Uploaded files

main.py

@@ -0,0 +1,40 @@
+"""Phi 2 Inference.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1qGlsJwf-rAF06cTMJU56iM9z3k_HZngS
+"""
+
+
+"""## Tokenizer and Model Prep"""
+
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+
+tokenizer = AutoTokenizer.from_pretrained(
+    "microsoft/phi-2",
+    trust_remote_code = True
+)
+
+model = AutoModelForCausalLM.from_pretrained(
+    "microsoft/phi-2",
+    torch_dtype = "auto",
+    device_map = "auto",
+    trust_remote_code = True
+)
+
+prompt = """Give me a list of 13 words that have 9 letters."""
+
+with torch.no_grad():
+  token_ids = tokenizer.encode(prompt, add_special_tokens=False ,return_tensors="pt")
+  output_ids = model.generate(
+      token_ids.to(model.device),
+      max_new_tokens=512,
+      do_sample=True,
+      temperature = 0.3
+  )
+
+output = tokenizer.decode(output_ids[0][token_ids.size(1) :])
+
+print(output)

readme.md

@@ -0,0 +1,56 @@
+# Microsoft Phi 2 Streamlit App
+
+Welcome to the Microsoft Phi 2 Streamlit App, a powerful and user-friendly application that harnesses the capabilities of the Microsoft Phi 2 language model for text generation. This app, built using Streamlit, allows users to interact with Phi 2 effortlessly and generate contextually rich text based on their prompts.
+
+## Getting Started
+
+1. **Clone the Repository:**
+   ```bash
+   git clone https://github.com/your-username/microsoft-phi2-streamlit-app.git
+   cd microsoft-phi2-streamlit-app
+   ```
+
+2. **Install Dependencies:**
+   ```bash
+   pip install -r requirements.txt
+   ```
+
+3. **Run the App:**
+   ```bash
+   streamlit run app.py
+   ```
+   This command will launch the app in your default web browser.
+
+## Usage
+
+1. Enter your text prompt in the provided text area.
+2. Click the "Generate Output" button to initiate text generation based on your prompt.
+3. Explore the generated output provided by the Microsoft Phi 2 language model.
+
+## Customize and Explore
+
+Feel free to explore and customize the app to suit your needs. You can experiment with different prompts, adjust parameters, or even integrate additional features. The codebase is open for further customization and serves as a starting point for various text generation applications.
+
+## Demonstration
+* Start screen
+
+![img.png](img.png)
+
+* The output
+
+![img_1.png](img_1.png)
+
+
+## Acknowledgments
+
+This app utilizes the Microsoft Phi 2 language model, and we appreciate the contributions from the Microsoft Research team in advancing language model capabilities. Special thanks to the Streamlit development team for providing an excellent framework for building interactive web applications.
+
+## License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+## Issues and Contributions
+
+If you encounter any issues or have suggestions for improvements, please feel free to open an issue or submit a pull request. Your contributions are highly valued!
+
+Enjoy generating text with Microsoft Phi 2!

requirements.txt

@@ -0,0 +1,6 @@
+transformers
+sentencepiece
+accelerate
+bitsandbytes
+einops
+streamlit

streamlit_app.py

@@ -0,0 +1,36 @@
+import streamlit as st
+from transformers import AutoTokenizer, AutoModelForCausalLM
+import torch
+
+# Load the Phi 2 model and tokenizer
+tokenizer = AutoTokenizer.from_pretrained(
+    "microsoft/phi-2",
+    trust_remote_code=True
+)
+
+model = AutoModelForCausalLM.from_pretrained(
+    "microsoft/phi-2",
+    device_map="auto",
+    trust_remote_code=True
+)
+
+# Streamlit UI
+st.title("Microsoft Phi 2 Streamlit App")
+
+# User input prompt
+prompt = st.text_area("Enter your prompt:", """Write a story about Nasa""")
+
+# Generate output based on user input
+if st.button("Generate Output"):
+    with torch.no_grad():
+        token_ids = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt")
+        output_ids = model.generate(
+            token_ids.to(model.device),
+            max_new_tokens=512,
+            do_sample=True,
+            temperature=0.3
+        )
+
+    output = tokenizer.decode(output_ids[0][token_ids.size(1):])
+    st.text("Generated Output:")
+    st.write(output)