Update Dockerfile to use pip cache for faster installation

Dockerfile

@@ -5,7 +5,7 @@ FROM python:3.12.0
 #     && rm -rf /var/lib/apt/lists/*
 COPY requirements.txt /app/requirements.txt
 WORKDIR /app
-RUN pip install -r requirements.txt
+RUN --mount=type=cache,target=/root/.cache/pip pip install -r requirements.txt
 RUN playwright install-deps 
 RUN useradd -m -u 1000 user
 USER user
@@ -27,4 +27,5 @@ RUN mv frpc_linux_amd64_v0.2 /usr/local/lib/python3.12/site-packages/gradio/
 RUN chmod +x /usr/local/lib/python3.12/site-packages/gradio/frpc_linux_amd64_v0.2
 EXPOSE 7860
 USER user
-CMD ["python", "main.py"]
+# CMD ["python", "main.py"]
+CMD ["gradio", "main.py"]

main.py

@@ -9,6 +9,7 @@ from utils import HuggingChat
 from langchain_core.prompts import PromptTemplate
 from langchain_community.embeddings import HuggingFaceEmbeddings
 import langchain
+from langchain_groq.chat_models import ChatGroq
 
 # import json
 # import shutil
@@ -68,7 +69,16 @@ Answer:
         self.db = Chroma(persist_directory="./pragetx_chroma", embedding_function=HuggingFaceEmbeddings(), collection_name="pragetx")  
         # self.llm = ChatOllama(model="phi3:3.8b", base_url="http://localhost:11434", num_gpu=16)
         # self.llm = HuggingChat(email = os.getenv("HF_EMAIL") , psw = os.getenv("HF_PASS") )
-        self.llm = HuggingChat(email = "brij1808" , psw = "Brijesh321@R" )
+        # self.llm = HuggingChat(email = "brij1808" , psw = "Brijesh321@R" )
+        os.environ["GROQ_API_KEY"] = os.getenv("GROQ_API_KEY")
+        self.llm = ChatGroq(
+                model="gemma2-9b-it",
+                temperature=1,
+                max_tokens=512,
+                timeout=5,
+                max_retries=2,
+            )
+
         self.chain = (
             {"chat_history": self.chat_history, "context": self.db.as_retriever(search_kwargs={"k":3}), "question": RunnablePassthrough()} |
             self.prompt |
@@ -87,6 +97,7 @@ Answer:
         print(history)
         prompt = history[-1][0] or ""
         for chunks in self.chain.stream(prompt):
+            print('chunks: ', chunks)
             history[-1][1] = history[-1][1] or ""
             history[-1][1] += chunks
             yield history

requirements.txt

@@ -42,6 +42,7 @@ googleapis-common-protos==1.63.0
 gradio==4.31.5
 gradio_client==0.16.4
 greenlet==3.0.3
+# groq==0.11.0
 grpcio==1.64.0
 h11==0.14.0
 httpcore==1.0.5
@@ -70,6 +71,7 @@ langchain==0.2.1
 langchain-chroma==0.1.1
 langchain-community==0.2.1
 langchain-core==0.2.1
+langchain-groq
 langchain-text-splitters==0.2.0
 langsmith==0.1.63
 lxml==5.2.2