Merge pull request #784 from NVIDIA/feat/bump-langchain-version

Feat: Upgrade LangChain to Version 0.3

examples/configs/rag/custom_rag_output_rails/config.py

@@ -13,8 +13,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from langchain.llms.base import BaseLLM
 from langchain.prompts import PromptTemplate
+from langchain_core.language_models.llms import BaseLLM
 from langchain_core.output_parsers import StrOutputParser
 
 from nemoguardrails import LLMRails

examples/configs/rag/multi_kb/config.py

@@ -23,9 +23,9 @@
 from gpt4pandas import GPT4Pandas
 from langchain.chains import RetrievalQA
 from langchain.embeddings import HuggingFaceEmbeddings
-from langchain.llms import BaseLLM
 from langchain.text_splitter import CharacterTextSplitter
 from langchain.vectorstores import FAISS
+from langchain_core.language_models.llms import BaseLLM
 from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
 
 from nemoguardrails import LLMRails, RailsConfig

examples/configs/rag/pinecone/config.py

@@ -21,8 +21,8 @@
 from langchain.chains import RetrievalQA
 from langchain.docstore.document import Document
 from langchain.embeddings.openai import OpenAIEmbeddings
-from langchain.llms import BaseLLM
 from langchain.vectorstores import Pinecone
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import LLMRails
 from nemoguardrails.actions import action

examples/scripts/demo_llama_index_guardrails.py

@@ -15,7 +15,7 @@
 
 from typing import Any, Callable, Coroutine
 
-from langchain.llms.base import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import LLMRails, RailsConfig
 

nemoguardrails/actions/llm/generation.py

@@ -28,7 +28,7 @@
 from typing import Callable, List, Optional, cast
 
 from jinja2 import Environment, meta
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions.actions import ActionResult, action
 from nemoguardrails.actions.llm.utils import (

nemoguardrails/actions/v2_x/generation.py

@@ -21,7 +21,7 @@
 from ast import literal_eval
 from typing import Any, List, Optional, Tuple
 
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 from rich.text import Text
 
 from nemoguardrails.actions.actions import action

nemoguardrails/library/content_safety/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import Dict, Optional
 
-from langchain.llms.base import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions.actions import action
 from nemoguardrails.actions.llm.utils import llm_call

nemoguardrails/library/factchecking/align_score/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import Optional
 
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import RailsConfig
 from nemoguardrails.actions import action

nemoguardrails/library/hallucination/actions.py

@@ -17,8 +17,8 @@
 from typing import Optional
 
 from langchain.chains import LLMChain
-from langchain.llms.base import BaseLLM
 from langchain.prompts import PromptTemplate
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import RailsConfig
 from nemoguardrails.actions import action

nemoguardrails/library/llama_guard/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import List, Optional, Tuple
 
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions import action
 from nemoguardrails.actions.llm.utils import llm_call

nemoguardrails/library/patronusai/actions.py

@@ -17,7 +17,7 @@
 import re
 from typing import List, Optional, Tuple, Union
 
-from langchain.llms.base import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions import action
 from nemoguardrails.actions.llm.utils import llm_call

nemoguardrails/library/self_check/facts/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import Optional
 
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import RailsConfig
 from nemoguardrails.actions import action

nemoguardrails/library/self_check/input_check/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import Optional
 
-from langchain.llms.base import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import RailsConfig
 from nemoguardrails.actions.actions import ActionResult, action

nemoguardrails/library/self_check/output_check/actions.py

@@ -16,7 +16,7 @@
 import logging
 from typing import Optional
 
-from langchain.llms.base import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails import RailsConfig
 from nemoguardrails.actions import action

nemoguardrails/llm/helpers.py

@@ -19,7 +19,7 @@
     AsyncCallbackManagerForLLMRun,
     CallbackManagerForLLMRun,
 )
-from langchain.llms.base import LLM, BaseLLM
+from langchain_core.language_models.llms import LLM, BaseLLM
 
 
 def get_llm_instance_wrapper(

nemoguardrails/llm/providers/_langchain_nvidia_ai_endpoints_patch.py

@@ -21,8 +21,8 @@
 from langchain_core.language_models.chat_models import generate_from_stream
 from langchain_core.messages import BaseMessage
 from langchain_core.outputs import ChatResult
-from langchain_core.pydantic_v1 import Field
 from langchain_nvidia_ai_endpoints import ChatNVIDIA as ChatNVIDIAOriginal
+from pydantic.v1 import Field
 
 log = logging.getLogger(__name__)
 

nemoguardrails/llm/providers/nemollm.py

@@ -24,15 +24,15 @@
     AsyncCallbackManagerForLLMRun,
     CallbackManagerForLLMRun,
 )
-from langchain.llms.base import BaseLLM
-from langchain.pydantic_v1 import BaseModel, root_validator
 from langchain.schema import Generation
 from langchain.schema.output import GenerationChunk, LLMResult
+from langchain_core.language_models.llms import BaseLLM
+from pydantic.v1 import root_validator
 
 log = logging.getLogger(__name__)
 
 
-class NeMoLLM(BaseLLM, BaseModel):
+class NeMoLLM(BaseLLM):
     """Wrapper around NeMo LLM large language models.
 
     If NGC_API_HOST, NGC_API_KEY and NGC_ORGANIZATION_ID environment variables are set,

nemoguardrails/llm/providers/providers.py

@@ -23,30 +23,37 @@
 
 import asyncio
 import logging
-from importlib.metadata import PackageNotFoundError, version
+import warnings
+from importlib.metadata import version
 from typing import Any, Dict, List, Optional, Type
 
-from langchain.base_language import BaseLanguageModel
 from langchain.callbacks.manager import (
     AsyncCallbackManagerForLLMRun,
     CallbackManagerForLLMRun,
 )
-from langchain.llms.base import LLM
 from langchain.schema.output import GenerationChunk
 from langchain_community import llms
 from langchain_community.llms import HuggingFacePipeline
-from packaging import version as pkg_version
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.rails.llm.config import Model
 
 from .nemollm import NeMoLLM
 from .trtllm.llm import TRTLLM
 
+# NOTE: this is temp
+# Suppress specific warnings related to protected namespaces in Pydantic models, they must update their code.
+warnings.filterwarnings(
+    "ignore",
+    message=r'Field "model_.*" in .* has conflict with protected namespace "model_"',
+    category=UserWarning,
+    module=r"pydantic\._internal\._fields",
+)
 log = logging.getLogger(__name__)
 
-# Initialize the providers with the default ones, for now only NeMo LLM.
+# Initialize the providers with the default ones
 # We set nvidia_ai_endpoints provider to None because it's only supported if `langchain_nvidia_ai_endpoints` is installed.
-_providers: Dict[str, Type[BaseLanguageModel]] = {
+_providers: Dict[str, Optional[Type[BaseLLM]]] = {
     "nemollm": NeMoLLM,
     "trt_llm": TRTLLM,
     "nvidia_ai_endpoints": None,
@@ -193,7 +200,7 @@ def discover_langchain_providers():
         # If the "_acall" method is not defined, we add it.
         if (
             provider_cls
-            and issubclass(provider_cls, LLM)
+            and issubclass(provider_cls, BaseLLM)
             and "_acall" not in provider_cls.__dict__
         ):
             log.debug("Adding async support to %s", provider_cls.__name__)
@@ -204,12 +211,12 @@ def discover_langchain_providers():
 discover_langchain_providers()
 
 
-def register_llm_provider(name: str, provider_cls: Type[BaseLanguageModel]):
+def register_llm_provider(name: str, provider_cls: Type[BaseLLM]):
     """Register an additional LLM provider."""
     _providers[name] = provider_cls
 
 
-def get_llm_provider(model_config: Model) -> Type[BaseLanguageModel]:
+def get_llm_provider(model_config: Model) -> Type[BaseLLM]:
     if model_config.engine not in _providers:
         raise RuntimeError(f"Could not find LLM provider '{model_config.engine}'")
 

nemoguardrails/llm/providers/trtllm/llm.py

@@ -21,8 +21,8 @@
 from typing import Any, Dict, List, Optional
 
 from langchain.callbacks.manager import CallbackManagerForLLMRun
-from langchain.llms.base import LLM
-from langchain.pydantic_v1 import Field, root_validator
+from langchain_core.language_models.llms import BaseLLM
+from pydantic.v1 import Field, root_validator
 
 from nemoguardrails.llm.providers.trtllm.client import TritonClient
 
@@ -31,7 +31,7 @@
 RANDOM_SEED = 0
 
 
-class TRTLLM(LLM):
+class TRTLLM(BaseLLM):
     """A custom Langchain LLM class that integrates with TRTLLM triton models.
 
     Arguments:

nemoguardrails/rails/llm/llmrails.py

@@ -25,8 +25,8 @@
 import warnings
 from typing import Any, AsyncIterator, Dict, List, Optional, Tuple, Type, Union, cast
 
-from langchain.llms.base import BaseLLM
 from langchain_core.language_models import BaseLanguageModel
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions.llm.generation import LLMGenerationActions
 from nemoguardrails.actions.llm.utils import get_colang_history

pyproject.toml

@@ -42,9 +42,9 @@ dependencies = [
   "httpx>=0.24.1",
   "jinja2>=3.1.4",
   # The 0.1.9 has a bug related to SparkLLM which breaks everything.
-  "langchain>=0.2.14,<0.3.0,!=0.1.9",
-  "langchain-core>=0.2.14,<0.3.0,!=0.1.26",
-  "langchain-community>=0.0.16,<0.3.0",
+  "langchain>=0.2.14,<0.4.0,!=0.1.9",
+  "langchain-core>=0.2.14,<0.4.0,!=0.1.26",
+  "langchain-community>=0.0.16,<0.4.0",
   "lark~=1.1.7",
   "nest-asyncio>=1.5.6",
   "prompt-toolkit>=3.0",

tests/test_configs/with_custom_llm_prompt_action_v2_x/actions.py

@@ -16,7 +16,7 @@
 from ast import literal_eval
 from typing import Optional
 
-from langchain.llms import BaseLLM
+from langchain_core.language_models.llms import BaseLLM
 
 from nemoguardrails.actions import action
 from nemoguardrails.actions.llm.utils import llm_call

tests/utils.py

@@ -23,7 +23,7 @@
     AsyncCallbackManagerForLLMRun,
     CallbackManagerForLLMRun,
 )
-from langchain.llms.base import LLM
+from langchain_core.language_models.llms import LLM
 
 from nemoguardrails import LLMRails, RailsConfig
 from nemoguardrails.colang import parse_colang_file