Initial commit: DeepSeek LLM chat application

README.md

@@ -1,13 +1,20 @@
 ---
-title: Test
-emoji: 📈
-colorFrom: purple
-colorTo: purple
+title: DeepSeek Coding Assistant
+emoji: 🤖
+colorFrom: blue
+colorTo: green
 sdk: streamlit
 sdk_version: 1.41.1
 app_file: app.py
 pinned: false
-short_description: this for testing
+short_description: Code-focused chat interface using DeepSeek 6.7B
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# DeepSeek Coding Assistant
+An AI coding assistant powered by DeepSeek 6.7B model.
+
+## Features
+- Specialized in coding and technical discussions
+- Based on DeepSeek 6.7B instruct model
+- Real-time response generation
+- Context-aware conversations

app.py

@@ -1,4 +1,52 @@
 import streamlit as st
+from transformers import AutoModelForCausalLM, AutoTokenizer
+import torch
 
-x = st.slider('Select a value')
-st.write(x, 'squared is', x * x)
+@st.cache_resource
+def load_model():
+    model_name = "deepseek-ai/deepseek-coder-6.7b-instruct"
+    tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
+    model = AutoModelForCausalLM.from_pretrained(
+        model_name,
+        torch_dtype=torch.float16,
+        trust_remote_code=True,
+        device_map="auto"
+    )
+    return model, tokenizer
+
+def generate_response(prompt, model, tokenizer):
+    formatted_prompt = f"Human: {prompt}\n\nAssistant:"
+    inputs = tokenizer(formatted_prompt, return_tensors="pt").to(model.device)
+    outputs = model.generate(
+        **inputs,
+        max_length=2048,
+        temperature=0.7,
+        num_return_sequences=1,
+        pad_token_id=tokenizer.eos_token_id
+    )
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response.split("Assistant:")[-1].strip()
+
+st.title("🤖 DeepSeek Chat")
+st.write("Using DeepSeek 6.7B - a powerful open source coding LLM")
+
+if 'messages' not in st.session_state:
+    st.session_state.messages = []
+
+for message in st.session_state.messages:
+    with st.chat_message(message["role"]):
+        st.markdown(message["content"])
+
+# Load model
+model, tokenizer = load_model()
+
+if prompt := st.chat_input("Ask me anything about coding!"):
+    st.session_state.messages.append({"role": "user", "content": prompt})
+    with st.chat_message("user"):
+        st.markdown(prompt)
+
+    with st.chat_message("assistant"):
+        with st.spinner("Thinking..."):
+            response = generate_response(prompt, model, tokenizer)
+            st.markdown(response)
+            st.session_state.messages.append({"role": "assistant", "content": response})

requirements.txt

@@ -0,0 +1,8 @@
+# Add any additional dependencies here
+# streamlit is already pre-installed
+pandas
+numpy
+transformers>=4.34.0
+torch>=2.0.0
+accelerate>=0.20.0
+einops