Transformer-based language model with 14 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources that consist of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. Additionally, Phi-3-Medium has been chat fine-tuned after pre-training through supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-Medium has exhibited significant improvements in several capabilities, particularly in alignment, robustness, and safety. The model family offers two variants, 4K and 128K, which represent the context length (in tokens) that it can support.
+![phi3modelsmall](../../imgs/01/phi3smallbenchmark.png)
 
-## **Phi-3-vision**
+![phi3modelsmall128k](../../imgs/01/phi3smallbenchmark128.png)
 
-The [Phi-3-vision](https://aka.ms/phi3-vision-128k-azure-ai) is a 4.2B parameter multimodal model with language and vision capabilities.  
+## Phi-3-medium
 
-Phi-3-vision is the first multimodal model in the Phi-3 family, bringing together text and images. Phi-3-vision can be used to reason over real-world images and extract and reason over text from images. It has also been optimized for chart and diagram understanding and can be used to generate insights and answer questions. Phi-3-vision builds on the language capabilities of the Phi-3-mini, continuing to pack strong language and image reasoning quality in a small size.  
+Phi-3-medium, a 14B parameter language model, available in two context lengths [128K](https://ai.azure.com/explore/models/Phi-3-medium-128k-instruct/version/2/registry/azureml) and [4K.](https://ai.azure.com/explore/models/Phi-3-medium-4k-instruct/version/2/registry/azureml), continues the trend by outperforming Gemini 1.0 Pro.
 
-## **Phi Silica**
-We are introducing Phi Silica which is built from the Phi series of models and is designed specifically for the NPUs in Copilot+ PCs. Windows is the first platform to have a state-of-the-art small language model (SLM) custom built for the NPU and shipping inbox.   
-Phi Silica API along with OCR, Studio Effects, Live Captions, Recall User Activity APIs will be available in Windows Copilot Library in June. More APIs like Vector Embedding, RAG API, Text Summarization will be coming later. 
+Phi-3-medium is a Transformer-based language model with 14 billion parameters. It was trained using high-quality data containing educationally useful information, augmented with new data sources that consist of various NLP synthetic texts, and both internal and external chat datasets, which significantly improve chat capabilities. Additionally, Phi-3-medium has been chat fine-tuned after pre-training through supervised fine-tuning (SFT) and Direct Preference Optimization (DPO). Following this post-training, Phi-3-medium has exhibited significant improvements in several capabilities, particularly in alignment, robustness, and safety. The model family offers two variants, 4K and 128K, which represent the context length (in tokens) that it can support.
+
+![phi3modelmedium](../../imgs/01/phi3mediumbenchmark.png)
+
+![phi3modelmedium128k](../../imgs/01/phi3mediumbenchmark128.png)
+
+## Phi-3-vision
+
+The [Phi-3-vision](https://ai.azure.com/explore/models/Phi-3-vision-128k-instruct/version/2/registry/azureml), a 4.2B parameter multimodal model with language and vision capabilities, outperforms larger models like Claude-3 Haiku and Gemini 1.0 Pro V in general visual reasoning, OCR, and table and chart understanding tasks.
+
+Phi-3-vision is the first multimodal model in the Phi-3 family, bringing together text and images. Phi-3-vision can be used to reason over real-world images and extract and reason over text from images. It has also been optimized for chart and diagram understanding and can be used to generate insights and answer questions. Phi-3-vision builds on the language capabilities of the Phi-3-mini, continuing to pack strong language and image reasoning quality in a small size.
+
+![phi3modelvision](../../imgs/01/phi3visionbenchmark.png)
+
+> [!NOTE]
+>
+> Phi-3 models do not perform as well on factual knowledge benchmarks (such as TriviaQA) as the smaller model size results in less capacity to retain facts.
+
+## Phi silica
+
+We are introducing Phi Silica which is built from the Phi series of models and is designed specifically for the NPUs in Copilot+ PCs. Windows is the first platform to have a state-of-the-art small language model (SLM) custom built for the NPU and shipping inbox. Phi Silica API along with OCR, Studio Effects, Live Captions, and Recall User Activity APIs will be available in Windows Copilot Library in June. More APIs like Vector Embedding, RAG API, and Text Summarization will be coming later.
 
 ## **Find all Phi-3 models** 
 
-- [Azure AI](https://aka.ms/phi3-azure-ai) 
-- [Hugging Face.](https://aka.ms/phi3-hf) 
+- [Azure AI](https://ai.azure.com/explore/models?selectedCollection=phi)
+- [Hugging Face](https://huggingface.co/collections/microsoft/phi-3-6626e15e9585a200d2d761e3) 
 
 ## ONNX Models