Skip to main content

promptwatch.io python client to trace langchain sessions

Project description

PromptWatch.io ... session tracking for LangChain

It enables you to:

  • track all the chain executions
  • track LLM Prompts and re-play the LLM runs with the same input parameters and model settings to tweak your prompt template
  • track your costs per project and per tenant (your customer)

Installation

pip install promptwatch

Basic usage

In order to enable session tracking wrap you chain executions in PromptWatch block

from langchain import OpenAI, LLMChain, PromptTemplate
from promptwatch import PromptWatch

prompt_template = PromptTemplate.from_template("Finish this sentence {input}")
my_chain = LLMChain(llm=OpenAI(), prompt=prompt_template)

with PromptWatch(api_key="<your-api-key>") as pw:
    my_chain("The quick brown fox jumped over")

Here you can get your API key: http://www.promptwatch.io/get-api-key (no registration needed)

You can set it directly into PromptWatch constructor, or set is as an ENV variable PROMPTWATCH_API_KEY

Comprehensive Chain Execution Tracking

With PromptWatch.io, you can track all chains, actions, retrieved documents, and more to gain complete visibility into your system. This makes it easy to identify issues with your prompts and quickly fix them for optimal performance.

What sets PromptWatch.io apart is its intuitive and visual interface. You can easily drill down into the chains to find the root cause of any problems and get a clear understanding of what's happening in your system.

Read more here: Chain tracing documentation

LLM Prompt caching

It is often tha case that some of the prompts are repeated over an over. It is costly and slow. With PromptWatch you just wrap your LLM model into our CachedLLM interface and it will automatically reuse previously generated values.

Read more here: Prompt caching documentation

LLM Prompt Template Tweaking

Tweaking prompt templates to find the optimal variation can be a time-consuming and challenging process, especially when dealing with multi-stage LLM chains. Fortunately, PromptWatch.io can help simplify the process!

With PromptWatch.io, you can easily experiment with different prompt variants by replaying any given LLM chain with the exact same inputs used in real scenarios. This allows you to fine-tune your prompts until you find the variation that works best for your needs.

Read more here: Prompt tweaking documentation

Keep Track of Your Prompt Template Changes

Making changes to your prompt templates can be a delicate process, and it's not always easy to know what impact those changes will have on your system. Version control platforms like GIT are great for tracking code changes, but they're not always the best solution for tracking prompt changes.

Read more here: Prompt template versioning documentation

Unit testing

Unit tests will help you understand what impact your changes in Prompt templates and your code can have on representative sessions examples.

Read more here: Unit tests documentation

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

promptwatch-0.4.0.tar.gz (37.1 kB view details)

Uploaded Source

Built Distribution

promptwatch-0.4.0-py3-none-any.whl (41.0 kB view details)

Uploaded Python 3

File details

Details for the file promptwatch-0.4.0.tar.gz.

File metadata

  • Download URL: promptwatch-0.4.0.tar.gz
  • Upload date:
  • Size: 37.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.9.16

File hashes

Hashes for promptwatch-0.4.0.tar.gz
Algorithm Hash digest
SHA256 6900fe8d0bbf91a984b2355c436cc600a09ae4bf5c0c84257ee3534319078c02
MD5 7126f8926b68465596456b04e99d49c0
BLAKE2b-256 4bac54667ecb5f812e3981672613293b2aed15a713932edc86859e84275076da

See more details on using hashes here.

File details

Details for the file promptwatch-0.4.0-py3-none-any.whl.

File metadata

  • Download URL: promptwatch-0.4.0-py3-none-any.whl
  • Upload date:
  • Size: 41.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.1 CPython/3.9.16

File hashes

Hashes for promptwatch-0.4.0-py3-none-any.whl
Algorithm Hash digest
SHA256 0a1cc18bdcd72267ea11b35f8c919c574a8d7159f39ce873f7182516bcfa627a
MD5 b83ecf01c127b6067a5655e3bc1b9807
BLAKE2b-256 a593a9009d092af6e98af6fbb1ea95e466bd0e4e00402d7b23f7b664b0cfbeff

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page