Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -9,6 +9,7 @@ from langchain.chat_models import ChatOpenAI
|
|
| 9 |
from langchain.tools import DuckDuckGoSearchRun
|
| 10 |
from langchain_community.llms import HuggingFaceHub
|
| 11 |
from langchain_huggingface import HuggingFaceEndpoint
|
|
|
|
| 12 |
|
| 13 |
# (Keep Constants as is)
|
| 14 |
# --- Constants ---
|
|
@@ -31,8 +32,16 @@ class BasicAgent: # Some times Inheritance is needed
|
|
| 31 |
|
| 32 |
# # Create the LLM # Temprature set to 0 because we need exact match
|
| 33 |
# llm = ChatOpenAI(temperature=0, model="gpt-3.5-turbo") # ChatML understands roles (user, assistant, system)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
|
| 35 |
-
|
|
|
|
| 36 |
repo_id="tiiuae/falcon-7b-instruct",
|
| 37 |
huggingfacehub_api_token=os.environ["HUGGINGFACEHUB_API_TOKEN"],
|
| 38 |
model_kwargs={"temperature": 0.5, "max_new_tokens": 512}
|
|
@@ -66,7 +75,8 @@ class BasicAgent: # Some times Inheritance is needed
|
|
| 66 |
def __call__(self, question: str) -> str:
|
| 67 |
print(f"Agent received question (first 50 chars): {question[:50]}...")
|
| 68 |
# Do not return intermidiate steps or thoughts
|
| 69 |
-
response = self.agent.run(question)
|
|
|
|
| 70 |
print(f"Agent response: {response}")
|
| 71 |
return response
|
| 72 |
|
|
|
|
| 9 |
from langchain.tools import DuckDuckGoSearchRun
|
| 10 |
from langchain_community.llms import HuggingFaceHub
|
| 11 |
from langchain_huggingface import HuggingFaceEndpoint
|
| 12 |
+
from langchain_community.chat_models import ChatHuggingFace
|
| 13 |
|
| 14 |
# (Keep Constants as is)
|
| 15 |
# --- Constants ---
|
|
|
|
| 32 |
|
| 33 |
# # Create the LLM # Temprature set to 0 because we need exact match
|
| 34 |
# llm = ChatOpenAI(temperature=0, model="gpt-3.5-turbo") # ChatML understands roles (user, assistant, system)
|
| 35 |
+
|
| 36 |
+
# # Define the LLM
|
| 37 |
+
# llm = HuggingFaceHub(
|
| 38 |
+
# repo_id="tiiuae/falcon-7b-instruct",
|
| 39 |
+
# huggingfacehub_api_token=os.environ["HUGGINGFACEHUB_API_TOKEN"],
|
| 40 |
+
# model_kwargs={"temperature": 0.5, "max_new_tokens": 512}
|
| 41 |
+
# )
|
| 42 |
|
| 43 |
+
# Define the LLM
|
| 44 |
+
llm = ChatHuggingFace(
|
| 45 |
repo_id="tiiuae/falcon-7b-instruct",
|
| 46 |
huggingfacehub_api_token=os.environ["HUGGINGFACEHUB_API_TOKEN"],
|
| 47 |
model_kwargs={"temperature": 0.5, "max_new_tokens": 512}
|
|
|
|
| 75 |
def __call__(self, question: str) -> str:
|
| 76 |
print(f"Agent received question (first 50 chars): {question[:50]}...")
|
| 77 |
# Do not return intermidiate steps or thoughts
|
| 78 |
+
# response = self.agent.run(question)
|
| 79 |
+
response = self.agent.invoke({"input": question})
|
| 80 |
print(f"Agent response: {response}")
|
| 81 |
return response
|
| 82 |
|