Skip to main content

Trubrics

Trubrics is an LLM user analytics platform that lets you collect, analyse and manage user prompts & feedback on AI models.

Check out Trubrics repo for more information on Trubrics.

In this guide, we will go over how to set up the TrubricsCallbackHandler.

Installation and Setup

%pip install --upgrade --quiet  trubrics

Getting Trubrics Credentials

If you do not have a Trubrics account, create one on here. In this tutorial, we will use the default project that is built upon account creation.

Now set your credentials as environment variables:

import os

os.environ["TRUBRICS_EMAIL"] = "***@***"
os.environ["TRUBRICS_PASSWORD"] = "***"

Usage

The TrubricsCallbackHandler can receive various optional arguments. See here for kwargs that can be passed to Trubrics prompts.

class TrubricsCallbackHandler(BaseCallbackHandler):

"""
Callback handler for Trubrics.

Args:
project: a trubrics project, default project is "default"
email: a trubrics account email, can equally be set in env variables
password: a trubrics account password, can equally be set in env variables
**kwargs: all other kwargs are parsed and set to trubrics prompt variables, or added to the `metadata` dict
"""

Examples

Here are two examples of how to use the TrubricsCallbackHandler with Langchain LLMs or Chat Models. We will use OpenAI models, so set your OPENAI_API_KEY key here:

os.environ["OPENAI_API_KEY"] = "sk-***"

1. With an LLM

from langchain.callbacks import TrubricsCallbackHandler
from langchain_openai import OpenAI
llm = OpenAI(callbacks=[TrubricsCallbackHandler()])
2023-09-26 11:30:02.149 | INFO     | trubrics.platform.auth:get_trubrics_auth_token:61 - User jeff.kayne@trubrics.com has been authenticated.
res = llm.generate(["Tell me a joke", "Write me a poem"])
2023-09-26 11:30:07.760 | INFO     | trubrics.platform:log_prompt:102 - User prompt saved to Trubrics.
2023-09-26 11:30:08.042 | INFO | trubrics.platform:log_prompt:102 - User prompt saved to Trubrics.
print("--> GPT's joke: ", res.generations[0][0].text)
print()
print("--> GPT's poem: ", res.generations[1][0].text)
--> GPT's joke:  

Q: What did the fish say when it hit the wall?
A: Dam!

--> GPT's poem:

A Poem of Reflection

I stand here in the night,
The stars above me filling my sight.
I feel such a deep connection,
To the world and all its perfection.

A moment of clarity,
The calmness in the air so serene.
My mind is filled with peace,
And I am released.

The past and the present,
My thoughts create a pleasant sentiment.
My heart is full of joy,
My soul soars like a toy.

I reflect on my life,
And the choices I have made.
My struggles and my strife,
The lessons I have paid.

The future is a mystery,
But I am ready to take the leap.
I am ready to take the lead,
And to create my own destiny.

2. With a chat model

from langchain.callbacks import TrubricsCallbackHandler
from langchain_core.messages import HumanMessage, SystemMessage
from langchain_openai import ChatOpenAI
chat_llm = ChatOpenAI(
callbacks=[
TrubricsCallbackHandler(
project="default",
tags=["chat model"],
user_id="user-id-1234",
some_metadata={"hello": [1, 2]},
)
]
)
chat_res = chat_llm(
[
SystemMessage(content="Every answer of yours must be about OpenAI."),
HumanMessage(content="Tell me a joke"),
]
)
2023-09-26 11:30:10.550 | INFO     | trubrics.platform:log_prompt:102 - User prompt saved to Trubrics.
print(chat_res.content)
Why did the OpenAI computer go to the party?

Because it wanted to meet its AI friends and have a byte of fun!