bert.json

{
    "code": "import matplotlib.pyplot as plt\r\nimport numpy as np\r\nfrom sagemaker import image_uris, model_uris, script_uris, instance_types\r\nfrom sagemaker.predictor import Predictor\r\nfrom sagemaker import get_execution_role\r\nimport json\r\n\r\n\r\nmodel_id, model_version = \"model-txt2img-stabilityai-stable-diffusion-v2-fp16\", \"*\"\r\n\r\ninference_instance_type = instance_types.retrieve_default(\r\n    model_id=model_id, model_version=model_version, scope=\"inference\"\r\n)\r\n\r\n# Retrieve the inference docker container uri. This is the base HuggingFace container image for the default model above.\r\ndeploy_image_uri = image_uris.retrieve(\r\n    region=None,\r\n    framework=None,  # automatically inferred from model_id\r\n    image_scope=\"inference\",\r\n    model_id=model_id,\r\n    model_version=model_version,\r\n    instance_type=inference_instance_type,\r\n)\r\n\r\n# Retrieve the inference script uri. This includes all dependencies and scripts for model loading, inference handling etc.\r\ndeploy_source_uri = script_uris.retrieve(\r\n    model_id=model_id, model_version=model_version, script_scope=\"inference\"\r\n)\r\n\r\n\r\n# Retrieve the model uri. This includes the pre-trained nvidia-ssd model and parameters.\r\nmodel_uri = model_uris.retrieve(\r\n    model_id=model_id, model_version=model_version, model_scope=\"inference\"\r\n)\r\n\r\n# To increase the maximum response size from the endpoint.\r\nenv = {\r\n    \"MMS_MAX_RESPONSE_SIZE\": \"20000000\",\r\n}\r\n\r\n# Create the SageMaker model instance\r\nmodel = Model(\r\n    image_uri=deploy_image_uri,\r\n    source_dir=deploy_source_uri,\r\n    model_data=model_uri,\r\n    entry_point=\"inference.py\",  # entry point file in source_dir and present in deploy_source_uri\r\n    role=get_execution_role(),\r\n    predictor_cls=Predictor,\r\n    name=endpoint_name,\r\n    env=env,\r\n)\r\n\r\n# deploy the Model. Note that we need to pass Predictor class when we deploy model through Model class,\r\n# for being able to run inference through the sagemaker API.\r\nmodel_predictor = model.deploy(\r\n    initial_instance_count=1,\r\n    instance_type=inference_instance_type,\r\n    predictor_cls=Predictor,\r\n    endpoint_name=endpoint_name,\r\n)\r\n\r\n\r\ndef query(model_predictor, text):\r\n    \"\"\"Query the model predictor.\"\"\"\r\n\r\n    encoded_text = json.dumps(text).encode(\"utf-8\")\r\n\r\n    query_response = model_predictor.predict(\r\n        encoded_text,\r\n        {\r\n            \"ContentType\": \"application\/x-text\",\r\n            \"Accept\": \"application\/json\",\r\n        },\r\n    )\r\n    return query_response\r\n\r\n\r\ndef parse_response(query_response):\r\n    \"\"\"Parse response and return generated image and the prompt\"\"\"\r\n\r\n    response_dict = json.loads(query_response)\r\n    return response_dict[\"generated_image\"], response_dict[\"prompt\"]\r\n\r\n\r\ndef display_img_and_prompt(img, prmpt):\r\n    \"\"\"Display hallucinated image.\"\"\"\r\n    plt.figure(figsize=(12, 12))\r\n    plt.imshow(np.array(img))\r\n    plt.axis(\"off\")\r\n    plt.title(prmpt)\r\n    plt.show()\r\n",
    "links": [{
        "label": "SageMaker JumpStart Examples",
        "href": "https://github.com/aws/amazon-sagemaker-examples/tree/main/sagemaker-jumpstart"
    }]
}