ODSC-58449: ads LangChain plugin update (#877)

Author: mingkang111

ads/llm/langchain/plugins/base.py

@@ -8,6 +8,7 @@
 from langchain.llms.base import LLM
 from langchain.pydantic_v1 import BaseModel, Field, root_validator
 
+from ads import logger
 from ads.common.auth import default_signer
 from ads.config import COMPARTMENT_OCID
 
@@ -95,6 +96,11 @@ def validate_environment(  # pylint: disable=no-self-argument
         """Validate that python package exists in environment."""
         # Initialize client only if user does not pass in client.
         # Users may choose to initialize the OCI client by themselves and pass it into this model.
+        logger.warning(
+            f"The ads langchain plugin {cls.__name__} will be deprecated soon. "
+            "Please refer to https://python.langchain.com/v0.2/docs/integrations/providers/oci/ "
+            "for the latest support."
+        )
         if not values.get("client"):
             auth = values.get("auth", {})
             client_kwargs = auth.get("client_kwargs") or {}

pyproject.toml

@@ -175,7 +175,7 @@ pii = [
   "spacy==3.6.1",
   "report-creator==1.0.9",
 ]
-llm = ["langchain-community<0.0.32", "langchain>=0.1.10,<0.1.14", "evaluate>=0.4.0", "langchain-core<0.1.51"]
+llm = ["langchain-community<0.0.32", "langchain>=0.1.10,<0.1.14", "evaluate>=0.4.0"]
 aqua = ["jupyter_server"]
 
 # To reduce backtracking (decrese deps install time) during test/dev env setup reducing number of versions pip is

tests/unitary/with_extras/langchain/test_serialization.py

@@ -7,20 +7,25 @@
 
 import os
 from copy import deepcopy
-from unittest import TestCase, mock, SkipTest
+from unittest import SkipTest, TestCase, mock, skipIf
 
-from langchain.llms import Cohere
+import langchain_core
 from langchain.chains import LLMChain
+from langchain.llms import Cohere
 from langchain.prompts import PromptTemplate
 from langchain.schema.runnable import RunnableParallel, RunnablePassthrough
 
-from ads.llm.serialize import load, dump
 from ads.llm import (
     GenerativeAI,
     GenerativeAIEmbeddings,
     ModelDeploymentTGI,
     ModelDeploymentVLLM,
 )
+from ads.llm.serialize import dump, load
+
+
+def version_tuple(version):
+    return tuple(map(int, version.split(".")))
 
 
 class ChainSerializationTest(TestCase):
@@ -142,6 +147,10 @@ def test_llm_chain_serialization_with_oci(self):
         self.assertEqual(llm_chain.llm.model, "my_model")
         self.assertEqual(llm_chain.input_keys, ["subject"])
 
+    @skipIf(
+        version_tuple(langchain_core.__version__) > (0, 1, 50),
+        "Serialization not supported in this langchain_core version",
+    )
     def test_oci_gen_ai_serialization(self):
         """Tests serialization of OCI Gen AI LLM."""
         try:
@@ -157,6 +166,10 @@ def test_oci_gen_ai_serialization(self):
         self.assertEqual(llm.compartment_id, self.COMPARTMENT_ID)
         self.assertEqual(llm.client_kwargs, self.GEN_AI_KWARGS)
 
+    @skipIf(
+        version_tuple(langchain_core.__version__) > (0, 1, 50),
+        "Serialization not supported in this langchain_core version",
+    )
     def test_gen_ai_embeddings_serialization(self):
         """Tests serialization of OCI Gen AI embeddings."""
         try:
@@ -201,10 +214,27 @@ def test_runnable_sequence_serialization(self):
         element_3 = kwargs.get("last")
         self.assertNotIn("_type", element_3)
         self.assertEqual(element_3.get("id"), ["ads", "llm", "ModelDeploymentTGI"])
-        self.assertEqual(
-            element_3.get("kwargs"),
-            {"endpoint": "https://modeldeployment.customer-oci.com/ocid/predict"},
-        )
+
+        if version_tuple(langchain_core.__version__) > (0, 1, 50):
+            self.assertEqual(
+                element_3.get("kwargs"),
+                {
+                    "max_tokens": 256,
+                    "temperature": 0.2,
+                    "p": 0.75,
+                    "endpoint": "https://modeldeployment.customer-oci.com/ocid/predict",
+                    "best_of": 1,
+                    "do_sample": True,
+                    "watermark": True,
+                },
+            )
+        else:
+            self.assertEqual(
+                element_3.get("kwargs"),
+                {
+                    "endpoint": "https://modeldeployment.customer-oci.com/ocid/predict",
+                },
+            )
 
         chain = load(serialized)
         self.assertEqual(len(chain.steps), 3)