add three new ‘Slangify’ chat styles and remove the previous TA’s functions.

.github/workflows/check.yml

@@ -13,4 +13,4 @@ jobs:
       - name: Check large files
         uses: ActionsDesk/lfs-warning@v2.0
         with:
-          filesizelimit: 10485760 # this is 10MB so we can sync to HF Spaces
+          filesizelimit: 10485760 # this is 10MB so we can sync to HF Spaces

.github/workflows/{sync.yml → main.yml}

@@ -1,10 +1,7 @@
 name: Sync to Hugging Face hub
 on:
-  workflow_run:
-    workflows: ["Run Pytest Tests"]  # This must match the name of your test workflow
-    types:
-      - completed
-
+  push:
+    branches: [main]
 
   # to run this workflow manually from the Actions tab
   workflow_dispatch:
@@ -20,4 +17,4 @@ jobs:
       - name: Push to hub
         env:
           HF_TOKEN: ${{ secrets.HF_TOKEN }}
-        run: git push -f https://oxmraz-mldo:$HF_TOKEN@huggingface.co/spaces/Group17WPIMLDO24/Case-Study-1 main
+        run: git push https://oxmraz-mldo:$HF_TOKEN@huggingface.co/spaces/Group17WPIMLDO24/Case-Study-1 main

.github/workflows/setupaccess.yml

@@ -1,25 +0,0 @@
-name: Setup Access
-
-on:
-  workflow_dispatch:  # Manual trigger
-
-jobs:
-  deploy:
-    runs-on: ubuntu-latest 
-
-    steps:
-      - name: Checkout Repository
-        uses: actions/checkout@v3
-
-      - name: Install expect
-        run: sudo apt-get install expect
-
-      - name: Add permission for script to run
-        run: chmod +x setupaccess.sh
-      
-      - name: Run Bash Scripts
-        env:
-          PASSPHRASE_GROUP17: ${{ secrets.PASSPHRASE_GROUP17 }}
-          GROUP17_PUBLICKKEY: ${{ secrets.GROUP17_PUBLICKKEY }}
-          GROUP17_PRIVATEKEY: ${{ secrets.GROUP17_PRIVATEKEY }}
-        run: expect setupaccess.exp "$PASSPHRASE_GROUP17" # this one has code triggering setupaccess.sh, so command ultimately runs multiple scripts :)

.github/workflows/test.yml

@@ -1,45 +0,0 @@
-# ----ATTRIBUTION-START----
-# LLM: Github Copilot
-# PROMPT: i have written  tests. i run them like this pytest test_blip_image_caption_large.py test_phi3_mini_4k_instruct.py test_musicgen_small.py - help me create a github runner that runs these tests - it also needs to create the environment variable "HF_API_TOKEN". it is added to the github repo under the name HF_API_TOKEN
-# EDITS: /
-
-name: Run Pytest Tests
-
-# Triggers the workflow on push or pull request to the main branch
-on:
-  push:
-    branches:
-      - main
-  pull_request:
-    branches:
-      - main
-  workflow_dispatch:  # Manual trigger
-
-
-jobs:
-  test:
-    runs-on: ubuntu-latest
-
-    env:
-      # Create the HF_API_TOKEN environment variable from the repository secrets
-      HF_API_TOKEN: ${{ secrets.HF_API_TOKEN }}
-
-    steps:
-    - name: Checkout code
-      uses: actions/checkout@v3
-      
-    - name: Set up Python
-      uses: actions/setup-python@v4
-      with:
-        python-version: "3.x"  # Set your preferred Python version here
-
-    - name: Install dependencies
-      run: |
-        python -m pip install --upgrade pip
-        pip install -r requirements.txt  # Ensure you have a requirements.txt in your repo
-
-    - name: Run Pytest tests
-      run: |
-        pytest test_blip_image_caption_large.py test_phi3_mini_4k_instruct.py test_musicgen_small.py
-
-# -----ATTRIBUTION-END-----

.gitignore

@@ -1,5 +0,0 @@
-/__pycache__
-*.wav
-.pytest_cache
-/audio_data
-.cache

README.md

@@ -1,19 +1,12 @@
 ---
-title: Case-Study-1 - Image-To-Music
-emoji: 🎼
-colorFrom: gray
-colorTo: blue
+title: CS553_Example
+emoji: 💬
+colorFrom: yellow
+colorTo: purple
 sdk: gradio
-sdk_version: 4.44.0
+sdk_version: 4.36.1
 app_file: app.py
 pinned: false
 ---
 
-## Case-Study-1: Image-To-Music 🎼
-
-An image to music converter, built with the following models:
-- https://huggingface.co/Salesforce/blip-image-captioning-large for Image Captioning
-- https://huggingface.co/microsoft/Phi-3-mini-4k-instruct       for Audio Prompt generation with Caption
-- https://huggingface.co/facebook/musicgen-small                for Music Generation
-
-Currently supports .jpg, .jpeg, and .png!
+An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).

app.py

@@ -1,164 +1,158 @@
-# external imports
-import gc
-import logging as log
-import time
-import uuid
 import gradio as gr
-
-# local imports
-from blip_image_caption_large import Blip_Image_Caption_Large
-from phi3_mini_4k_instruct import Phi3_Mini_4k_Instruct
-from musicgen_small import Musicgen_Small
-import config
-
-log.basicConfig(level=log.INFO)
-
-
-class Image_To_Music:
-    def __init__(self, use_local_caption=False, use_local_llm=False, use_local_musicgen=False):
-
-        self.use_local_llm = use_local_llm
-        self.use_local_caption = use_local_caption
-        self.use_local_musicgen = use_local_musicgen
-
-        self.image_path = None
-        self.generated_caption = None
-        self.generated_description = None
-        self.audio_path = config.AUDIO_DIR + str(uuid.uuid4()) + ".wav"
-
-        self.caption_generation_duration = -1
-        self.description_generation_duration = -1
-        self.music_generation_duration = -1
-
-    def caption_image(self, image_path):
-        log.info("Captioning Image...")
-        caption_start_time = time.time()
-
-        # load model
-        self.image_caption_model = Blip_Image_Caption_Large()
-
-        self.image_path = image_path
-        self.generated_caption = self.image_caption_model.caption_image(self.image_path, self.use_local_caption)
-
-        # delete model to free up ram
-        del self.image_caption_model
-        gc.collect()
-
-        self.caption_generation_duration = time.time() - caption_start_time
-        log.info(f"Captioning Complete in {self.caption_generation_duration:.2f} seconds: {self.generated_caption} - used local model: {self.use_local_caption}")
-        return self.generated_caption
-    
-    def generate_description(self):
-        log.info("Generating Music Description...")
-        description_start_time = time.time()
-
-        # load model
-        self.text_generation_model = Phi3_Mini_4k_Instruct()
-
-        messages = [
-            {"role": "system", "content": "You are an image caption to song description converter with a deep understanding of Music and Art. You are given the caption of an image. Your task is to generate a textual description of a musical piece that fits the caption. The description should be detailed and vivid, and should include the genre, mood, instruments, tempo, and other relevant information about the music. You should also use your knowledge of art and visual aesthetics to create a musical piece that complements the image. Only output the description of the music, without any explanation or introduction. Be concise."},
-            {"role": "user", "content": self.generated_caption},
-        ]
-        self.generated_description = self.text_generation_model.generate_text(messages, self.use_local_llm)
-
-        # delete model to free up ram
-        del self.text_generation_model
-        gc.collect()
-
-        self.description_generation_duration = time.time() - description_start_time
-        log.info(f"Description Generation Complete in {self.description_generation_duration:.2f} seconds: {self.generated_description} - used local model: {self.use_local_llm}")
-        return self.generated_description
-    
-    def generate_music(self):
-        log.info("Generating Music...")
-        music_start_time = time.time()
-        
-        # load model
-        self.music_generation_model = Musicgen_Small()
-
-        self.music_generation_model.generate_music(self.generated_description, self.audio_path, self.use_local_musicgen)
-        
-        # delete model to free up ram
-        del self.music_generation_model
-        gc.collect()
-
-        self.music_generation_duration = time.time() - music_start_time
-        log.info(f"Music Generation Complete in {self.music_generation_duration:.2f} seconds: {self.audio_path} - used local model: {self.use_local_musicgen}")
-        return self.audio_path
-    
-    def get_durations(self):
-            return f"Caption Generation Time: {self.caption_generation_duration:.2f} seconds\nDescription Generation Time: {self.description_generation_duration:.2f} seconds\nMusic Generation Time: {self.music_generation_duration:.2f} seconds\nTotal Time: {self.caption_generation_duration + self.description_generation_duration + self.music_generation_duration:.2f} seconds"
-
-    def run_yield(self, image_path):
-
-        self.caption_image(image_path)
-        yield [self.generated_caption, None, None, None]
-        self.generate_description()
-        yield [self.generated_caption, self.generated_description, None, None]
-        self.generate_music()
-        yield [self.generated_caption, self.generated_description, self.audio_path, None]
-        return [self.generated_caption, self.generated_description, self.audio_path,self.get_durations()]
-    
-    def run(self, image_path):
-        self.caption_image(image_path)
-        self.generate_description()
-        self.generate_music()
-        return [self.generated_caption, self.generated_description, self.audio_path, self.get_durations()]
-
-
-def run_image_to_music(image_path, llm_max_new_tokens, llm_temperature, llm_top_p, musicgen_max_seconds, use_local_caption, use_local_llm, use_local_musicgen):
-    config.LLM_MAX_NEW_TOKENS = llm_max_new_tokens
-    config.LLM_TEMPERATURE = llm_temperature
-    config.LLM_TOP_P = llm_top_p
-    config.MUSICGEN_MAX_NEW_TOKENS = musicgen_max_seconds * 51
-    itm = Image_To_Music(use_local_caption=use_local_caption, use_local_llm=use_local_llm, use_local_musicgen=use_local_musicgen)
-    return itm.run(image_path)
-
-# Gradio UI 
-def gradio():
-    # Define Gradio Interface, information from (https://www.gradio.app/docs/chatinterface)
-    with gr.Blocks() as demo:
-        gr.Markdown("<h1 style='text-align: center;'> ⛺ Image to Music Generator 🎼</h1>")
-        image_input = gr.Image(type="filepath", label="Upload Image")
-
-
-        # ----ATTRIBUTION-START----
-        # LLM: ChatGPT4o
-        # PROMPT: i need 3 checkbox fields that pass booleans to the run_image_to_music function. it should be  "Use local Image Captioning" "Use local LLM" "Use local Music Generation". please make it a nice parameter selector
-        # EDITS: /
-
-        # Checkbox parameters
-        with gr.Row():
-            local_captioning = gr.Checkbox(label="Use local Image Captioning", value=False)
-            local_llm = gr.Checkbox(label="Use local LLM", value=False)
-            local_music_gen = gr.Checkbox(label="Use local Music Generation", value=False)
-        # -----ATTRIBUTION-END-----
-
-        # ----ATTRIBUTION-START----
-        # LLM: ChatGPT4o
-        # PROMPT: Now, I need sliders for the different models that are used in the product:\n LLM_MAX_NEW_TOKENS = 50\nLLM_TEMPERATURE = 0.7\nLLM_TOP_P = 0.95\nMUSICGEN_MAX_NEW_TOKENS = 256 # 256 =  5 seconds of audio\n they should be in a hidden menu that opens when I click on "advanced options"\nPlease label them for the end user and fit them nicely in the following UI: <code>
-        # EDITS: added interactive flags
-        # Advanced options with sliders
-        with gr.Accordion("Advanced Options", open=False):
-            gr.Markdown("<h3>LLM Settings</h3>")
-            llm_max_new_tokens = gr.Slider(1, 200, value=50, step=1, label="LLM Max Tokens", interactive=True)
-            llm_temperature = gr.Slider(0.0, 1.0, value=0.7, step=0.01, label="LLM Temperature", interactive=True)
-            llm_top_p = gr.Slider(0.01, 0.99, value=0.95, step=0.01, label="LLM Top P", interactive=True)
-
-            gr.Markdown("<h3>Music Generation Settings</h3>")
-            musicgen_max_seconds = gr.Slider(1, 30, value=5, step=1, label="MusicGen Duration in Seconds (local model only)", interactive=True)
-        # -----ATTRIBUTION-END-----
-
-        with gr.Row():
-            caption_output = gr.Textbox(label="Image Caption")
-            music_description_output = gr.Textbox(label="Music Description")
-            durations = gr.Textbox(label="Processing Times", interactive=False, placeholder="Time statistics will appear here")
-
-        music_output = gr.Audio(label="Generated Music")
-        # Button to trigger the process
-        generate_button = gr.Button("Generate Music")
-        generate_button.click(fn=run_image_to_music, inputs=[image_input, llm_max_new_tokens, llm_temperature, llm_top_p, musicgen_max_seconds, local_captioning, local_llm, local_music_gen], outputs=[caption_output, music_description_output, music_output, durations])
-    # Launch Gradio app
-    demo.launch(server_port=config.SERVICE_PORT, server_name=config.SERVER_NAME)
-
-gradio()
+from huggingface_hub import InferenceClient
+import torch
+from transformers import pipeline
+
+# Inference client setup
+client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
+pipe = pipeline("text-generation", "microsoft/Phi-3-mini-4k-instruct", torch_dtype=torch.bfloat16, device_map="auto")
+
+# Global flag to handle cancellation
+stop_inference = False
+
+def style_response(style, response):
+    """Modify response style based on the selected style."""
+    if style == "Nautical Marauder":
+        response = response.replace("you", "ye").replace("hello", "ahoy").replace("friend", "matey")
+        response = response.replace("is", "be").replace("my", "me").replace("the", "th'").replace("am", "be")
+    elif style == "Elizabethan Prose":
+        response = response.replace("you", "thou").replace("are", "art").replace("is", "be").replace("my", "mine")
+        response = response.replace("your", "thy").replace("the", "thee").replace("has", "hath").replace("do", "doth")
+    elif style == "Cyber Elite":
+        response = response.replace("e", "3").replace("a", "4").replace("t", "7").replace("o", "0").replace("i", "1")
+    return response
+
+def get_css(style):
+    """Return corresponding CSS based on the selected style."""
+    if style == "Nautical Marauder":
+        return """
+        body {
+            background-color: #2b2b2b;
+            font-family: 'Trebuchet MS', sans-serif;
+            color: #f4e9c9;
+            background-image: url('https://www.transparenttextures.com/patterns/old-map.png');
+        }
+        .gradio-container {
+            background: rgba(0, 0, 0, 0.7);
+            border: 2px solid #d4af37;
+            box-shadow: 0 4px 8px rgba(255, 255, 255, 0.1);
+        }
+        .gr-chat {
+            font-size: 16px;
+            color: #f4e9c9;
+        }
+        """
+    elif style == "Elizabethan Prose":
+        return """
+        body {
+            background-color: #f5f0e1;
+            font-family: 'Dancing Script', cursive;
+            color: #5c4033;
+            background-image: url('https://www.transparenttextures.com/patterns/old-paper.png');
+        }
+        .gradio-container {
+            background: rgba(255, 255, 255, 0.9);
+            border: 2px solid #a0522d;
+            box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
+        }
+        .gr-chat {
+            font-size: 18px;
+            color: #5c4033;
+        }
+        """
+    elif style == "Cyber Elite":
+        return """
+        body {
+            background-color: #000000;
+            font-family: 'Courier New', Courier, monospace;
+            color: #00ff00;
+        }
+        .gradio-container {
+            background: #1a1a1a;
+            border: 2px solid #00ff00;
+            box-shadow: 0 4px 8px rgba(0, 255, 0, 0.3);
+        }
+        .gr-chat {
+            font-size: 16px;
+            color: #00ff00;
+        }
+        """
+    else:
+        # Default style
+        return """
+        body {
+            background-color: #f0f0f0;
+            font-family: 'Arial', sans-serif;
+            color: #333;
+        }
+        .gradio-container {
+            background: white;
+            box-shadow: 0 4px 8px rgba(0, 0, 0, 0.1);
+            border-radius: 10px;
+        }
+        .gr-chat {
+            font-size: 16px;
+            color: #333;
+        }
+        """
+
+def respond(message, history: list[tuple[str, str]], style="Standard Conversational"):
+    global stop_inference
+    stop_inference = False  # Reset cancellation flag
+
+    # Initialize history if it's None
+    if history is None:
+        history = []
+
+    # API-based inference
+    messages = [{"role": "user", "content": message}]
+
+    response = ""
+    for message_chunk in client.chat_completion(
+        messages,
+        max_tokens=512,  # Default max tokens for response
+        stream=True,
+        temperature=0.7,  # Default temperature
+        top_p=0.95,  # Default top-p
+    ):
+        if stop_inference:
+            response = "Inference cancelled."
+            yield history + [(message, response)]
+            return
+        token = message_chunk.choices[0].delta.content
+        response += token
+        yield history + [(message, style_response(style, response))]  # Apply selected style to the response
+
+
+def cancel_inference():
+    global stop_inference
+    stop_inference = True
+
+# Define the interface
+with gr.Blocks() as demo:
+    gr.Markdown("<h1 style='text-align: center;'>🌟 Fancy AI Chatbot 🌟</h1>")
+    gr.Markdown("Interact with the AI chatbot using customizable settings below.")
+
+    chat_history = gr.Chatbot(label="Chat")
+
+    user_input = gr.Textbox(show_label=False, placeholder="Type your message here...")
+
+    cancel_button = gr.Button("Cancel Inference", variant="danger")
+
+    # New feature: Style selection with more formal names
+    style_selection = gr.Dropdown(
+        label="Response Style", 
+        choices=["Standard Conversational", "Nautical Marauder", "Elizabethan Prose", "Cyber Elite"], 
+        value="Standard Conversational"
+    )
+
+    def apply_css(style):
+        return get_css(style)
+
+    # Adjusted to ensure history is maintained and passed correctly
+    user_input.submit(respond, [user_input, chat_history, style_selection], chat_history)
+    style_selection.change(apply_css, style_selection, gr.CSS())
+    cancel_button.click(cancel_inference)
+
+if __name__ == "__main__":
+    demo.launch(share=False)  # Remove share=True because it's not supported on HF Spaces

attribution_example.py

@@ -1,8 +0,0 @@
-# Example Code Attribution for AI-Generated Code
-
-# ----ATTRIBUTION-START----
-# LLM: Github Copilot
-# PROMPT: write a hello world example
-# EDITS: changed the wording to make it more personal
-print("Hello, World! This is your Copilot speaking!")
-# -----ATTRIBUTION-END-----

blip_image_caption_large.py

@@ -1,29 +0,0 @@
-# external imports
-from transformers import pipeline
-from huggingface_hub import InferenceClient
-
-# local imports
-import config
-
-class Blip_Image_Caption_Large:
-    def __init__(self):
-        pass
-
-    def caption_image(self, image_path, use_local_caption):
-        if use_local_caption:
-            return self.caption_image_local_pipeline(image_path)
-        else:
-            return self.caption_image_api(image_path)
-    
-    def caption_image_local_pipeline(self, image_path):
-        self.local_pipeline = pipeline("image-to-text", model=config.IMAGE_CAPTION_MODEL)
-        result = self.local_pipeline(image_path)[0]['generated_text']
-        return result
-
-    def caption_image_api(self, image_path):
-        client = InferenceClient(config.IMAGE_CAPTION_MODEL, token=config.HF_API_TOKEN)
-        try:
-            result = client.image_to_text(image_path).generated_text
-        except Exception as e:
-            result = f"Error: {e}"
-        return result

config.py

@@ -1,38 +0,0 @@
-import os
-import logging as log
-
-log.basicConfig(level=log.INFO)
-
-SERVICE_PORT = 7860
-SERVER_NAME = "0.0.0.0"
-
-IMAGE_CAPTION_MODEL = "Salesforce/blip-image-captioning-large"
-
-LLM_MODEL = "microsoft/Phi-3-mini-4k-instruct"
-LLM_MAX_LENGTH = 50
-LLM_MAX_NEW_TOKENS = 50
-LLM_TEMPERATURE = 0.7
-LLM_TOP_P = 0.95
-
-MUSICGEN_MODEL = "facebook/musicgen-small"
-MUSICGEN_MODEL_API_URL = f"https://api-inference.huggingface.co/models/{MUSICGEN_MODEL}"
-MUSICGEN_MAX_NEW_TOKENS = 256 # 5 seconds of audio
-
-AUDIO_DIR = "Case-Study-1/data/"
-
-HF_API_TOKEN = os.getenv("HF_API_TOKEN")
-
-if HF_API_TOKEN:
-    log.info(f"Read HF_API_TOKEN: {HF_API_TOKEN[0:4]}...")
-else:
-    print("HF_API_TOKEN not found in environment variables.")
-
-# ----ATTRIBUTION-START----
-# LLM: Github Copilot
-# PROMPT: create an output folder for the generated audio files
-# EDITS: /
-def create_output_folder():
-    os.makedirs(AUDIO_DIR, exist_ok=True)
-# -----ATTRIBUTION-END-----
-
-create_output_folder()

musicgen_small.py

@@ -1,48 +0,0 @@
-# external imports
-from transformers import pipeline
-from io import BytesIO
-import requests
-import scipy
-
-# local imports
-import config
-
-class Musicgen_Small:
-    def __init__(self):
-        pass
-
-    def generate_music(self, prompt, audio_path, use_local_musicgen):
-        if use_local_musicgen:
-            self.generate_music_local_pipeline(prompt, audio_path)
-        else:
-            self.generate_music_api(prompt, audio_path)
-    
-    def generate_music_local_pipeline(self, prompt, audio_path):
-        self.local_pipeline = pipeline("text-to-audio", model=config.MUSICGEN_MODEL)
-        music = self.local_pipeline(prompt, forward_params={"do_sample": True, "max_new_tokens": config.MUSICGEN_MAX_NEW_TOKENS})
-        scipy.io.wavfile.write(audio_path, rate=music["sampling_rate"], data=music["audio"])
-
-    def generate_music_api(self, prompt, audio_path):
-        headers =  {"Authorization": f"Bearer {config.HF_API_TOKEN}"}
-        payload = {
-            "inputs": prompt
-        }
-
-        response = requests.post(config.MUSICGEN_MODEL_API_URL, headers=headers, json=payload)
-
-        # ----ATTRIBUTION-START----
-        # LLM: ChatGPT4o
-        # PROMPT: please save the audio to a .wav file
-        # EDITS: changed variables to match the code
-
-        # Convert the byte content into an audio array
-        try:
-            audio_buffer = BytesIO(response.content)
-            # Use scipy to save the audio, assuming it's a WAV format audio stream
-            # If it's raw PCM audio, you would need to decode it first.
-            with open(audio_path, "wb") as f:
-                f.write(audio_buffer.read())
-            # -----ATTRIBUTION-END-----
-        except Exception as e:
-            print(f"Error: {e}")
-

phi3_mini_4k_instruct.py

@@ -1,45 +0,0 @@
-# external imports
-from transformers import pipeline
-from huggingface_hub import InferenceClient
-import torch
-# local imports
-import config
-from llama_cpp import Llama
-
-
-class Phi3_Mini_4k_Instruct:
-    def __init__(self):
-        pass
-
-    def generate_text(self, messages, use_local_llm):
-        if use_local_llm:
-            return self.generate_text_llama_cpp(messages)
-        else:
-            return self.generate_text_api(messages)
-
-    def generate_text_llama_cpp(self, messages):
-        model = Llama.from_pretrained(
-            repo_id="microsoft/Phi-3-mini-4k-instruct-gguf",
-            filename="Phi-3-mini-4k-instruct-q4.gguf"
-        )
-        response = model.create_chat_completion(messages)
-        generated_message = response['choices'][0]['message']['content']
-
-        return generated_message
-
-    def generate_text_local_pipeline(self, messages):
-        self.local_pipeline = pipeline("text-generation", model=config.LLM_MODEL, trust_remote_code=True, torch_dtype=torch.bfloat16, device_map="auto")
-        self.local_pipeline.model.config.max_length = config.LLM_MAX_LENGTH
-        self.local_pipeline.model.config.max_new_tokens = config.LLM_MAX_NEW_TOKENS
-        self.local_pipeline.model.config.temperature = config.LLM_TEMPERATURE
-        self.local_pipeline.model.config.top_p = config.LLM_TOP_P
-        result = self.local_pipeline(messages)[-1]['generated_text'][-1]['content']
-        return result
-
-    def generate_text_api(self, messages):
-        client = InferenceClient(config.LLM_MODEL, token=config.HF_API_TOKEN)
-        try:
-            result = client.chat_completion(messages, max_tokens=config.LLM_MAX_NEW_TOKENS, temperature=config.LLM_TEMPERATURE, top_p=config.LLM_TOP_P).choices[0].message.content
-        except Exception as e:
-            result = f"Error: {e}"
-        return result

requirements.txt

@@ -1,74 +1,5 @@
-accelerate==0.34.2
-aiofiles==23.2.1
-annotated-types==0.7.0
-anyio==4.4.0
-certifi==2024.8.30
-charset-normalizer==3.3.2
-click==8.1.7
-contourpy==1.3.0
-cycler==0.12.1
-fastapi==0.114.2
-ffmpy==0.4.0
-filelock==3.16.0
-fonttools==4.53.1
-fsspec==2024.9.0
-gradio==4.44.0
-gradio_client==1.3.0
-h11==0.14.0
-httpcore==1.0.5
-httpx==0.27.2
-huggingface-hub==0.24.6
-idna==3.8
-importlib_resources==6.4.5
-iniconfig==2.0.0
-Jinja2==3.1.4
-kiwisolver==1.4.7
-llama_cpp_python==0.3.1
-markdown-it-py==3.0.0
-MarkupSafe==2.1.5
-matplotlib==3.9.2
-mdurl==0.1.2
-mpmath==1.3.0
-networkx==3.3
-numpy==2.1.1
-orjson==3.10.7
-packaging==24.1
-pandas==2.2.2
-pillow==10.4.0
-pluggy==1.5.0
-psutil==6.0.0
-pydantic==2.9.1
-pydantic_core==2.23.3
-pydub==0.25.1
-Pygments==2.18.0
-pyparsing==3.1.4
-pytest==8.3.3
-python-dateutil==2.9.0.post0
-python-multipart==0.0.9
-pytz==2024.2
-PyYAML==6.0.2
-regex==2024.9.11
-requests==2.32.3
-rich==13.8.1
-ruff==0.6.5
-safetensors==0.4.5
-scipy==1.14.1
-semantic-version==2.10.0
-shellingham==1.5.4
-six==1.16.0
-sniffio==1.3.1
-starlette==0.38.5
-sympy==1.13.2
-tokenizers==0.19.1
-tomlkit==0.12.0
-torch==2.4.1
-torchaudio==2.4.1
-torchvision==0.19.1
-tqdm==4.66.5
-transformers==4.44.2
-typer==0.12.5
-typing_extensions==4.12.2
-tzdata==2024.1
-urllib3==2.2.2
-uvicorn==0.30.6
-websockets==12.0
+huggingface_hub==0.23.*
+gradio==4.39.*
+torch==2.4.*
+transformers==4.43.*
+accelerate==0.33.*

setup.sh

@@ -1,52 +0,0 @@
-#!/bin/bash
-
-VENV_DIR="venv"
-
-#ensure package list
-sudo add-apt-repository -y universe
-#ensure python and requirements is installed
-sudo apt install -qq -y python3-venv
-sudo apt install -qq -y python3-pip
-sudo apt install -y build-essential
-sudo apt install -y gcc g++ 
-sudo apt install -y screen
-
-
-# Check if the virtual environment exists
-if [ ! -d "$VENV_DIR" ]; then
-  echo "Virtual environment not found. Creating a new one..."
-  # Create a virtual environment
-  python3 -m venv "$VENV_DIR"
-  echo "Virtual environment created."
-
-else
-    echo "Virtual environment found."
-fi
-
-# Activate the virtual environment
-source "$VENV_DIR/bin/activate"
-echo "Virtual environment $VENV_DIR activated."
-
-pip install --upgrade pip
-
-
-if git pull | grep -q 'Already up to date.'; then
-    echo "Repository is up to date. Proceeding with setup."
-
-else
-    echo "Repository updated successfully. Proceeding to next step."
-fi
-
-echo "Checking if http://127.0.0.1:7860 is running..."
-if curl -s --head http://127.0.0.1:7860 | grep "200 OK" > /dev/null; then
-    echo "URL is running.No further action required. Exiting."
-    exit 0  # Exit script since the service is already running
-else
-    echo "URL is not running.Proceeding with setup."
-    # Install dependencies and run the application
-    pip install -r requirements.txt
-
-    screen -S "app" -d -m bash -c 'python3 app.py'
-fi
-deactivate
-exit 0

setupaccess.exp

@@ -1,13 +0,0 @@
-#!/usr/bin/expect -f
-set PASSPHRASE_GROUP17 [lindex $argv 0]
-spawn ./setupaccess.sh
-set timeout 5
-expect "The authenticity of host"
-send "yes\r"
-expect "Enter passphrase for key 'group17':\r"
-send "$PASSPHRASE_GROUP17\r"
-expect "Enter passphrase for key 'group17':\r"
-send "$PASSPHRASE_GROUP17\r"
-expect "Enter passphrase for key 'group17':\r"
-send "$PASSPHRASE_GROUP17\r"
-expect eof

setupaccess.sh

@@ -1,45 +0,0 @@
-#!/bin/bash
-
-touch group17.pub
-echo "$GROUP17_PUBLICKKEY" > group17.pub
-echo "setupaccess.sh: make group17.pub file"
-
-touch group17
-echo "$GROUP17_PRIVATEKEY" > group17
-echo "setupaccess.sh: make group17 file"
-
-chmod 600 group17
-echo "setupaccess.sh: change permissions of group17 file"
-
-ssh-keygen -R "[paffenroth-23.dyn.wpi.edu]:22017"
-echo "setupaccess.sh: remove known host keys for the server to avoid the REMOTE HOST IDENTIFICATION HAS CHANGED error"
-
-cat group17.pub > authorized_keys
-echo "setupaccess.sh: make an authorized_keys file with group17.pub as an authorized key"
-
-rm group17.pub
-echo "setupaccess.sh: remove group17.pub file from host"
-
-scpOutput=$(scp -o StrictHostKeyChecking=no -i group17 -P 22017 authorized_keys student-admin@paffenroth-23.dyn.wpi.edu:/home/student-admin/.ssh 2>&1)
-echo "setupaccess.sh: try to copy authorized_keys file to server"
-
-if [[ "$scpOutput" = *"student-admin@paffenroth-23.dyn.wpi.edu: Permission denied (publickey)."* ]];
-then touch student-admin_key
-    echo "$STUDENT_ADMIN_KEY" > student-admin_key
-    echo "setupaccess.sh: make student-admin_key file"
-    scp -o StrictHostKeyChecking=no -i student-admin_key -P 22017 authorized_keys student-admin@paffenroth-23.dyn.wpi.edu:/home/student-admin/.ssh
-    echo "setupaccess.sh: copied authorized_keys file to server with student-admin_key"
-    rm student-admin_key
-    echo "setupaccess.sh: remove student-admin_key from host"
-else
-    echo "setupaccess.sh: copied authorized_keys file to server with our private key"
-fi
-
-rm authorized_keys
-echo "setupaccess.sh: remove authorized_keys file from host"
-
-ssh -p 22017 -i group17 -o StrictHostKeyChecking=no student-admin@paffenroth-23.dyn.wpi.edu
-echo "setupaccess.sh: try to ssh in"
-
-rm group17
-echo "setupaccess.sh: remove group17 file from host"

test_blip_image_caption_large.py

@@ -1,16 +0,0 @@
-from blip_image_caption_large import Blip_Image_Caption_Large
-
-# Test the local image caption pipeline with wikipedia image 
-def test_blip_image_caption_local_model():
-    image_caption_model = Blip_Image_Caption_Large()
-    image_path = "data/Students_taking_computerized_exam.jpg"
-    result = image_caption_model.caption_image(image_path, use_local_caption=True)
-    assert result == "several people sitting at desks with computers in a classroom"
-
-# Test the image caption API with wikipedia image
-def test_blip_image_caption_api():
-    image_caption_model = Blip_Image_Caption_Large()
-    image_path = "data/Students_taking_computerized_exam.jpg"
-    result = image_caption_model.caption_image(image_path, use_local_caption=False)
-    assert result == "several people sitting at desks with computers in a classroom"
-    

test_musicgen_small.py

@@ -1,26 +0,0 @@
-from musicgen_small import Musicgen_Small
-
-import config
-import os
-
-# Test the local Musicgen_Small class with a 5 second music generation and assert file creation
-def test_musicgen_small_local_model():
-    musicgen_model = Musicgen_Small()
-    prompt = "a very testy song, perfect to test the music generation model"
-    audio_path = f"{config.AUDIO_DIR}/test_musicgen_small_local.wav"
-    musicgen_model.generate_music(prompt, audio_path, use_local_musicgen=True)
-    assert os.path.exists(audio_path)
-    assert os.path.getsize(audio_path) > 0
-    os.remove(audio_path)
-    assert not os.path.exists(audio_path)
-
-# Test the Musicgen_Small API with a 30 second music generation and assert file creation
-def test_musicgen_small_api():
-    musicgen_model = Musicgen_Small()
-    prompt = "a very testy song, perfect to test the music generation model"
-    audio_path = f"{config.AUDIO_DIR}/test_musicgen_small_api.wav"
-    musicgen_model.generate_music(prompt, audio_path, use_local_musicgen=False)
-    assert os.path.exists(audio_path)
-    assert os.path.getsize(audio_path) > 0
-    os.remove(audio_path)
-    assert not os.path.exists(audio_path)

test_phi3_mini_4k_instruct.py

@@ -1,20 +0,0 @@
-from phi3_mini_4k_instruct import Phi3_Mini_4k_Instruct
-
-# Test the local Phi3_Mini_4k_Instruct Model with default values
-def test_phi3_mini_4k_instruct_local():
-    phi3_mini_4k_instruct = Phi3_Mini_4k_Instruct()
-    messages = [
-        {"role": "system", "content": "You are an image caption to song description converter with a deep understanding of Music and Art. You are given the caption of an image. Your task is to generate a textual description of a musical piece that fits the caption. The description should be detailed and vivid, and should include the genre, mood, instruments, tempo, and other relevant information about the music. You should also use your knowledge of art and visual aesthetics to create a musical piece that complements the image. Only output the description of the music, without any explanation or introduction. Be concise."},
-        {"role": "user", "content": "several people sitting at desks with computers in a classroom"},
-    ]
-    generated_description = phi3_mini_4k_instruct.generate_text(messages, use_local_llm=True)
-    assert isinstance(generated_description, str) and generated_description != ""
-
-def test_phi3_mini_4k_instruct_api():
-    phi3_mini_4k_instruct = Phi3_Mini_4k_Instruct()
-    messages = [
-        {"role": "system", "content": "You are an image caption to song description converter with a deep understanding of Music and Art. You are given the caption of an image. Your task is to generate a textual description of a musical piece that fits the caption. The description should be detailed and vivid, and should include the genre, mood, instruments, tempo, and other relevant information about the music. You should also use your knowledge of art and visual aesthetics to create a musical piece that complements the image. Only output the description of the music, without any explanation or introduction. Be concise."},
-        {"role": "user", "content": "several people sitting at desks with computers in a classroom"},
-    ]
-    generated_description = phi3_mini_4k_instruct.generate_text(messages, use_local_llm=False)
-    assert isinstance(generated_description, str) and generated_description != ""

tmp.txt

@@ -0,0 +1 @@
+ponytails are really cool 3