Free AI web copilot to create summaries, insights and extended knowledge, download it at here

1517

Abstract

gues without connecting to the frontend.</p><p id="dd72">The configurability feature allows for the configuration of runnables, making experimentation with different parameters possible. This feature has been exposed in the playground as well as in the main WebLangChain app.</p><p id="06c2">Here’s an example of using the LangChain Expression Language to configure a component:</p><div id="6d30"><pre><span class="hljs-keyword">from</span> langchain <span class="hljs-keyword">import</span> expression

<span class="hljs-comment"># Configuring a component</span> configurable_component = expression.configure(<span class="hljs-string">"component_name"</span>, { <span class="hljs-string">"param1"</span>: <span class="hljs-string">"value1"</span>, <span class="hljs-string">"param2"</span>: <span class="hljs-string">"value2"</span> })</pre></div><p id="afe2">With configuration, you can save different versions of configurations via a URL. For example, WebLangChain allows end users to use configurations in the UI, and it can also be used for internal apps to let users choose between different configuration options.</p><p id="583e">To get a URL for a given configuration, use the following:</p><div id="1b10"><pre>configurable_component<span class="hljs-selector-class">.get_configuration_url</span>()</pre></div><p id="ff9c">Please note that the configuration options are not currently persisted, but there are plans for alpha testing a deployment platform where the configuration can be persisted for m

Options

ultiple users.</p><p id="c00b">The common architectures for LLM applications are also easily configurable. For example, the WebLangChain app is a cognitive architecture aimed at retrieval augmented generation, and it can be configured with parameters such as the LLM to use to answer, prompt to use to generate the search query for follow-ups, prompt to use to generate the answer, and retriever to use to look up queries.</p><p id="7959">This insight into common architectures is driving a lot of the work towards making application development easier and more collaborative.</p><div id="59c0" class="link-block"> <a href="https://readmedium.com/langchain-what-is-the-x-language-chain-a94e8a45e409"> <div> <div> <h2>LANGCHAIN — What Is the X-Language Chain?</h2> <div><h3>Every once in a while, a new technology, an old problem, and a big idea turn into an innovation. — Dean Kamen</h3></div> <div><p>medium.com</p></div> </div> <div> <div style="background-image: url(https://miro.readmedium.com/v2/resize:fit:320/1*nu7ZXSdSXeo6aCLEJYoZpg.jpeg)"></div> </div> </div> </a> </div><p id="13ea">In summary, LangServe Playground and Configurability provide a seamless and user-friendly way to experiment with different parameters and configurations, facilitating collaboration and making application development more efficient.</p></article></body>

LANGCHAIN — Is LangServe Playground Configurable?

Digital design is like painting, except the paint never dries. — Neville Brody

LANGCHAIN — Constructing Knowledge Graphs from Text Using OpenAI Functions

The great myth of our times is that technology is communication. — Libby Larsen

medium.com

LangServe Playground and Configurability

LangServe provides a playground experience that allows you to change configurable parameters and try out different inputs with real-time streamed responses. The playground offers a simple UI with streaming outputs, a full log of intermediate steps, and configurable options. This facilitates collaboration by providing a way for engineers to expose chains/agents to non-technical colleagues without connecting to the frontend.

The configurability feature allows for the configuration of runnables, making experimentation with different parameters possible. This feature has been exposed in the playground as well as in the main WebLangChain app.

Here’s an example of using the LangChain Expression Language to configure a component:

from langchain import expression

# Configuring a component
configurable_component = expression.configure("component_name", {
    "param1": "value1",
    "param2": "value2"
})

With configuration, you can save different versions of configurations via a URL. For example, WebLangChain allows end users to use configurations in the UI, and it can also be used for internal apps to let users choose between different configuration options.

To get a URL for a given configuration, use the following:

configurable_component.get_configuration_url()

Please note that the configuration options are not currently persisted, but there are plans for alpha testing a deployment platform where the configuration can be persisted for multiple users.

The common architectures for LLM applications are also easily configurable. For example, the WebLangChain app is a cognitive architecture aimed at retrieval augmented generation, and it can be configured with parameters such as the LLM to use to answer, prompt to use to generate the search query for follow-ups, prompt to use to generate the answer, and retriever to use to look up queries.

This insight into common architectures is driving a lot of the work towards making application development easier and more collaborative.

LANGCHAIN — What Is the X-Language Chain?

Every once in a while, a new technology, an old problem, and a big idea turn into an innovation. — Dean Kamen

medium.com

In summary, LangServe Playground and Configurability provide a seamless and user-friendly way to experiment with different parameters and configurations, facilitating collaboration and making application development more efficient.