Llama 4 household of fashions from Meta at the moment are out there in SageMaker JumpStart

from sagemaker.jumpstart.mannequin import JumpStartModel

mannequin = JumpStartModel(model_id = "meta-vlm-llama-4-scout-17b-16e-instruct")

predictor = mannequin.deploy(accept_eula=False)

payload = {
"messages": [
 {"role": "system", "content": "You are a helpful assistant"},
 {"role": "user", "content": "How are you doing today"},
 {"role": "assistant", "content": "Good, what can i help you with today?"},
 {"role": "user", "content": "Give me 5 steps to become better at tennis?"}
],
 "temperature": 0.6,
 "top_p": 0.9,
 "max_tokens": 512,
 "logprobs": False
} 
response = predictor.predict(payload) 
response_message = response['choices'][0]['message']['content']

payload = {
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful AI assistant."
        },
        {
            "role": "user",
            "content": "What are three key benefits of large language models for businesses?"
        }
    ],
    "max_tokens": 2048,
    "temperature": 0.7,
    "top_p": 0.9,
    "stream": False
}

Giant language fashions (LLMs) supply quite a few advantages to companies, remodeling how they function, 
work together with clients, and generate insights. Listed below are three key advantages:

1. **Enhanced Buyer Expertise by way of Automation and Personalization**:
   - **Automated Buyer Assist**: LLMs can energy chatbots and digital assistants 
	that present 24/7 buyer assist. These fashions can perceive and reply to a variety of buyer queries, assist with troubleshooting, and even course of transactions, considerably decreasing the workload on human buyer assist brokers.
   - **Personalization**: By analyzing buyer knowledge, LLMs can assist companies personalize their 
	communications and choices. As an example, they'll generate customized emails, product suggestions, and content material, enhancing buyer engagement and satisfaction.

2. **Effectivity and Productiveness Good points**:
   - **Content material Era**: LLMs can automate the creation of varied varieties of content material, equivalent to 
   weblog posts, studies, product descriptions, and social media updates. This not solely accelerates content material manufacturing but additionally permits human writers to concentrate on extra inventive and strategic duties.
   - **Information Evaluation and Summarization**: These fashions can shortly analyze giant volumes of information, extract related info, and summarize findings in a readable format. This functionality can considerably scale back the effort and time required for market analysis, aggressive evaluation, and inner reporting.

3. **Improved Determination Making with Information-Pushed Insights**:
   - **Market and Pattern Evaluation**: LLMs can course of and analyze huge quantities of information from numerous sources, together with information articles, social media, and market studies. This helps companies keep knowledgeable about market traits, client sentiment, and competitor exercise, enabling extra knowledgeable strategic choices.
   - **Threat Administration and Compliance**: By analyzing regulatory paperwork and monitoring communications, LLMs can assist companies determine and mitigate compliance dangers. They will additionally help within the creation of compliance studies and documentation, decreasing the danger of human error.

In abstract, giant language fashions supply companies the potential to automate and improve buyer interactions, 
enhance operational effectivity, and achieve deeper insights from their knowledge. These capabilities can result in value financial savings, 
elevated income, and a aggressive edge out there.

import requests
import base64
import boto3
import json

s3 = boto3.shopper("s3")

s3_bucket = f"jumpstart-cache-prod-{area}"
key_prefix = "inference-notebook-assets"

def download_from_s3(key_filenames):
    for key_filename in key_filenames:
        s3.download_file(s3_bucket, f"{key_prefix}/{key_filename}", key_filename)

# Outline picture names
heat_map = "heatmap_semantic_similarity_search.png"

# Obtain and show the heatmap picture
download_from_s3(key_filenames=[heat_map])

def img_to_base64(image_path):
    with open(image_path, "rb") as f:
        img = f.learn()
    enc_img = base64.b64encode(img).decode('utf-8')
    return enc_img

b64_img = img_to_base64(heat_map)

knowledge = {
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant",
        },
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": "What is in this image?",
                },
                {
                    "type": "image_url",
                    "image_url": {
                        "url": b64_img
                    }
                },
            ],
        },
    ],
    "temperature": 0.6,
    "top_p": 0.9,
    "max_tokens": 512
}

payload = json.dumps(knowledge)

The picture presents a heatmap illustrating the semantic textual similarity between 
sentences. The heatmap is split into two axes, with the x-axis itemizing six sentences 
and the y-axis itemizing the identical six sentences in a distinct order.nn**Sentences:**nn* 
Your canine is so cute.n* How cute your canine is!n* You have got such a cute canine!n* New York Metropolis 
is the place the place I work.n* I work in New York Metropolis.n* What colour do you want probably the most?n* 
What's your favorite colour?nn**Heatmap:**nnThe heatmap shows the semantic textual 
similarity between every pair of sentences, with darker shades of purple indicating greater similarity 
and lighter shades of orange indicating decrease similarity. The colour bar on the suitable facet of the 
heatmap offers a scale for decoding the similarity values, starting from -0.4 to 1.0.nn
**Observations:**nn* The three sentences associated to canines ("Your canine is so cute.", "How cute your canine is!", 
and "You have got such a cute canine!") present excessive similarity with one another, as indicated by the darkish purple squares 
within the corresponding part of the heatmap.n* The 2 sentences associated to New York Metropolis ("New York Metropolis is 
the place the place I work." and "I work in New York Metropolis.") additionally present excessive similarity with one another.n* The 2 
sentences associated to favourite colours ("What colour do you want probably the most?" and "What's your favourite colour?") 
present excessive similarity with one another.n* The sentences from completely different classes (canines, New York Metropolis, and 
favourite colours) present decrease similarity with one another, as indicated by the lighter shades of orange within the heatmap.nn
General, the heatmap successfully visualizes the semantic textual similarity between the given sentences, permitting 
for straightforward identification of comparable and dissimilar sentence pairs.

def url_to_base64(image_url):
    # Obtain the picture
    response = requests.get(image_url)
    if response.status_code != 200:
        return None

    # Encode the picture content material to base64
    image_base64 = base64.b64encode(response.content material).decode('utf-8')
    return image_base64

image_url_2 = "https://uncooked.githubusercontent.com/mathvision-cuhk/MATH-V/refs/heads/most important/pictures/13.jpg"
b64_img_url_2 = url_to_base64(image_url_2)

knowledge = {
    "messages": [
        {
            "role": "system",
            "content": "You are a helpful assistant",
        },
        {
            "role": "user",
            "content": [
                {
                    "type": "text",
                    "text": "There are two images. Describe each one and why they are related",
                },
                {
                    "type": "image_url",
                    "image_url": {
                        "url": b64_img,
                    }
                },
                {
                    "type": "image_url",
                    "image_url": {
                        "url": b64_img_url_2,
                    }
                },
            ],
        },
    ],
    "temperature": 0.6,
    "top_p": 0.9,
    "max_tokens": 512
}

The primary picture is a heatmap illustrating the semantic textual similarity between 
sentences. The heatmap is titled "Semantic Textual Similarity Between Sentences" 
and incorporates a colour scale starting from -0.4 to 1.0, with darker shades indicating 
greater similarity. The sentences being in contrast are:nn* Your canine is so cute.n* 
How cute your canine is!n* You have got such a cute canine!n* New York Metropolis is the place 
the place I work.n* I work in New York Metropolis.n* What colour do you want probably the most?n* 
What's your favorite colour?nnThe heatmap reveals that the sentences about canines 
have excessive similarity scores, with values above 0.9, indicating that they convey 
comparable meanings. Equally, the sentences about work areas and favourite colours 
additionally present excessive similarity scores. Nevertheless, there's a low similarity rating between 
the sentences about canines and people about work areas or favourite colours.nn
The second picture presents 5 stick figures labeled A, B, C, D, and E, every representing
an individual. The figures are drawn utilizing easy strains and geometric shapes, equivalent to circles
for heads, squares for our bodies, and triangles for skirts. The principle distinction between the
figures lies of their arm and leg positions:nn* Determine A has its arms at its sides and 
straight legs.n* Determine B has its arms at its sides and straight legs, however with a barely 
completely different posture.n* Determine C has its arms at its sides and straight legs, much like Determine A.n*
Determine D has one arm raised and straight legs.n* Determine E has two arms raised and bent legs.nn
General, the 2 pictures are distinct when it comes to their content material and goal. The heatmap offers a 
visible illustration of semantic similarities between sentences, whereas the stick figures seem to 
be easy illustrations of individuals with various poses.

payload = {
    "messages": [
        {
            "role": "system",
            "content": """You are a helpful AI assistant that summarizes codebases to me to help me understand
            how to analyze code by synthesizing through the entire codebase before responding.
            
            Be thorough in your search as the file may be nested within a markdown code block or within a directory listing.
            """
        },
        {
            "role": "user",
            "content": f"Can you explain to me the buildspec-1-10-2.yml file and how it relates to the rest of the huggingface directory? Here is the AWS Deep Learning Container Repository for reference:nn{text}"
        }
    ],
    "max_tokens": 2048,
    "temperature": 0.7,
    "top_p": 0.9,
    "stream": False
}

## Overview of the buildspec-1-10-2.yml File

The offered info seems to be part of a big, merged illustration of 
the Hugging Face repository, generated by Repomix. The `buildspec-1-10-2.yml` file 
is probably going a construct specification file utilized in a Steady Integration/Steady 
Deployment (CI/CD) pipeline.

## Goal and Format

The file comprises a packed illustration of all the repository's contents, designed
 to be simply consumable by AI programs for evaluation, code evaluation, or different automated processes. 
The content material is organized into sections:

1.  **Abstract**: Offers an summary of the file's goal, format, and utilization pointers.
2.  **Repository Data**: Not proven within the offered excerpt, however seemingly comprises metadata concerning the repository.
3.  **Listing Construction**: Lists the listing construction of the repository, together with file paths.
4.  **File Entries**: Every file entry consists of a header with the file path and the total contents 
of the file in a code block.

## Relationship to the Hugging Face Listing

The `buildspec-1-10-2.yml` file is a part of the Hugging Face repository, which seems to be a 
giant challenge with a number of subdirectories and information. The listing construction part lists numerous 
subdirectories, together with:

*   `.github/`: Comprises GitHub-related configuration information, equivalent to subject templates, workflows, 
and CODEOWNERS.
*   `.release_templates/`: Contains launch picture templates for numerous frameworks like MXNet, PyTorch, 
and TensorFlow.
*   `autogluon/`: A subdirectory with construct artifacts, inference configurations, and different associated information.

## buildspec-1-10-2.yml File Contents

Sadly, the offered excerpt doesn't comprise the precise contents of the `buildspec-1-10-2.yml` file. 
Nevertheless, primarily based on the file title and its presence in a CI/CD context, it's seemingly that this file specifies 
the construct course of, dependencies, and atmosphere configurations for the Hugging Face challenge.

## Utilization Tips

When working with this file, remember:

*   Deal with this file as read-only and make modifications to the unique repository information as a substitute.
*   Use the file path to tell apart between completely different information within the repository.
*   Deal with this file with the identical stage of safety as the unique repository, as it could 
comprise delicate info.

## Further Notes

*   Some information is perhaps excluded primarily based on `.gitignore` guidelines and Repomix's configuration.
*   Binary information will not be included on this packed illustration.
*   Information are sorted by Git change depend, with information having extra modifications showing on the backside.

## Instance Use Case

To investigate the `buildspec-1-10-2.yml` file, you'll:

1.  Find the file within the repository listing construction.
2.  Evaluate the file contents to grasp the construct course of, dependencies, and atmosphere configurations.
3.  Use this info to optimize the construct course of, troubleshoot points, or make modifications to the repository.

Remember that the precise contents of the `buildspec-1-10-2.yml` file will not be offered, so this 
is a normal rationalization primarily based on the file title and context. 

To get the precise content material you'll search for 

## File: buildspec-1-10-2.yml 

 block and skim its content material 

Let me know if I can help with the rest.

payload = {
    "messages": [
        {
            "role": "system",
            "content": "You are a financial analyst. Extract key financial data from this 10-K report."
        },
        {
            "role": "user",
            "content": f"""Extract the following information from this Amazon 10-K report:
            1. What year is this report for?
            2. Total revenue for the year
            3. Net income
            4. Key business segments and their performance
            5. Important trends mentioned
            
            Here's the 10-K text:
            {pdf_text[:25000]}
            
            Format your response as a JSON object with these fields: 12 months, income, net_income, segments, traits.
            """
        }
    ],
    "max_tokens": 2048,
    "temperature": 0.7,
    "top_p": 0.9,
    "stream": False
}

To investigate Amazon's development from 2017 to 2024, we'll look at the income and internet revenue 
knowledge out there from the 10-Okay studies.

### Income Progress

| Yr | Income (in billion USD) |
| --- | --- |
| 2014 | 89 |
| 2015 | 107 |
| 2016 | 135.99 |
| 2017 | 177.87 |
| 2018 | 232.88 |
| 2019 | 280.52 |
| 2020 | 386.06 |
| 2021 | 478 |
| 2022 | 513 |
| 2023 | 513.98 |
| 2024 | 574.78 (projected) |

From 2017 to 2024, Amazon's income has grown from $177.87 billion to a projected $574.78 billion, 
representing a major enhance of roughly 223% over the 7-year interval.

### Internet Earnings Progress

| Yr | Internet Earnings (in billion USD) |
| --- | --- |
| 2014 | 0.236 |
| 2015 | 0.596 |
| 2016 | 2.37 |
| 2017 | 3.03 |
| 2018 | 10.07 |
| 2019 | 11.59 |
| 2020 | 18.68 |
| 2021 | 33.4 |
| 2022 | 18.7 (decrease on account of financial circumstances) |
| 2023 | 33.36 |
| 2024 | 37.68 (projected) |

Amazon's internet revenue has grown from $3.03 billion in 2017 to a projected $37.68 billion in 2024, 
representing a considerable enhance of roughly 1143% over the 7-year interval.

### Phase and Pattern Evaluation

All through the years, Amazon has maintained its three most important enterprise segments: North America, 
Worldwide, and Amazon Internet Companies (AWS). The corporate has continued to increase into new merchandise, 
companies, applied sciences, and geographic areas, driving development and rising competitors.

The traits talked about throughout the 10-Okay studies spotlight the corporate's:

1. **Seasonality**: Increased gross sales quantity throughout the fourth quarter.
2. **Intense competitors**: Throughout geographies and industries, together with e-commerce, cloud computing, 
and digital content material.
3. **Growth**: Into new merchandise, companies, applied sciences, and geographic areas, which topics 
the corporate to further enterprise, authorized, monetary, and aggressive dangers.
4. **Worldwide operations**: Topic to dangers equivalent to native financial and political circumstances, 
authorities regulation, and restrictions on gross sales or distribution.

### Conclusion

Amazon has demonstrated outstanding development from 2017 to 2024, with income rising by 
roughly 223% and internet revenue rising by roughly 1143%. The corporate's continued 
enlargement into new areas, its sturdy presence in cloud computing by way of AWS, and its means 
to adapt to altering market circumstances have contributed to its success. Nevertheless, the corporate 
additionally faces intense competitors, seasonality, and dangers related to worldwide operations.

--------------------------------------------------
Ask a query concerning the Amazon 10-Okay studies throughout years.

predictor.delete_model()
predictor.delete_endpoint()