Impulse SDK for Python

The Impulse SDK for Python provides a convenient way to interact with the Impulse API for managing large language model (LLM) fine-tuning processes and related resources.

Features

• User management
• API key management
• Dataset operations
• Fine-tuning job management
• Model information retrieval
• Usage tracking

Installation

You can install the Impulse SDK using pip:

pip install impulse-api-sdk-python

Setting up API Key

Generate an API key by visiting the Impulse Dashboard at this settings page.

Setting environment variable

export IMPSDK_API_KEY=xxxxx

Usage – Python Client

User Management

import os
import asyncio
from impulse.api_sdk.sdk import ImpulseSDK


async def main():
    async with ImpulseSDK(os.environ.get("IMPSDK_API_KEY")) as client:
        user = await client.user.get_current_user()
        print(user)


asyncio.run(main())

Datasets

The files API is used for fine-tuning and allows developers to upload data to fine-tune on. It also has several methods to list all files, retrive files, and delete files. Please refer to our fine-tuning docs here.

import asyncio
import os

from impulse.api_sdk.sdk import ImpulseSDK


async def main():
    async with ImpulseSDK(os.environ.get("IMPSDK_API_KEY")) as client:
        files = await client.dataset.list_datasets()
        await client.dataset.upload_dataset("somefile.jsonl")
        await client.dataset.get_dataset("somefile.jsonl")
        await client.dataset.delete("somefile.jsonl")

        print(files)


asyncio.run(main())

Models

This lists all the models that Impulse supports.

import os
import asyncio
from impulse.api_sdk.sdk import ImpulseSDK


async def main():
    async with ImpulseSDK(os.environ.get("IMPSDK_API_KEY")) as client:
        models = await client.model.list_base_models()
        print(models)


asyncio.run(main())

Fine-tunes

The finetune API is used for fine-tuning and allows developers to create finetuning jobs. It also has several methods to list all jobs, retrive statuses and get checkpoints. Please refer to our fine-tuning docs here.

import os
import asyncio
from impulse.api_sdk.sdk import ImpulseSDK
from impulse.api_sdk.models import FineTuningJobCreate, FineTuningJobParameters


async def main():
    async with ImpulseSDK(os.environ.get("IMPSDK_API_KEY")) as client:
        files = await client.dataset.list_datasets()
        job = await client.fine_tuning.create_fine_tuning_job(FineTuningJobCreate(
            base_model_name="llm_llama3_1_8b",
            dataset_name=files[0].name,
            name="test-fine-tuning-job",
            parameters=FineTuningJobParameters(
                batch_size=2,
                shuffle=True,
                num_epochs=1,
                use_lora=True,
                use_qlora=False
            )
        ))

        print(job)

        jobs = await client.fine_tuning.list_fine_tuning_jobs()
        print(jobs)

        job_details = await client.fine_tuning.get_fine_tuning_job(job.name)
        print(job_details)


asyncio.run(main())

Usage Tracking

import os
import asyncio
from impulse.api_sdk.sdk import ImpulseSDK
from datetime import timedelta, date


async def main():
    async with ImpulseSDK(os.environ.get("IMPSDK_API_KEY")) as client:
        end_date = date.today()
        start_date = end_date - timedelta(days=30)
        total_cost = await client.usage.get_total_cost(start_date, end_date)
        print(total_cost)

        usage_records = await client.usage.list_usage_records(start_date, end_date)
        print(usage_records)


asyncio.run(main())

Support

If you encounter any problems or have any questions, please open an issue on the GitHub repository.