autonomous-app 0.3.47__tar.gz → 0.3.49__tar.gz

{autonomous_app-0.3.47 → autonomous_app-0.3.49}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: autonomous-app
-Version: 0.3.47
+Version: 0.3.49
 Summary: Containerized application framework built on Flask with additional libraries and tools for rapid development of web applications.
 Author-email: Steven A Moore <samoore@binghamton.edu>
 Project-URL: homepage, https://github.com/Sallenmoore/autonomous

{autonomous_app-0.3.47 → autonomous_app-0.3.49}/src/autonomous/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.3.47"
+__version__ = "0.3.49"
 
 from dotenv import load_dotenv
 

{autonomous_app-0.3.47 → autonomous_app-0.3.49}/src/autonomous/ai/models/local_model.py

@@ -4,6 +4,7 @@ import os
 import random
 
 import requests
+from PIL import Image
 from pydub import AudioSegment
 
 from autonomous import log
@@ -119,7 +120,7 @@ class LocalAIModel(AutoModel):
             },
         }
 
-        # log("==== LocalAI JSON Payload ====", payload, _print=True)
+        log("==== LocalAI JSON Payload ====", payload, _print=True)
 
         result_text = ""
         try:
@@ -139,7 +140,7 @@ class LocalAIModel(AutoModel):
             ):
                 params = result_dict.pop("parameters")
                 result_dict.update(params)
-
+            log("==== LocalAI JSON Result ====", result_dict, _print=True)
             return result_dict
 
         except Exception as e:
@@ -242,64 +243,60 @@ class LocalAIModel(AutoModel):
             log(f"TTS Error: {e}", _print=True)
             return None
 
-    # ... inside LocalAIModel class ...
-
     def _get_dimensions(self, aspect_ratio):
         """
-        Maps abstract aspect ratios to optimal SDXL resolutions.
-        SDXL performs best at ~1024x1024 total pixels.
+        Returns a tuple: ((base_w, base_h), (target_w, target_h))
+
+        1. base_*: The resolution sent to SDXL (approx 1024x1024).
+        2. target_*: The final resolution to resize/upscale to.
         """
-        resolutions = {
+        # Standard SDXL buckets (approx 1MP)
+        # We use these for the initial generation to ensure good composition.
+        sdxl_base = {
             "1:1": (1024, 1024),
-            "3:4": (896, 1152),
-            "4:3": (1152, 896),
-            "16:9": (1216, 832),
-            "2K": (2048, 1080),
-            "2KPortrait": (1080, 2048),
-            "Portrait": (1080, 2048),
-            "4K": (3840, 2160),
-            "Landscape": (3840, 2160),
-            "4KPortrait": (2160, 3840),
-            "9:16": (832, 1216),
-            "3:2": (1216, 832),
-            "2:3": (832, 1216),
+            "Portrait": (896, 1152),  # 3:4
+            "Landscape": (1216, 832),  # 3:2 or 16:9 approx
+        }
+
+        # The Logic: Define the target, map it to the closest SDXL base
+        # Format: "Key": ((Base_W, Base_H), (Target_W, Target_H))
+        resolutions = {
+            # Standard
+            "1:1": ((832, 832), (1024, 1024)),
+            "3:4": ((832, 1152), (1664, 2304)),
+            "4:3": ((1152, 832), (2304, 1664)),
+            # High Res (The logic changes here)
+            "16:9": ((1216, 832), (2048, 1152)),
+            "9:16": ((832, 1216), (1152, 2048)),
+            # 2K Tier
+            "2K": ((1216, 832), (2048, 1152)),  # Base is 1216x832 -> Upscale to 2K
+            "2KPortrait": ((832, 1216), (1152, 2048)),
+            # 4K Tier (The generated image will be upscaled ~3x)
+            "4K": ((1216, 832), (3840, 2160)),
+            "4KPortrait": ((832, 1216), (2160, 3840)),
         }
-        # Default to 1:1 (1024x1024) if unknown
-        return resolutions.get(aspect_ratio, (1024, 1024))
+
+        # Default to 1:1 if unknown
+        return resolutions.get(aspect_ratio, ((832, 832), (1024, 1024)))
 
     def generate_image(
         self, prompt, negative_prompt="", files=None, aspect_ratio="2KPortrait"
     ):
-        # # 1. CLIP Token Limit Fix (Auto-Summarize)
-        # if len(prompt) > 800:
-        #     log("⚠️ Prompt exceeds CLIP limit. rewriting...", _print=True)
-        #     summary_instruction = (
-        #         "Convert the description into a comma-separated Stable Diffusion prompt. "
-        #         "Keep visual elements and style. Under 50 words."
-        #     )
-        #     new_prompt = self.generate_text(
-        #         message=prompt, additional_instructions=summary_instruction, context={}
-        #     )
-        #     if new_prompt and len(new_prompt) > 10:
-        #         prompt = new_prompt
-
-        # 2. Resolution Calculation
-        width, height = self._get_dimensions(aspect_ratio)
-
-        # 3. Construct Payload
-        # We send both the abstract params (for logging/metadata)
-        # and the concrete pixels (for the engine).
+        # 1. Resolution Calculation
+        (base_w, base_h), (target_w, target_h) = self._get_dimensions(aspect_ratio)
+
+        # 2. Construct Base Generation Payload
+        # We tell the AI to generate the smaller, stable size first.
         data = {
             "prompt": prompt,
             "negative_prompt": negative_prompt,
             "aspect_ratio": aspect_ratio,
-            "width": width,  # <--- Calculated Pixel Width
-            "height": height,  # <--- Calculated Pixel Height
+            "width": base_w,
+            "height": base_h,
         }
 
         try:
-            # Handle Files (Corrected List Logic)
-            # requests.post expects a list of tuples for multiple files with same key
+            # Handle Input Files (for Img2Img)
             files_list = []
             if files and isinstance(files, dict):
                 for fn, f_bytes in files.items():
@@ -307,20 +304,45 @@ class LocalAIModel(AutoModel):
                         file_obj = io.BytesIO(f_bytes)
                     else:
                         file_obj = f_bytes
-                    # Appending to list instead of overwriting dict key
                     files_list.append(("files", (fn, file_obj, "image/png")))
 
-            # Send Request
+            # 3. Step 1: Generate Base Image
+            url = f"{self._media_url}/generate-image"
             if files_list:
-                response = requests.post(
-                    f"{self._media_url}/generate-image", data=data, files=files_list
-                )
+                response = requests.post(url, data=data, files=files_list)
             else:
-                response = requests.post(f"{self._media_url}/generate-image", data=data)
+                response = requests.post(url, data=data)
 
             response.raise_for_status()
-            log("==== LocalAI Image Payload ====", data, _print=True)
-            return response.content
+            image_content = response.content
+
+            # 4. Step 2: Upscale (If necessary)
+            if (base_w, base_h) != (target_w, target_h):
+                log(
+                    f"Requesting AI Upscale: {base_w}x{base_h} -> {target_w}x{target_h}...",
+                    _print=True,
+                )
+
+                # Prepare payload for the /upscale route
+                upscale_data = {
+                    "prompt": prompt,  # Reuse prompt to guide texture generation
+                    "width": target_w,  # Explicitly tell server the target size
+                    "height": target_h,
+                }
+
+                # Send the image we just generated back to the server as a file
+                upscale_files = {
+                    "file": ("generated.png", io.BytesIO(image_content), "image/png")
+                }
+
+                upscale_response = requests.post(
+                    f"{self._media_url}/upscale", data=upscale_data, files=upscale_files
+                )
+                upscale_response.raise_for_status()
+                image_content = upscale_response.content
+
+            log("==== LocalAI Image Generation Complete ====", data, _print=True)
+            return image_content
 
         except Exception as e:
             log(f"Image Gen Error: {e}", _print=True)

autonomous_app-0.3.49/src/autonomous/ai/models/mock_model.py

@@ -0,0 +1,78 @@
+import base64
+import io
+import json
+
+from autonomous import log
+from autonomous.model.autoattr import ListAttr, StringAttr
+from autonomous.model.automodel import AutoModel
+
+
+class MockAIModel(AutoModel):
+    """
+    A Offline Mock Model for development.
+    Returns valid, structure-compliant dummy data instantly.
+    """
+
+    messages = ListAttr(StringAttr(default=[]))
+    name = StringAttr(default="mock_agent")
+    instructions = StringAttr(default="You are a Mock AI.")
+    description = StringAttr(default="Offline Mock Provider for testing.")
+
+    # A 1x1 Red Pixel PNG (Base64 decoded to bytes)
+    # This ensures PIL/Frontend code receives valid image data.
+    _MOCK_IMAGE_BYTES = base64.b64decode(
+        "iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8z8BQDwAEhQGAhKmMIQAAAABJRU5ErkJggg=="
+    )
+
+    # A 1-second silent WAV file (Base64 decoded to bytes)
+    _MOCK_AUDIO_BYTES = base64.b64decode(
+        "UklGRigAAABXQVZFZm10IBAAAAABAAEAQB8AAEAfAAABAAgAZGF0YQAAAAEA"
+    )
+
+    def generate_json(self, message, system_prompt=None, uri="", context={}):
+        """
+        Returns a rich dummy JSON object.
+        It attempts to guess the context based on keys found in the prompt,
+        otherwise returns a safe default TTRPG object.
+        """
+        log(f"⚡ [MOCK] Generating JSON for prompt: {message[:50]}...", _print=True)
+
+        # 1. Default Mock Object
+        mock_response = {}
+
+        # 2. Heuristic: If the user provided a context with 'name', use it.
+        # This makes the mock feel slightly responsive.
+        if context:
+            if "name" in context:
+                mock_response["name"] = f"Mock {context['name']}"
+            if "title" in context:
+                mock_response["name"] = f"Mock {context['title']}"
+
+        return mock_response
+
+    def generate_text(self, message, additional_instructions="", uri="", context={}):
+        log(f"⚡ [MOCK] Generating Text...", _print=True)
+        return (
+            "Lorem ipsum dolor sit amet, consectetur adipiscing elit. "
+            "This is MOCK TEXT generated because you are currently offline. "
+            "The quick brown fox jumps over the lazy dog."
+        )
+
+    def summarize_text(self, text, primer=""):
+        log(f"⚡ [MOCK] Summarizing {len(text)} chars...", _print=True)
+        return "This is a mock summary of the provided text. It is concise and offline."
+
+    def generate_transcription(self, audio_file, prompt=""):
+        log(f"⚡ [MOCK] Transcribing Audio...", _print=True)
+        return "This is a mock transcription of the audio file."
+
+    def generate_audio(self, prompt, voice=None):
+        log(f"⚡ [MOCK] Generating Audio...", _print=True)
+        return self._MOCK_AUDIO_BYTES
+
+    def generate_image(
+        self, prompt, negative_prompt="", files=None, aspect_ratio="2KPortrait"
+    ):
+        log(f"⚡ [MOCK] Generating Image ({aspect_ratio})...", _print=True)
+        # Returns a valid 1x1 pixel PNG so your UI displays something red instead of crashing
+        return self._MOCK_IMAGE_BYTES

{autonomous_app-0.3.47 → autonomous_app-0.3.49}/src/autonomous_app.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: autonomous-app
-Version: 0.3.47
+Version: 0.3.49
 Summary: Containerized application framework built on Flask with additional libraries and tools for rapid development of web applications.
 Author-email: Steven A Moore <samoore@binghamton.edu>
 Project-URL: homepage, https://github.com/Sallenmoore/autonomous

{autonomous_app-0.3.47 → autonomous_app-0.3.49}/src/autonomous_app.egg-info/SOURCES.txt

@@ -14,6 +14,7 @@ src/autonomous/ai/textagent.py
 src/autonomous/ai/models/__init__.py
 src/autonomous/ai/models/gemini.py
 src/autonomous/ai/models/local_model.py
+src/autonomous/ai/models/mock_model.py
 src/autonomous/apis/version_control/GHCallbacks.py
 src/autonomous/apis/version_control/GHOrganization.py
 src/autonomous/apis/version_control/GHRepo.py