piperguy / llama-2-7b-vllm

Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 7B fine-tuned model, optimized for dialogue use cases and converted for the Hugging Face Transformers format. Links to other models can be found in the index at the bottom.

Llama 2 - 7b with vLLM

vLLM is a fast and easy-to-use library for LLM inference and serving. You can use this template to use vLLM in Inferless.

---

Prerequisites

• Git. You would need git installed on your system if you wish to customize the repo after forking.
• Python>=3.8. You would need Python to customize the code in the app.py according to your needs.
• Curl. You would need Curl if you want to make API calls from the terminal itself.

---

Quick Start

Here is a quick start to help you get up and running with this template on Inferless.

Download the config-vllm and Create a runtime

Get started by downloading the config-vllm.yaml file and go to Inferless dashboard and create a custom runtime

Quickly add this as a Custom runtime

Fork the Repository

Get started by forking the repository. You can do this by clicking on the fork button in the top right corner of the repository page.

This will create a copy of the repository in your own GitHub account, allowing you to make changes and customize it according to your needs.

Add Your Hugging Face Auth Token

Go into the app.py and replace <your_token> with your hugging face api key. This token will be used to download weights from hugging face if you're using Llama-2. Make sure to check the repo is private to protect your hugging face key.

Import the Model in Inferless

Log in to your inferless account, select the workspace you want the model to be imported into and click the Add Model button.

Select the PyTorch as framework and choose Repo(custom code) as your model source and use the forked repo URL as the Model URL.

After the create model step, while setting the configuration for the model make sure to select the appropriate runtime.

Enter all the required details to Import your model. Refer this link for more information on model import.

The following is a sample Input and Output JSON for this model which you can use while importing this model on Inferless.

Input

{
  "inputs": [
    {
      "data": [
        "What is Quantum Computing?"
      ],
      "name": "questions",
      "shape": [
        1
      ],
      "datatype": "BYTES"
    }
  ]
}

Output

{
  "outputs": [
    {
      "data": [
        "data"
      ],
      "name": "result",
      "shape": [
        1
      ],
      "datatype": "BYTES"
    }
  ]
}

---

Curl Command

Following is an example of the curl command you can use to make inference. You can find the exact curl command in the Model's API page in Inferless.

curl --location '<your_inference_url>' \
          --header 'Content-Type: application/json' \
          --header 'Authorization: Bearer <your_api_key>' \
          --data '{
                "inputs": [
                    {
                    "data": [
                        "What is Quantum Computing?"
                    ],
                    "name": "questions",
                    "shape": [
                        1
                    ],
                    "datatype": "BYTES"
                    }
                ]
                }
            '

---

Customizing the Code

Open the app.py file. This contains the main code for inference. It has three main functions, initialize, infer and finalize.

Initialize - This function is executed during the cold start and is used to initialize the model. If you have any custom configurations or settings that need to be applied during the initialization, make sure to add them in this function.

Infer - This function is where the inference happens. The argument to this function inputs, is a dictionary containing all the input parameters. The keys are the same as the name given in inputs. Refer to input for more.

def infer(self, inputs):
    prompt = inputs["prompt"]

Finalize - This function is used to perform any cleanup activity for example you can unload the model from the gpu by setting self.pipe = None.

For more information refer to the Inferless docs.