# Files

"Files" are used to upload documents that can be used with features like "**Fine-tuning**", and "**Inference Batch API**". You need to first create the file object using "**/v1/files**" and get the **presigned upload url**. Then upload the file through presigned url.

<table><thead><tr><th>Functionality</th><th width="173">Max File Size</th><th width="173">File Extension</th><th data-type="checkbox">File Format Required</th></tr></thead><tbody><tr><td>Fine-tuning</td><td>512 MB</td><td>.jsonl</td><td>true</td></tr><tr><td>Inference Batch API</td><td>100 MB</td><td>.jsonl</td><td>true</td></tr></tbody></table>

{% hint style="info" %}
The total storage limit for all files uploaded by a single organization is capped at **100 GB**.

For Fine-tuning, we only support data in the [**OpenAI fine-tuning format**](https://platform.openai.com/docs/guides/fine-tuning#preparing-your-dataset).&#x20;
{% endhint %}

The file upload API returns a **presigned upload URL,** and the file needs to be uploaded using this presigned URL. The presigned URL is valid for **2 hours**.

## Step 1: Create a File Object

In this step, you create a file object, which initializes the upload process by providing essential metadata like the file name and type. The "purpose" parameter can be set to either "fine-tune" or "batch".

{% hint style="info" %}
Please note that when passing the "**file\_name**" parameter, you need to include the **file extension**. Otherwise, we will default to treating it as a JSONL file.
{% endhint %}

**Example Request:**

{% hint style="info" %}
Replace `{{API_TOKEN}}` with your actual token.
{% endhint %}

{% tabs %}
{% tab title="Curl" %}

```bash
curl --location 'https://api.netmind.ai/v1/files' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer {{API_TOKEN}}' \
--data '{"file_name":"text.jsonl","purpose":"fine-tune"}'
```

{% endtab %}

{% tab title="Python" %}

```python
import requests
import json

url = "https://api.netmind.ai/v1/files"

payload = json.dumps({
  "file_name": "text.jsonl",
  "purpose": "fine-tune"
})
headers = {
  'Content-Type': 'application/json',
  'Authorization': 'Bearer {{API_TOKEN}}'
}

response = requests.request("POST", url, headers=headers, data=payload)

print(response.text)
```

{% endtab %}
{% endtabs %}

**Example Response:**

```json
{
    "id": "c67b7...298f3",
    "presigned_url": "https://FAKE_PRESIGNED_URL/..."
}
```

***

## Step 2: Upload File Using Presigned URL

**Example Request:**

{% hint style="info" %}
Replace `{{`DATA\_FILE\_PATH`}}` with your actual data file path.
{% endhint %}

{% tabs %}
{% tab title="Curl" %}

```bash
curl --location --request PUT 'https://FAKE_PRESIGNED_URL/...' \
--data '@{{DATA_FILE_PATH}}'
```

{% endtab %}

{% tab title="Python" %}

```python
import requests
import os


presigned_url = "https://FAKE_PRESIGNED_URL/..."
file_name = "{{DATA_FILE_PATH}}"
file_size = os.path.getsize(file_name)


with open(file_name, "rb") as f:
    response = requests.put(presigned_url, data=f)
    print(response.status_code)
    print(response.text)
```

{% endtab %}
{% endtabs %}

**Example Response:**

```json
{}
```


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://netmind-power.gitbook.io/netmind-power-documentation/api/files.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.