Story/internal 501 update to langchain to 0 3 1 on chatbot ex #17
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
The assistant gets a bit to funny in it’s answers 😄  https://bitcrowd.atlassian.net/browse/INTERNAL-499
LLM frequently use Markdown to structure their output. sth like \
# lib/chatbot_web/live/chat_live.ex
# (..)
markdown_html =
String.trim(assigns.content)
|> Earmark.as_html!()
|> Phoenix.HTML.raw()
assigns =
assigns
|> assign(:class, "u-max-width-75 u-bg-white " <> justify_self)
|> assign(:markdown, markdown_html)
~H"""
<.ui_card id={@id} class={@Class}>
<%= @markdown %>
https://bitcrowd.atlassian.net/browse/INTERNAL-500
…chatbot.ex
Langchain.ex has progressed to 0.3.1. We should update to that. However,
it introduces a breaking change:
Compiling 10 files (.ex)
error: module LangChain.Utils.ApiOverride is not loaded and could not be found
│
12 │ import LangChain.Utils.ApiOverride
│ ^
│
└─ lib/chatbot/llm_mock.ex:12:3: Chatbot.LLMMock (module)
The background is
[this PR](brainlid/langchain#155):
> This change includes the Mimic testing library for mocking out live
> calls and forcing a specific return value.
>
> Originally, was using `LangChain.Utils.ApiOverride` which provided
> process-level result overrides. However, it had the negative
> requirement of requiring testing-aware to be in the production code,
> specifically in the ChatXYZ.call functions.
>
> Using Mimic removes that requirement, keeping the production level
> code cleaner.
The reasoning is sound, however, it crashes with what we do in
`Chatbot.LLMMock`. As LLM Mock is not used in tests (and `LLMMock` does
not have tests itself), I would take it out for now. 
https://bitcrowd.atlassian.net/browse/INTERNAL-501
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
rag 0.2 is not released yet. We should do that by Wednesday. When done, we should update the blogpost (see below) and check that code examples still run
Add
ragas dependency and runmix deps.get.