Update app.py

app.py

@@ -15,12 +15,11 @@ from langchain_community.embeddings import HuggingFaceEmbeddings
 #from transformers import pipeline
 # Load model directly
 #from transformers import AutoModelForCausalLM
-from getpass import getpass
+from langchain_community.llms.huggingface_pipeline import HuggingFacePipeline
+
 
-HUGGINGFACEHUB_API_TOKEN = getpass()
 
 
-os.environ["HUGGINGFACEHUB_API_TOKEN"] = HUGGINGFACEHUB_API_TOKEN
 #access_token = os.getenv("HUGGINGFACE_API_KEY")
 
 
@@ -92,12 +91,16 @@ def get_conversational_chain(retriever):
     #repo_id='meta-llama/Meta-Llama-3-70B'
     #repo_id = 'mistralai/Mixtral-8x7B-Instruct-v0.1'
     #repo_id= 'nvidia/Llama3-ChatQA-1.5-8B'
-    repo_id= 'google/gemma-1.1-2b-it'
-    llm = HuggingFaceEndpoint(repo_id=repo_id, temperature=0.3,token = HUGGINGFACEHUB_API_TOKEN)
+    #repo_id= 'google/gemma-1.1-2b-it'
+    #llm = HuggingFaceEndpoint(repo_id=repo_id, temperature=0.3,token = HUGGINGFACEHUB_API_TOKEN)
     #tokenizer = AutoTokenizer.from_pretrained("google/gemma-1.1-2b-it")
     #llm = AutoModelForCausalLM.from_pretrained("google/gemma-1.1-2b-it")
     #llm = AutoModelForCausalLM.from_pretrained("microsoft/Phi-3-mini-128k-instruct", trust_remote_code=True, token=access_token)
     #llm = pipeline("text-generation", model="google/gemma-1.1-2b-it")
+    llm = HuggingFacePipeline.from_model_id(
+    model_id="Phi-3-mini-128k-instruct",
+    task="text-generation",
+    pipeline_kwargs={"max_new_tokens": 10})
 
     pt = ChatPromptTemplate.from_template(prompt_template)
     # Retrieve and generate using the relevant snippets of the blog.