Dependencies issues when installing new Packages

[latifs]

Checked other resources

• I added a very descriptive title to this question.
• I searched the LangChain documentation with the integrated search.
• I used the GitHub search to find a similar question and didn't find it.

Commit to Help

• I commit to help with one of those options 👆

Example Code

none

Description

Hey guys,
I'd love to start using Langserve as it seems to be an amazing tool.
I'm also switching from LangchainJS to Langchain (python) since a lot of things are still missing in JS.
So the complexity is at the max.

The problem that I am having is that when I try to add a package, it comes with an extra pyproject.toml, on top of the root one for Langserve, that has extra dependencies and stuff, so far so good.

But when trying to install these a lot of version errors start to pop out of everywhere and when you try to modify one file the other one starts complaining about something else.

So my question is, do we need 2 pyproject files, do we merge the dependencies somehow?
is there a tool to fix that. Poetry doesn't seem to help right now.

Thanks

System Info

none

[dosubot[bot]]

It sounds like you're encountering version conflicts due to multiple pyproject.toml files when trying to install new packages. Here are some steps and tools that might help resolve these issues:

1. Single pyproject.toml File:

   • Ideally, you should have a single pyproject.toml file at the root of your project. This file should include all the dependencies required for your project, including those for Langserve and any other packages you need.

2. Merging Dependencies:

   • Manually merge the dependencies from the extra pyproject.toml file into the root pyproject.toml file. Ensure that you resolve any version conflicts by specifying compatible versions for each dependency.

3. Using Poetry:

   • Poetry is a great tool for managing dependencies. If you haven't already, you can install Poetry with pip install poetry.
   • Use poetry add <package> to add new dependencies. Poetry will handle version conflicts and update the pyproject.toml file accordingly.

4. Common Version Conflicts:

   • Pydantic Version Conflicts: Ensure that your dependencies are compatible with Pydantic v2. Update your Pydantic version constraints to pydantic>=2,<3.
   • Python Version Conflicts: Ensure your Python version is at least 3.9, as Python 3.8 is no longer supported.

5. Automated Tools:

   • Use langchain-cli to help with migration and updating deprecated imports. Install it with pip install -U langchain-cli and run langchain-cli migrate [path to code].

For more detailed steps on resolving common version conflicts, you can refer to the LangChain documentation ^[1].

If you need further assistance, feel free to provide specific error messages, and I can help troubleshoot those as well.

To continue talking to Dosu, mention @dosu.

[latifs]

@dosu
I've tried a lot of different things i've made sure to remove the pyproject files inside the added package and have only one at the root now.

I've also tried to start from a clean slate doing this:

poetry env remove python
rm poetry.lock  # If it exists
poetry cache clear pypi --all
poetry install

Here is my pyproject.toml file:

[tool.poetry]
name = "langserve-demo"
version = "0.1.0"
description = ""
authors = ["Your Name <[email protected]>"]
readme = "README.md"
packages = [
    { include = "app" },
]

[tool.langserve]
export_module = "pirate_speak.chain"
export_attr = "chain"

[tool.templates-hub]
use-case = "chatbot"
author = "LangChain"
integrations = ["OpenAI"]
tags = ["getting-started"]

[tool.poetry.dependencies]
python = "^3.11"
uvicorn = "^0.23.2"
langserve = {extras = ["server"], version = ">=0.0.30"}
pydantic = "^2"
fastapi = "^0.115.0"
pirate-speak = {path = "packages/pirate-speak", develop = true}
langchain-openai = "^0.0.2"
langchain = "^0.1.0"
langchain-core = "^0.1.7"

[tool.poetry.group.dev.dependencies]
langchain-cli = "^0.0.15"

[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"

and here is the crazy error I get after poetry install:

Because no versions of langchain match >0.1.0,<0.1.1 || >0.1.1,<0.1.2 || >0.1.2,<0.1.3 || >0.1.3,<0.1.4 || >0.1.4,<0.1.5 || >0.1.5,<0.1.6 || >0.1.6,<0.1.7 || >0.1.7,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0
 and langchain (0.1.5) depends on langchain-core (>=0.1.16,<0.2), langchain (>0.1.0,<0.1.1 || >0.1.1,<0.1.2 || >0.1.2,<0.1.3 || >0.1.3,<0.1.4 || >0.1.4,<0.1.6 || >0.1.6,<0.1.7 || >0.1.7,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.16,<0.2).
And because langchain (0.1.4) depends on langchain-core (>=0.1.16,<0.2), langchain (>0.1.0,<0.1.1 || >0.1.1,<0.1.2 || >0.1.2,<0.1.3 || >0.1.3,<0.1.6 || >0.1.6,<0.1.7 || >0.1.7,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.16,<0.2).
And because langchain (0.1.3) depends on langchain-core (>=0.1.14,<0.2)
 and langchain (0.1.2) depends on langchain-core (>=0.1.14,<0.2), langchain (>0.1.0,<0.1.1 || >0.1.1,<0.1.6 || >0.1.6,<0.1.7 || >0.1.7,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.14,<0.2).
And because langchain (0.1.1) depends on langchain-core (>=0.1.9,<0.2)
 and langchain (0.1.0) depends on langchain-core (>=0.1.7,<0.2), langchain (>=0.1.0,<0.1.6 || >0.1.6,<0.1.7 || >0.1.7,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2).
And because langchain (0.1.6) depends on langchain-core (>=0.1.22,<0.2)
 and langchain (0.1.7) depends on langchain-core (>=0.1.22,<0.2), langchain (>=0.1.0,<0.1.8 || >0.1.8,<0.1.9 || >0.1.9,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2).
And because langchain (0.1.8) depends on langchain-core (>=0.1.24,<0.2)
 and langchain (0.1.9) depends on langchain-core (>=0.1.26,<0.2), langchain (>=0.1.0,<0.1.10 || >0.1.10,<0.1.11 || >0.1.11,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2).
And because langchain (0.1.10) depends on langchain-core (>=0.1.28,<0.2)
 and langchain (0.1.11) depends on langchain-core (>=0.1.29,<0.2), langchain (>=0.1.0,<0.1.12 || >0.1.12,<0.1.13 || >0.1.13,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2).
And because langchain (0.1.12) depends on langchain-core (>=0.1.31,<0.2.0)
 and langchain (0.1.13) depends on langchain-core (>=0.1.33,<0.2.0), langchain (>=0.1.0,<0.1.14 || >0.1.14,<0.1.15 || >0.1.15,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2.0).
And because langchain (0.1.14) depends on langchain-core (>=0.1.37,<0.2.0)
 and langchain (0.1.15) depends on langchain-core (>=0.1.41,<0.2.0), langchain (>=0.1.0,<0.1.16 || >0.1.16,<0.1.17 || >0.1.17,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2.0).
And because langchain (0.1.16) depends on langchain-core (>=0.1.42,<0.2.0)
 and langchain (0.1.17) depends on langchain-core (>=0.1.48,<0.2.0), langchain (>=0.1.0,<0.1.19 || >0.1.19,<0.1.20 || >0.1.20,<0.2.0) requires langchain-core (>=0.1.7,<0.2.0).
And because langchain (0.1.19) depends on langchain-core (>=0.1.52,<0.2.0)
 and langchain (0.1.20) depends on langchain-core (>=0.1.52,<0.2.0), langchain (>=0.1.0,<0.2.0) requires langchain-core (>=0.1.7,<0.2.0).
Because no versions of langserve match >0.3.0,<0.4.0
 and langserve[server] (0.3.0) depends on langchain-core (>=0.3,<0.4), langserve[server] (>=0.3.0,<0.4.0) requires langchain-core (>=0.3,<0.4).
Thus, langserve[server] (>=0.3.0,<0.4.0) is incompatible with langchain (>=0.1.0,<0.2.0).
So, because langserve-pirate-speak depends on both langserve[server] (^0.3.0) and langchain (^0.1.0), version solving failed.

[dosubot[bot]]

Could you provide the specific versions of the langchain and langserve packages you are trying to use? Additionally, are there any other dependencies or constraints that might be affecting the version resolution?

To continue talking to Dosu, mention @dosu.

[latifs]

@dosu,
at this point I'd be fine with any combination of recent versions of langchain and langserve.
I'd just like to be able to do langchain serve and start testing the tool.
So if you have a started template that includes poetry, langserve and one package (like pirate-speak) i'd be fine with that.

[dosubot[bot]]

Here's a starter template that includes Poetry, Langserve, and the Pirate-Speak package with compatible versions of LangChain and Langserve:

[tool.poetry]
name = "pirate-speak"
version = "0.0.1"
description = "Get started with a simple template that speaks like a pirate"
authors = []
readme = "README.md"

[tool.poetry.dependencies]
python = ">=3.9,<4.0"
langchain-core = "^0.3.0"
langchain-openai = ">=0.0.1"
openai = "<2"
anthropic = "^0.6.0"
cohere = "^4.34"

[tool.poetry.group.dev.dependencies]
langchain-cli = ">=0.0.21"
fastapi = "^0.104.0"
sse-starlette = "^1.6.5"

[tool.langserve]
export_module = "pirate_speak_configurable"
export_attr = "chain"

[tool.templates-hub]
use-case = "chatbot"
author = "LangChain"
integrations = ["Cohere", "Anthropic", "OpenAI"]
tags = ["configurable"]

[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"

Environment Setup

Set the following environment variables to access all 3 configurable alternative model providers:

• OPENAI_API_KEY
• ANTHROPIC_API_KEY
• COHERE_API_KEY

Usage

To use this package, you should first have the LangChain CLI installed:

pip install -U langchain-cli

To create a new LangChain project and install this as the only package, you can do:

langchain app new my-app --package pirate-speak-configurable

If you want to add this to an existing project, you can just run:

langchain app add pirate-speak-configurable

And add the following code to your server.py file:

from pirate_speak_configurable import chain as pirate_speak_configurable_chain

add_routes(app, pirate_speak_configurable_chain, path="/pirate-speak-configurable")

(Optional) Let's now configure LangSmith.
LangSmith will help us trace, monitor and debug LangChain applications.
You can sign up for LangSmith here.
If you don't have access, you can skip this section

export LANGCHAIN_TRACING_V2=true
export LANGCHAIN_API_KEY=<your-api-key>
export LANGCHAIN_PROJECT=<your-project>  # if not specified, defaults to "default"

If you are inside this directory, then you can spin up a LangServe instance directly by:

langchain serve

This will start the FastAPI app with a server running locally at
http://localhost:8000

We can see all templates at http://127.0.0.1:8000/docs
We can access the playground at http://127.0.0.1:8000/pirate-speak-configurable/playground

We can access the template from code with:

from langserve.client import RemoteRunnable

runnable = RemoteRunnable("http://localhost:8000/pirate-speak-configurable")

This setup should help you get started with LangServe and the Pirate-Speak package ^{[1][2][3][4][5]}.

To continue talking to Dosu, mention @dosu.