Merge pull request #43 from jina-ai/docs-reader-lm-v2

docs: reader lm v2
jina-ai · Jan 15, 2025 · 42cc069 · 42cc069
2 parents 4ad0f90 + aa9de94
commit 42cc069
Show file tree

Hide file tree

Showing 3 changed files with 25 additions and 26 deletions.
diff --git a/examples/sample-reader-lm-v2-inference-input.json b/examples/sample-reader-lm-v2-inference-input.json
@@ -0,0 +1,5 @@
+{
+    "model": "ReaderLM-v2",
+    "prompt": "<html><head><title>Minimal Bullet Points</title></head><body><ul><li>hello</li><li>jina.ai</li></ul></body></html>",
+    "stream": false
+}
diff --git a/examples/sample-reader-lm-v2-inference-output.json b/examples/sample-reader-lm-v2-inference-output.json
@@ -0,0 +1 @@
+{"id":"cmpl-45e7ff341f3d4084a33c0fe3ece009c6","object":"text_completion","created":1736918982,"model":"ReaderLM-v2","choices":[{"index":0,"text":"The provided HTML content is a simple list with two bullet points. Here's the minimal representation in Markdown:\n\n```markdown\n- hello\n- jina.ai\n```\n\nThis version only includes the essential elements of the original HTML, excluding any decorative or supplementary information. If you need to include additional text or formatting, please provide that as well.","logprobs":null,"finish_reason":"stop","stop_reason":null,"prompt_logprobs":null}],"usage":{"prompt_tokens":61,"total_tokens":131,"completion_tokens":70}}
diff --git a/notebooks/Reader-LM.ipynb b/notebooks/Reader-LM.ipynb
@@ -9,15 +9,18 @@
     "This notebook shows you how to deploy the [Jina Reader-LM](https://jina.ai/news/reader-lm-small-language-models-for-cleaning-and-converting-html-to-markdown) models using Amazon SageMaker and perform inference with it:\n",
     "  - [reader-lm-500m](https://aws.amazon.com/marketplace/pp/prodview-nli7b6dueo424?sr=0-1&ref_=beagle&applicationId=AWSMPContessa)\n",
     "  - [reader-lm-1500m](https://aws.amazon.com/marketplace/pp/prodview-ms27ixcwq3wjk?sr=0-2&ref_=beagle&applicationId=AWSMPContessa)\n",
+    "  - [reader-lm-v2](TBD)\n",
     "\n",
-    "## Pre-requisites:\n",
-    "1. Ensure that IAM role used has **AmazonSageMakerFullAccess**\n",
-    "1. To deploy this ML model successfully, ensure that:\n",
-    "    1. Either your IAM role has these three permissions and you have authority to make AWS Marketplace subscriptions in the AWS account used: \n",
-    "        1. **aws-marketplace:ViewSubscriptions**\n",
-    "        1. **aws-marketplace:Unsubscribe**\n",
-    "        1. **aws-marketplace:Subscribe**  \n",
-    "    2. or your AWS account has a subscription to this model."
+    "## Prerequisites:\n",
+    "\n",
+    "1. This notebook should be rendered correctly in the Jupyter interface and can be executed either within an Amazon SageMaker Notebook Instance or Amazon SageMaker Studio.\n",
+    "2. Ensure that the IAM role being used has **AmazonSageMakerFullAccess**.\n",
+    "3. To successfully deploy this ML model, ensure that:\n",
+    "    1. Either your IAM role has the following three permissions and you have authority to make AWS Marketplace subscriptions in the AWS account:\n",
+    "        - **aws-marketplace:ViewSubscriptions**\n",
+    "        - **aws-marketplace:Unsubscribe**\n",
+    "        - **aws-marketplace:Subscribe**\n",
+    "    2. Or, your AWS account already has a subscription to this model."
    ]
   },
   {
@@ -50,19 +53,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Install `jina-sagemaker` package \n",
-    "\n",
-    "\n",
-    "```bash\n",
-    "pip install --upgrade jina-sagemaker\n",
-    "```"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "And then get the model package ARN(s)."
+    "And then let's install `jina-sagemaker` package and get the model package ARN(s) using code below."
    ]
   },
   {
@@ -71,6 +62,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "!pip install --upgrade jina-sagemaker\n",
+    "\n",
     "import boto3\n",
     "\n",
     "region = boto3.Session().region_name\n",
@@ -83,10 +76,11 @@
     "model_name_map = {\n",
     "    \"reader-lm-500m\": \"reader-lm-500m-186f9d30c561356c92a721dbf9540212\",\n",
     "    \"reader-lm-1500m\": \"reader-lm-1500m-cc71e40a204537f2a2fdbbd0a03d88e8\",\n",
+    "    \"reader-lm-v2\": \"reader-lm-v2-tbd\",\n",
     "}\n",
     "\n",
-    "# Specify the model name, reader-lm-500m is picked here for example\n",
-    "model_package_name = model_name_map[\"reader-lm-500m\"]\n",
+    "# Specify the model name, `reader-lm-v2` picked here for example\n",
+    "model_package_name = model_name_map[\"reader-lm-v2\"]\n",
     "\n",
     "# Mapping for Model Packages\n",
     "def get_arn_for_model(region_name, model_name):\n",
@@ -239,8 +233,7 @@
     "\n",
     "Create a input file `input.json` with the following content.\n",
     "\n",
-    "Note that if `reader-lm-500m` is the used, then `\"reader-lm-0.5b\"` should be used for `model`, \n",
-    "if `reader-lm-1500m` is the used, then `\"reader-lm-1.5b\"` should be used for `model`.  "
+    "Note that If using `reader-lm-v2`, set `model` to `ReaderLM-v2`; for `reader-lm-500m`, set `model` to `reader-lm-0.5b`; and for `reader-lm-1500m`, set `model` to `reader-lm-1.5b`."
    ]
   },
   {
@@ -334,7 +327,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.14"
+   "version": "3.10.16"
   }
  },
  "nbformat": 4,
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		{"id":"cmpl-45e7ff341f3d4084a33c0fe3ece009c6","object":"text_completion","created":1736918982,"model":"ReaderLM-v2","choices":[{"index":0,"text":"The provided HTML content is a simple list with two bullet points. Here's the minimal representation in Markdown:\n\n```markdown\n- hello\n- jina.ai\n```\n\nThis version only includes the essential elements of the original HTML, excluding any decorative or supplementary information. If you need to include additional text or formatting, please provide that as well.","logprobs":null,"finish_reason":"stop","stop_reason":null,"prompt_logprobs":null}],"usage":{"prompt_tokens":61,"total_tokens":131,"completion_tokens":70}}