FinetuneDB

https://docs.finetunedb.com/
Admin
FinetuneDB is an end-to-end platform for creating datasets, fine-tuning LLMs, evaluating model...
Tokens:8,606
Snippets:150
Trust Score:7.7
Update:1 month ago
Show doc for...
Tokens:
Raw
### Install FinetuneDB SDK

Source: https://docs.finetunedb.com/logging-data/node-js-sdk

Install the FinetuneDB Node.js SDK using npm or yarn.

```sh
npm install --save finetunedb
# or
yarn add finetunedb
```

--------------------------------

### OpenAI Completion Output Example

Source: https://docs.finetunedb.com/logging-data/web-api

Example of formatting the response from openai.chat.completions.create for log output.

```javascript
const response = await openai.chat.completions.create(...);
const output = [response.choices[0].message];
```

--------------------------------

### OpenAI Embedding Output Example

Source: https://docs.finetunedb.com/logging-data/web-api

Example of formatting the response.data from openai.embeddings.create for log output.

```javascript
const response = await openai.embeddings.create(...);
const output = response.data;
```

--------------------------------

### OpenAI Completion Input Example

Source: https://docs.finetunedb.com/logging-data/web-api

Example of the messages array structure used for CHATCOMPLETION type inputs.

```json
[
  {
    "role": "system",
    "content": "You are an expert assistant"
  },
  {
    "role": "user",
    "content": "Hello, how are you?"
  }
]
```

--------------------------------

### JSONL Export Format Example

Source: https://docs.finetunedb.com/features/dataset-manager/export-datasets

Shows the JSONL format used when exporting a dataset. Each line is a JSON object with a 'messages' array containing system, user, and assistant roles. This example includes two sample conversations.

```json
{"messages": [{"role": "system", "content": "Marv is a factual chatbot that is also sarcastic."}, {"role": "user", "content": "What's the capital of France?"}, {"role": "assistant", "content": "Paris, as if everyone doesn't know that already."}]}
{"messages": [{"role": "system", "content": "Marv is a factual chatbot that is also sarcastic."}, {"role": "user", "content": "Who wrote 'Romeo and Juliet'?"}, {"role": "assistant", "content": "Oh, just some guy named William Shakespeare. Ever heard of him?"}]}
```

--------------------------------

### OpenAI Embedding Input Example (String)

Source: https://docs.finetunedb.com/logging-data/web-api

Example of a single string input for EMBEDDING type logs.

```json
"Hello world!"
```

--------------------------------

### OpenAI Embedding Input Example (Array)

Source: https://docs.finetunedb.com/logging-data/web-api

Example of an array of strings input for EMBEDDING type logs.

```json
["Hello", "world!"]
```

--------------------------------

### Initialize and Use FinetuneDB Client

Source: https://docs.finetunedb.com/logging-data/node-js-sdk

Initialize the OpenAI client with FinetuneDB logging enabled, create a chat completion, and update the log with custom tags.

```js
// import OpenAI from 'openai'
import OpenAI from "finetunedb/openai";

// Fully compatible with original OpenAI initialization
const openai = new OpenAI({
  apiKey: "your api key", // defaults to process.env["OPENAI_API_KEY"]
  // Optional: Step 1 - Initialize FinetuneDB to enable logging
  finetunedb: {
    apiKey: "your api key", // defaults to process.env["FINETUNEDB_API_KEY"]
  },
});

async function main() {
  // Allows optional finetunedb object
  const completion = await openai.chat.completions.create({
    messages: [{ role: "user", content: "Hello world!" }],
    model: "gpt-3.5-turbo",
    // Optional: Step 2 - Log a chat completion
    finetunedb: {
      // Define the project ID
      projectId: "cloio7t90000...",
      // Enable/disable data collection. Defaults to true.
      logRequest: true,
      // Optional: Add custom searchable tags
      tags: ["test-prompt"],
    },
  });

  // Optional: Update log
  await completion.finetunedb.updateLastLog({
    tags: ["test-prompt", "new-tag"],
  });
}

main();
```

--------------------------------

### POST /logs

Source: https://docs.finetunedb.com/logging-data/web-api

Creates a new log entry in FineTuneDB. Accepts parameters for project identification, model configuration, input/output data, and metadata.

```APIDOC
## POST /logs

### Description
Creates a new log entry in FineTuneDB.

### Method
POST

### Endpoint
https://app.finetunedb.com/api/v1/logs

### Parameters
#### Request Body
- **projectId** (string) - Required - The project identifier.
- **parentId** (string) - Optional - Parent log identifier.
- **name** (string) - Optional - Name of the log.
- **provider** (enum: openai) - Required - The provider.
- **tags** (array of strings) - Optional - Tags for the log.
- **model** (string) - Required - The model used.
- **source** (string) - Required - The source of the log.
- **modelParameters** (object) - Optional - Model configuration parameters.
  - **model** (string) - Required - The model name.
  - **provider** (string) - Optional - The provider.
  - **max_tokens** (number) - Optional - Maximum tokens.
  - **temperature** (number) - Optional - Sampling temperature.
  - **top_p** (number) - Optional - Top-p sampling.
  - **stop_sequences** (array of strings) - Optional - Stop sequences.
  - **presence_penalty** (number) - Optional - Presence penalty.
  - **frequency_penalty** (number) - Optional - Frequency penalty.
  - **other** (record of any type) - Optional - Additional parameters.
- **input** (union of input types) - Optional - Input data (see Input Types).
- **output** (union of output types) - Optional - Output data (see Output Types).
- **metadata** (record of any type) - Optional - Metadata.
- **type** (enum: CHATCOMPLETION, EMBEDDING) - Required - The type of log.
- **error** (string) - Optional - Error message.
- **latencyMs** (number) - Optional - Latency in milliseconds.

### Input Types
#### OpenAI Completion (If type=CHATCOMPLETION)
Pass the messages value used in openai.chat.completions.create.
Example:
```json
[
  {
    "role": "system",
    "content": "You are an expert assistant"
  },
  {
    "role": "user",
    "content": "Hello, how are you?"
  }
]
```

#### OpenAI Embedding (If type=EMBEDDING)
Pass the input value sent to openai.embeddings.create.
Example:
```json
"Hello world!"
```
or
```json
["Hello", "world!"]
```

### Output Types
#### OpenAI Completion (If type=CHATCOMPLETION)
Pass the response value from openai.chat.completions.create as an array.
Example:
```javascript
const response = await openai.chat.completions.create(...);
const output = [response.choices[0].message];
```

#### OpenAI Embedding (If type=EMBEDDING)
Pass the response.data value from openai.embeddings.create.
Example:
```javascript
const response = await openai.embeddings.create(...);
const output = response.data;
```

```
Tokens:
Raw