Files
langchain/libs/community/langchain_community/llms/manifest.py
Eugene Yurtsev 844955d6e1 community[patch]: assign missed default (#26326)
Assigning missed defaults in various classes. Most clients were being
assigned during the `model_validator(mode="before")` step, so this
change should amount to a no-op in those cases.

---

This PR was autogenerated using gritql

```shell

grit apply 'class_definition(name=$C, $body, superclasses=$S) where {    
    $C <: ! "Config", // Does not work in this scope, but works after class_definition
    $body <: block($statements),
    $statements <: some bubble assignment(left=$x, right=$y, type=$t) as $A where {
        or {
            $y <: `Field($z)`,
            $x <: "model_config"
        }
    },
    // And has either Any or Optional fields without a default
    $statements <: some bubble assignment(left=$x, right=$y, type=$t) as $A where {
        $t <: or {
            r"Optional.*",
            r"Any",
            r"Union[None, .*]",
            r"Union[.*, None, .*]",
            r"Union[.*, None]",
        },
        $y <: ., // Match empty node        
        $t => `$t = None`,
    },    
}
' --language python .

```
2024-09-11 11:13:11 -04:00

64 lines
1.9 KiB
Python

from typing import Any, Dict, List, Mapping, Optional
from langchain_core.callbacks import CallbackManagerForLLMRun
from langchain_core.language_models.llms import LLM
from langchain_core.utils import pre_init
from pydantic import ConfigDict
class ManifestWrapper(LLM):
"""HazyResearch's Manifest library."""
client: Any = None #: :meta private:
llm_kwargs: Optional[Dict] = None
model_config = ConfigDict(
extra="forbid",
)
@pre_init
def validate_environment(cls, values: Dict) -> Dict:
"""Validate that python package exists in environment."""
try:
from manifest import Manifest
if not isinstance(values["client"], Manifest):
raise ValueError
except ImportError:
raise ImportError(
"Could not import manifest python package. "
"Please install it with `pip install manifest-ml`."
)
return values
@property
def _identifying_params(self) -> Mapping[str, Any]:
kwargs = self.llm_kwargs or {}
return {
**self.client.client_pool.get_current_client().get_model_params(),
**kwargs,
}
@property
def _llm_type(self) -> str:
"""Return type of llm."""
return "manifest"
def _call(
self,
prompt: str,
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> str:
"""Call out to LLM through Manifest."""
if stop is not None and len(stop) != 1:
raise NotImplementedError(
f"Manifest currently only supports a single stop token, got {stop}"
)
params = self.llm_kwargs or {}
params = {**params, **kwargs}
if stop is not None:
params["stop_token"] = stop
return self.client.run(prompt, **params)