* Remove unnecessary cast of id -> str (can do with a field setting)
* Remove unnecessary `set_text` model validator (can be done with a
computed field - though we had to make some changes to the `Generation`
class to make this possible
Before: ~2.4s
Blue circles represent time spent in custom validators :(
<img width="1337" alt="Screenshot 2025-05-14 at 10 10 12 AM"
src="https://github.com/user-attachments/assets/bb4f477f-4ee3-4870-ae93-14ca7f197d55"
/>
After: ~2.2s
<img width="1344" alt="Screenshot 2025-05-14 at 10 11 03 AM"
src="https://github.com/user-attachments/assets/99f97d80-49de-462f-856f-9e7e8662adbc"
/>
We still want to optimize the backwards compatible tool calls model
validator, though I think this might involve breaking changes, so wanted
to separate that into a different PR. This is circled in green.
When aggregating AIMessageChunks in a stream, core prefers the leftmost
non-null ID. This is problematic because:
- Core assigns IDs when they are null to `f"run-{run_manager.run_id}"`
- The desired meaningful ID might not be available until midway through
the stream, as is the case for the OpenAI Responses API.
For the OpenAI Responses API, we assign message IDs to the top-level
`AIMessage.id`. This works in `.(a)invoke`, but during `.(a)stream` the
IDs get overwritten by the defaults assigned in langchain-core. These
IDs
[must](https://community.openai.com/t/how-to-solve-badrequesterror-400-item-rs-of-type-reasoning-was-provided-without-its-required-following-item-error-in-responses-api/1151686/9)
be available on the AIMessage object to support passing reasoning items
back to the API (e.g., if not using OpenAI's `previous_response_id`
feature). We could add them elsewhere, but seeing as we've already made
the decision to store them in `.id` during `.(a)invoke`, addressing the
issue in core lets us fix the problem with no interface changes.
We only need to rebuild model schemas if type annotation information
isn't available during declaration - that shouldn't be the case for
these types corrected here.
Need to do more thorough testing to make sure these structures have
complete schemas, but hopefully this boosts startup / import time.
Most easily reviewed with the "hide whitespace" option toggled.
Seeing 10-50% speed ups in import time for common structures 🚀
The general purpose of this PR is to lazily import structures within
`langchain_core.XXX_module.__init__.py` so that we're not eagerly
importing expensive dependencies (`pydantic`, `requests`, etc).
Analysis of flamegraphs generated with `importtime` motivated these
changes. For example, the one below demonstrates that importing
`HumanMessage` accidentally triggered imports for `importlib.metadata`,
`requests`, etc.
There's still much more to do on this front, and we can start digging
into our own internal code for optimizations now that we're less
concerned about external imports.
<img width="1210" alt="Screenshot 2025-04-11 at 1 10 54 PM"
src="https://github.com/user-attachments/assets/112a3fe7-24a9-4294-92c1-d5ae64df839e"
/>
I've tracked the improvements with some local benchmarks:
## `pytest-benchmark` results
| Name | Before (s) | After (s) | Delta (s) | % Change |
|-----------------------------|------------|-----------|-----------|----------|
| Document | 2.8683 | 1.2775 | -1.5908 | -55.46% |
| HumanMessage | 2.2358 | 1.1673 | -1.0685 | -47.79% |
| ChatPromptTemplate | 5.5235 | 2.9709 | -2.5526 | -46.22% |
| Runnable | 2.9423 | 1.7793 | -1.163 | -39.53% |
| InMemoryVectorStore | 3.1180 | 1.8417 | -1.2763 | -40.93% |
| RunnableLambda | 2.7385 | 1.8745 | -0.864 | -31.55% |
| tool | 5.1231 | 4.0771 | -1.046 | -20.42% |
| CallbackManager | 4.2263 | 3.4099 | -0.8164 | -19.32% |
| LangChainTracer | 3.8394 | 3.3101 | -0.5293 | -13.79% |
| BaseChatModel | 4.3317 | 3.8806 | -0.4511 | -10.41% |
| PydanticOutputParser | 3.2036 | 3.2995 | 0.0959 | 2.99% |
| InMemoryRateLimiter | 0.5311 | 0.5995 | 0.0684 | 12.88% |
Note the lack of change for `InMemoryRateLimiter` and
`PydanticOutputParser` is just random noise, I'm getting comparable
numbers locally.
## Local CodSpeed results
We're still working on configuring CodSpeed on CI. The local usage
produced similar results.
Looks like `pyupgrade` was already used here but missed some docs and
tests.
This helps to keep our docs looking professional and up to date.
Eventually, we should lint / format our inline docs.
- Support thinking blocks in core's `convert_to_openai_messages` (pass
through instead of error)
- Ignore thinking blocks in ChatOpenAI (instead of error)
- Support Anthropic-style image blocks in ChatOpenAI
---
Standard integration tests include a `supports_anthropic_inputs`
property which is currently enabled only for tests on `ChatAnthropic`.
This test enforces compatibility with message histories of the form:
```
- system message
- human message
- AI message with tool calls specified only through `tool_use` content blocks
- human message containing `tool_result` and an additional `text` block
```
It additionally checks support for Anthropic-style image inputs if
`supports_image_inputs` is enabled.
Here we change this test, such that if you enable
`supports_anthropic_inputs`:
- You support AI messages with text and `tool_use` content blocks
- You support Anthropic-style image inputs (if `supports_image_inputs`
is enabled)
- You support thinking content blocks.
That is, we add a test case for thinking content blocks, but we also
remove the requirement of handling tool results within HumanMessages
(motivated by existing agent abstractions, which should all return
ToolMessage). We move that requirement to a ChatAnthropic-specific test.
Description:
The change allows you to use the overloaded `+` operator correctly when
`+`ing two BaseMessageChunk subclasses. Without this you *must*
instantiate a subclass for it to work.
Which feels... wrong. Base classes should be decoupled from sub classes
and should have in no way a dependency on them.
Issue:
You can't `+` a BaseMessageChunk with a BaseMessageChunk
e.g. this will explode
```py
from langchain_core.outputs import (
ChatGenerationChunk,
)
from langchain_core.messages import BaseMessageChunk
chunk1 = ChatGenerationChunk(
message=BaseMessageChunk(
type="customChunk",
content="HI",
),
)
chunk2 = ChatGenerationChunk(
message=BaseMessageChunk(
type="customChunk",
content="HI",
),
)
# this will throw
new_chunk = chunk1 + chunk2
```
In case anyone ran into this issue themselves, it's probably best to use
the AIMessageChunk:
a la
```py
from langchain_core.outputs import (
ChatGenerationChunk,
)
from langchain_core.messages import AIMessageChunk
chunk1 = ChatGenerationChunk(
message=AIMessageChunk(
content="HI",
),
)
chunk2 = ChatGenerationChunk(
message=AIMessageChunk(
content="HI",
),
)
# No explosion!
new_chunk = chunk1 + chunk2
```
Dependencies:
None!
Twitter handle:
`aaron_vogler`
Keeping these for later if need be:
```
baskaryan
efriis
eyurtsev
ccurme
vbarda
hwchase17
baskaryan
efriis
```
Co-authored-by: Erick Friis <erick@langchain.dev>
TRY004 ("use TypeError rather than ValueError") existing errors are
marked as ignore to preserve backward compatibility.
LMK if you prefer to fix some of them.
Co-authored-by: Erick Friis <erick@langchain.dev>
- Convert developer openai messages to SystemMessage
- store additional_kwargs={"__openai_role__": "developer"} so that the
correct role can be reconstructed if needed
- update ChatOpenAI to read in openai_role
---------
Co-authored-by: Erick Friis <erick@langchain.dev>
