Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the 7B Python specialist version. This model is designed for general code synthesis and understanding.
- Deployment of Codellama - 7B model using vLLM.
- By using the vLLM, you can expect an average latency of 2.09 sec.
- 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.
Here is a quick start to help you get up and running with this template on Inferless.
Get started by downloading the config.yaml file and go to Inferless dashboard and create a custom runtime
Quickly add this as a Custom runtime
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.
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.
{
"inputs": [
{
"data": [
"def factorial(int n):"
],
"name": "prompt",
"shape": [
1
],
"datatype": "BYTES"
}
]
}
{
"outputs": [
{
"data": [
"data"
],
"name": "result",
"shape": [
1
],
"datatype": "BYTES"
}
]
}
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": [
"def factorial(int n):"
],
"name": "prompt",
"shape": [
1
],
"datatype": "BYTES"
}
]
}
'
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.