first commit for the project

.gitignore

@@ -210,5 +210,8 @@ pyrightconfig.json
 config.py
 img/
 output_img/
+test.py
+test2.py
+
 
 # End of https://www.toptal.com/developers/gitignore/api/python,macos

.streamlit/config.toml

@@ -0,0 +1,6 @@
+[theme]
+primaryColor="#E694FF"
+backgroundColor="#0E1117"
+secondaryBackgroundColor="#31333F"
+textColor="#FAFAFA"
+font="sans serif"

app.py

@@ -1 +1,83 @@
-# TODO: Streamlit demo
+import streamlit as st
+from gpt_vision_prompt import generate_prompt_with_vision
+import tempfile
+from dalle_generate_img import generate_img_with_dalle
+from stability_generate_img import generate_image_with_stability
+from finetune_generate_img import generate_finetuned_img
+from midjourney_generate_img import midjourney_generate_img
+
+# Page configuration
+st.set_page_config(layout="wide")
+st.sidebar.title("API Keys")
+st.markdown("<h1 style='text-align: center; color: grey;'>Image Generation App</h1>", unsafe_allow_html=True)
+st.text("Prepared by [email protected] for fomo.ai")
+
+# List of API key names
+api_key_names = ["OPENAI_API_KEY", "MIDJOURNEY_GOAPI_TOKEN", "REPLICATE_API_TOKEN", "STABILITY_API_KEY"]
+
+# Initialize session state if it does not exist
+if 'api_keys' not in st.session_state:
+    st.session_state['api_keys'] = {key_name: "" for key_name in api_key_names}
+if 'editable_prompt' not in st.session_state:
+    st.session_state['editable_prompt'] = ""
+if 'upload_completed' not in st.session_state:
+    st.session_state['upload_completed'] = False
+
+# Define a function to request and update API keys
+def request_and_update_api_keys():
+    all_keys_entered = True
+    for key_name in api_key_names:
+        key_value = st.sidebar.text_input(f"Enter {key_name}:", value=st.session_state['api_keys'].get(key_name, ""), type="password", key=key_name)
+        st.session_state['api_keys'][key_name] = key_value
+        if not key_value:
+            all_keys_entered = False
+    return all_keys_entered
+
+all_keys_entered = request_and_update_api_keys()
+
+# Check if all API keys have been entered
+if all_keys_entered:
+    # Section to upload the image
+    uploaded_file = st.file_uploader("Upload Image to analyze", type=['jpg', 'jpeg', 'png'], on_change=lambda: setattr(st.session_state, 'upload_completed', True))
+    if uploaded_file is not None:
+        st.session_state['upload_completed'] = True
+        left_co, cent_co, _ = st.columns([1, 2, 1])
+        with cent_co:
+            st.image(uploaded_file, caption="Uploaded Image")
+        _, right_co = st.columns([5, 1])
+        if right_co.button("Generate Prompt"):
+            with st.spinner("Generating Prompt..."):
+                with tempfile.NamedTemporaryFile(delete=False, suffix='.png') as temp_file:
+                    temp_file.write(uploaded_file.getvalue())
+                    temp_path = temp_file.name
+                api_key = st.session_state['api_keys']['OPENAI_API_KEY']
+                prompt = generate_prompt_with_vision(temp_path, api_key=api_key)
+                st.session_state['editable_prompt'] = prompt
+                st.session_state['upload_completed'] = False
+
+    if st.session_state['upload_completed']:
+        # Maintain the content of the editable prompt after generating images
+        editable_prompt = st.text_area("Edit the prompt as needed:", value=st.session_state['editable_prompt'], placeholder="Enter your prompt here...", height=150, key='editable_prompt', on_change=lambda: st.session_state.update(editable_prompt=editable_prompt))
+        if st.button("Generate New Image", key='generate_image_btn'):
+            col1, col2, col3, col4 = st.columns(4)
+            with col1:
+                with st.spinner("Generating DALL·E Image..."):
+                    result_path_1 = generate_img_with_dalle(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['OPENAI_API_KEY'])
+                    st.image(result_path_1, caption="DALL·E Image")
+            with col2:
+                with st.spinner("Generating Stable Diffusion Image..."):
+                    result_path_2 = generate_image_with_stability(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['STABILITY_API_KEY'])
+                    st.image(result_path_2, caption="Stable Diffusion Image")
+            with col3:
+                with st.spinner("Generating Finetuning Image..."):
+                    result_path_3 = generate_finetuned_img(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['REPLICATE_API_TOKEN'])
+                    st.image(result_path_3, caption="Finetuned SDXL Image")
+            with col4:
+                with st.spinner("Generating Midjourney Image..."):
+                    result_path_4 = midjourney_generate_img(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['MIDJOURNEY_GOAPI_TOKEN'])
+                    st.image(result_path_4, caption="Midjourney Image")
+            # Update the prompt in session state to keep the text
+            st.session_state['editable_prompt'] = editable_prompt
+            st.session_state['upload_completed'] = False  # Disable the text area and button after generating images
+else:
+    st.warning('Please enter all required API keys to proceed.', icon="⚠️")

dalle_generate_img.py

@@ -13,10 +13,9 @@ import requests
 from config import OPENAI_API_KEY
 
 
-client = OpenAI(api_key=OPENAI_API_KEY)
-
-def generate_img_with_dalle(prompt="", ):
+def generate_img_with_dalle(prompt="", api_key=OPENAI_API_KEY):
     """Generate an image using the DALL-E API"""
+    client = OpenAI(api_key=api_key)
     # DALL-E model parameters
     size = '1024x1024'  # Choose between '1024x1024', '512x512', '256x256'
     quality = 'hd'  # Choose between 'standard', 'hd'
@@ -36,4 +35,4 @@ def generate_img_with_dalle(prompt="", ):
     img = Image.open(io.BytesIO(response.content))
     img.save('output_img/dalle_generated_img.png')  # Save the image as a .png file
     print('Image saved in output_img/dalle_generated_img.png')
-    return "Image saved in output_img/dalle_generated_img.png"
+    return "output_img/dalle_generated_img.png"

finetune_generate_img.py

@@ -10,11 +10,8 @@ import requests
 from PIL import Image
 from config import REPLICATE_API_TOKEN
 
-# Set up environment variables for Replicate API
-os.environ['REPLICATE_API_TOKEN'] = REPLICATE_API_TOKEN
 
-
-def generate_finetuned_img(prompt):
+def generate_finetuned_img(prompt, api_key=REPLICATE_API_TOKEN):
     """
     Generate a finetuned image based on the given prompt.
 
@@ -24,6 +21,8 @@ def generate_finetuned_img(prompt):
     Returns:
     str: The file path of the saved finetuned image.
     """
+    # Set up environment variables for Replicate API
+    os.environ['REPLICATE_API_TOKEN'] = api_key
     # Create finetuned image
     print('Creating finetuned image...')
     output = replicate.run(

gpt_vision_prompt.py

@@ -55,4 +55,7 @@ def generate_prompt_with_vision(image_path, prompt=PROMPT, api_key=OPENAI_API_KE
         headers=headers,
         json=payload,
         timeout=30)
+    print(response.status_code)
+    print(response.text)
+    print(response.json())
     return response.json()['choices'][0]['message']['content']

midjourney_generate_img.py

@@ -1,5 +1,6 @@
 """ Generate an image using the Midjourney API"""
 import io
+import random
 import requests
 from PIL import Image
 from progress_bar import print_progress_bar
@@ -12,13 +13,14 @@ headers = {
     "X-API-KEY": GOAPIKEY
 }
 
-def midjourney_generate_img(prompt):
+def midjourney_generate_img(prompt, api_key=GOAPIKEY):
     """Generate an image using the Midjourney API
 
     Keyword arguments:
     prompt -- The prompt to generate the image from
     Return: An image saved in a .png file
     """
+    headers["X-API-KEY"] = api_key
     img_generation_data = {
         "prompt": prompt,
         "aspect_ratio": "16:9",
@@ -66,4 +68,21 @@ def midjourney_generate_img(prompt):
     img = Image.open(io.BytesIO(image_response.content))
     img.save('output_img/midjourney_generated_img.png')
     print("Image saved in output_img/midjourney_generated_img.png")
-    return "output_img/midjourney_generated_img.png"
+
+    # divide img by 4 and save only one part
+    img_width, img_height = img.size
+    target_width = img_width // 2
+    target_height = img_height // 2
+    part = random.randint(1, 4) # select a random part
+    if part == 1:
+        img_cropped = img.crop((0, 0, target_width, target_height))  # Superior izquierda
+    elif part == 2:
+        img_cropped = img.crop((target_width, 0, img_width, target_height))  # Superior derecha
+    elif part == 3:
+        img_cropped = img.crop((0, target_height, target_width, img_height))  # Inferior izquierda
+    else:
+        img_cropped = img.crop((target_width, target_height, img_width, img_height))  # Inferior derecha
+    # save the selected img
+    img_cropped.save('output_img/midjourney_single_img.png')
+    print("Single image saved in output_img/midjourney_single_img.png")
+    return "output_img/midjourney_single_img.png"

requirements.txt

@@ -0,0 +1,124 @@
+adbc_driver_manager==0.10.0
+adbc_driver_postgresql==0.10.0
+adbc_driver_sqlite==0.10.0
+altair_saver==0.5.0
+altair_viewer==0.4.0
+anywidget==0.9.2
+AppKit==0.2.8
+atheris==2.3.0
+beautifulsoup4==4.12.3
+bokeh==3.2.1
+boto3==1.34.54
+botocore==1.29.76
+brotlicffi==1.1.0.0
+brotlipy==0.7.0
+cached_property==1.5.2
+chart_studio==1.1.0
+ConfigParser==6.0.1
+contextlib2==21.6.0
+cryptography==41.0.3
+ctypes_snappy==1.03
+curio==1.6
+cycler==0.11.0
+Cython==3.0.8
+cytoolz==0.12.0
+defusedxml==0.7.1
+diffusers==0.26.3
+disco==1.40.4
+dl==0.1.0
+docutils==0.18.1
+docutils==0.18.1
+email_validator==2.1.1
+eval_type_backport==0.1.3
+exceptiongroup==1.2.0
+fastparquet==2024.2.0
+filelock==3.13.1
+Foundation==0.1.0a0.dev1
+fqdn==1.5.1
+fsspec==2023.4.0
+gitdb_speedups==0.1.0
+gradio==4.19.2
+grpc_reflection==1.0.0
+h2==4.1.0
+HTMLParser==0.0.2
+hypothesis==6.98.15
+ipython==8.12.3
+ipywidgets==8.0.4
+isoduration==20.11.0
+jnius==1.1.0
+JPype1==1.5.0
+jsonpointer==2.1
+keyframed==0.3.15
+keyring==23.13.1
+linkify_it_py==2.0.0
+lxml==4.9.3
+lz4==4.3.2
+matplotlib==3.7.2
+moto==5.0.2
+mtrand==0.1
+numarray==1.5.1
+Numeric==24.2
+numexpr==2.8.4
+odfpy==1.4.1
+olefile==0.47
+openpyxl==3.0.10
+outcome==1.3.0.post0
+pickle5==0.0.12
+pkgutil_resolve_name==1.3.10
+plotly==5.19.0
+psutil==5.9.0
+pycares==4.4.0
+pycurl==7.45.2
+PyInstaller==6.4.0
+pynvml==11.5.0
+pyobjc_framework_Cocoa==9.0
+pyodide==0.0.2
+pyOpenSSL==23.2.0
+pyOpenSSL==24.0.0
+pyperf==2.6.2
+PyQt4==4.11.4
+PyQt5==5.15.10
+PyQt5_sip==12.11.0
+PyQt6==6.6.1
+PySide6==6.6.2
+pytest==7.4.0
+python_calamine==0.2.0
+pyxlsb==1.0.10
+PyYAML==6.0
+PyYAML==6.0.1
+QtPy==2.2.0
+railroad==0.5.0
+redis==5.0.2
+rfc3339_validator==0.1.4
+rfc3986_validator==0.1.1
+rfc3987==1.3.8
+s3fs==2023.4.0
+scikit_learn==1.3.0
+scipy==1.12.0
+sets==0.3.2
+setuptools_scm==8.0.4
+simplejson==3.19.2
+slack_sdk==3.27.1
+snowflake==0.6.0
+socksio==1.0.0
+Sphinx==5.0.2
+SQLAlchemy==1.4.39
+sympy==1.11.1
+tables==3.8.0
+testbench==0.1.2
+threadpoolctl==3.3.0
+torch==2.2.1
+traitlets==5.14.1
+transformers==4.32.1
+trove_classifiers==2024.2.23
+uri_template==1.3.0
+urllib3_secure_extra==0.1.0
+uvloop==0.19.0
+vegafusion==1.6.5
+watchdog==2.1.6
+webcolors==1.13
+xarray==2023.6.0
+xlrd==2.0.1
+xlsxwriter==3.2.0
+xmlrpclib==1.0.1
+zstandard==0.19.0

stability_generate_img.py

@@ -9,24 +9,13 @@ Return: An image saved in a .png file
 import os
 import io
 import warnings
-
 from stability_sdk import client
 import stability_sdk.interfaces.gooseai.generation.generation_pb2 as generation
 from PIL import Image
 from config import STABILITY_API_KEY
 
-# Set up environment variables for Stability API
-os.environ['STABILITY_HOST'] = 'grpc.stability.ai:443'
-os.environ['STABILITY_KEY'] = STABILITY_API_KEY
-
-# Set up our connection to the Stability API.
-stability_api = client.StabilityInference(
-    key=os.environ['STABILITY_KEY'],
-    verbose=True,
-    engine="stable-diffusion-xl-1024-v1-0",
-)
 
-def generate_image_with_stability(prompt, seed=42, steps=50, cfg_scale=7.0, width=1024, height=1024, samples=1):
+def generate_image_with_stability(prompt, seed=42, steps=50, cfg_scale=7.0, width=1024, height=1024, samples=1, api_key=STABILITY_API_KEY):
     """
     Generates an image based on the given prompt using Stability API.
 
@@ -39,6 +28,15 @@ def generate_image_with_stability(prompt, seed=42, steps=50, cfg_scale=7.0, widt
     :param samples: Number of images to generate.
     :return: A PIL.Image object of the generated image.
     """
+    os.environ['STABILITY_HOST'] = 'grpc.stability.ai:443'
+    os.environ['STABILITY_KEY'] = api_key
+    # Set up our connection to the Stability API.
+    stability_api = client.StabilityInference(
+        key=os.environ['STABILITY_KEY'],
+        verbose=True,
+        engine="stable-diffusion-xl-1024-v1-0",
+        )
+
     print("Creating Stability Image...")
     answers = stability_api.generate(
         prompt=prompt,
@@ -63,6 +61,6 @@ def generate_image_with_stability(prompt, seed=42, steps=50, cfg_scale=7.0, widt
                 img = Image.open(io.BytesIO(artifact.binary))
                 img.save("output_img/sd_generated_img.png")
                 print("Image saved in output_img/sd_generated_img.png")
-                return "Image saved in output_img/sd_generated_img.png"
+                return "output_img/sd_generated_img.png"
 
     raise ValueError("No image was generated.")

test.py

@@ -0,0 +1,86 @@
+import streamlit as st
+from gpt_vision_prompt import generate_prompt_with_vision
+import tempfile
+from dalle_generate_img import generate_img_with_dalle
+from stability_generate_img import generate_image_with_stability
+from finetune_generate_img import generate_finetuned_img
+from midjourney_generate_img import midjourney_generate_img
+
+
+# Configuración de la página
+st.set_page_config(layout="wide")
+st.sidebar.title("API Keys")
+st.markdown("<h1 style='text-align: center; color: grey;'>Image Generation App</h1>", unsafe_allow_html=True)
+st.text("Prepared by [email protected] for fomo.ai")
+
+# Lista de nombres de las API keys
+api_key_names = ["OPENAI_API_KEY", "MIDJOURNEY_GOAPI_TOKEN", "REPLICATE_API_TOKEN", "STABILITY_API_KEY"]
+
+# Inicializar el estado de la sesión si no existe
+if 'api_keys' not in st.session_state:
+    st.session_state['api_keys'] = {key_name: "" for key_name in api_key_names}
+if 'editable_prompt' not in st.session_state:
+    st.session_state['editable_prompt'] = ""
+
+# Definir una función para solicitar y actualizar las API keys
+def request_and_update_api_keys():
+    all_keys_entered = True
+    for key_name in api_key_names:
+        key_value = st.sidebar.text_input(f"Enter {key_name}:", value=st.session_state['api_keys'].get(key_name, ""), type="password", key=key_name)
+        st.session_state['api_keys'][key_name] = key_value
+        if not key_value:
+            all_keys_entered = False
+    return all_keys_entered
+
+all_keys_entered = request_and_update_api_keys()
+
+# Revisar si todas las API keys han sido ingresadas
+if all_keys_entered:
+    # Sección para subir la imagen
+    uploaded_file = st.file_uploader("Upload Image to analyze", type=['jpg', 'jpeg', 'png'])
+    if uploaded_file is not None:
+        left_co, cent_co,last_co = st.columns(3)
+        with cent_co:
+            st.image(uploaded_file, caption="Uploaded Image")
+        # Botón para generar el prompt solo si hay una imagen subida
+        if st.button("Generate Prompt"):
+            with st.spinner("Generating Prompt..."):
+                with tempfile.NamedTemporaryFile(delete=False, suffix='.png') as temp_file:
+                    temp_file.write(uploaded_file.getvalue())
+                    temp_path = temp_file.name
+                api_key = st.session_state['api_keys']['OPENAI_API_KEY']
+                prompt = generate_prompt_with_vision(temp_path, api_key=api_key)
+                st.success("Done!")
+                st.session_state['editable_prompt'] = prompt  # Actualizar el prompt en el estado de la sesión
+
+    editable_prompt = st.text_area("Edit the prompt as needed:", placeholder="Enter your prompt here...", height=150, key='editable_prompt', label_visibility='hidden')
+
+    col1, col2, col3, col4 = st.columns(4)
+    
+    if st.session_state['editable_prompt'] and st.button("Generate New Image"):
+        with col1:
+            with st.spinner("Generating DALL·E Image..."):
+                result_path_1 = generate_img_with_dalle(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['OPENAI_API_KEY'])
+                st.success("Generated DALL·E Image!")
+                st.image(result_path_1, caption="DALL·E Image")
+                
+        with col2:
+            with st.spinner("Generating Stable Diffusion Image..."):
+                result_path_2 = generate_image_with_stability(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['STABILITY_API_KEY'])
+                st.success("Generated Stable Diffusion Image!")
+                st.image(result_path_2, caption="Stable Diffusion Image")
+        
+        with col3:
+            with st.spinner("Generating Finetuning Image..."):
+                result_path_3 = generate_finetuned_img(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['REPLICATE_API_TOKEN'])
+                st.success("Generated Image using a finetuned model!")
+                st.image(result_path_3, caption="Finetuned SDXL Image")
+        
+        with col4:
+            with st.spinner("Generating Midjourney Image..."):
+                result_path_4 = midjourney_generate_img(st.session_state['editable_prompt'], api_key=st.session_state['api_keys']['MIDJOURNEY_GOAPI_TOKEN'])
+                st.success("Generated Midjourney Image!")
+                st.image(result_path_4, caption="Midjourney Image")
+
+else:
+    st.warning('Please enter all required API keys to proceed.', icon="⚠️")