Spaces:
Running
on
Zero
Running
on
Zero
Update models.py
Browse files
models.py
CHANGED
@@ -1,22 +1,16 @@
|
|
1 |
"""
|
2 |
Model management for Frame 0 Laboratory for MIA
|
3 |
-
BAGEL 7B integration
|
4 |
"""
|
5 |
|
6 |
import logging
|
|
|
7 |
import os
|
8 |
-
import subprocess
|
9 |
-
import spaces
|
10 |
-
import torch
|
11 |
from typing import Optional, Dict, Any, Tuple
|
12 |
from PIL import Image
|
13 |
-
from
|
14 |
-
from accelerate import infer_auto_device_map, load_checkpoint_and_dispatch, init_empty_weights
|
15 |
|
16 |
-
from config import
|
17 |
-
BAGEL_CONFIG, get_device_config, get_bagel_device_map,
|
18 |
-
BAGEL_PROMPTS, FLASH_ATTN_INSTALL
|
19 |
-
)
|
20 |
from utils import clean_memory, safe_execute
|
21 |
|
22 |
logger = logging.getLogger(__name__)
|
@@ -26,7 +20,6 @@ class BaseImageAnalyzer:
|
|
26 |
"""Base class for image analysis models"""
|
27 |
|
28 |
def __init__(self):
|
29 |
-
self.model = None
|
30 |
self.is_initialized = False
|
31 |
self.device_config = get_device_config()
|
32 |
|
@@ -40,235 +33,153 @@ class BaseImageAnalyzer:
|
|
40 |
|
41 |
def cleanup(self) -> None:
|
42 |
"""Clean up model resources"""
|
43 |
-
if hasattr(self, 'model') and self.model is not None:
|
44 |
-
del self.model
|
45 |
-
self.model = None
|
46 |
clean_memory()
|
47 |
|
48 |
|
49 |
-
class
|
50 |
-
"""BAGEL 7B model
|
51 |
|
52 |
def __init__(self):
|
53 |
super().__init__()
|
54 |
-
self.
|
55 |
-
self.
|
56 |
-
self.
|
57 |
-
self.vae_transform = None
|
58 |
-
self.vit_transform = None
|
59 |
-
self._install_flash_attn()
|
60 |
|
61 |
-
def _install_flash_attn(self):
|
62 |
-
"""Install flash attention dynamically"""
|
63 |
-
try:
|
64 |
-
logger.info("Installing flash attention...")
|
65 |
-
result = subprocess.run(
|
66 |
-
FLASH_ATTN_INSTALL["command"],
|
67 |
-
env=FLASH_ATTN_INSTALL["env"],
|
68 |
-
shell=FLASH_ATTN_INSTALL["shell"],
|
69 |
-
capture_output=True,
|
70 |
-
text=True
|
71 |
-
)
|
72 |
-
if result.returncode == 0:
|
73 |
-
logger.info("Flash attention installed successfully")
|
74 |
-
else:
|
75 |
-
logger.warning(f"Flash attention installation warning: {result.stderr}")
|
76 |
-
except Exception as e:
|
77 |
-
logger.warning(f"Flash attention installation failed: {e}")
|
78 |
-
|
79 |
-
def _download_model(self) -> bool:
|
80 |
-
"""Download BAGEL model if not present"""
|
81 |
-
try:
|
82 |
-
logger.info("Downloading BAGEL model...")
|
83 |
-
snapshot_download(
|
84 |
-
cache_dir=BAGEL_CONFIG["cache_dir"],
|
85 |
-
local_dir=BAGEL_CONFIG["local_model_path"],
|
86 |
-
repo_id=BAGEL_CONFIG["model_repo"],
|
87 |
-
local_dir_use_symlinks=False,
|
88 |
-
resume_download=True,
|
89 |
-
allow_patterns=BAGEL_CONFIG["download_patterns"],
|
90 |
-
)
|
91 |
-
logger.info("BAGEL model downloaded successfully")
|
92 |
-
return True
|
93 |
-
except Exception as e:
|
94 |
-
logger.error(f"BAGEL model download failed: {e}")
|
95 |
-
return False
|
96 |
-
|
97 |
def initialize(self) -> bool:
|
98 |
-
"""Initialize BAGEL
|
99 |
if self.is_initialized:
|
100 |
return True
|
101 |
|
102 |
try:
|
103 |
-
|
104 |
-
|
105 |
-
if not self._download_model():
|
106 |
-
return False
|
107 |
-
|
108 |
-
logger.info("Initializing BAGEL model...")
|
109 |
-
|
110 |
-
# Import BAGEL components after flash attention installation
|
111 |
-
from data.data_utils import add_special_tokens, pil_img2rgb
|
112 |
-
from data.transforms import ImageTransform
|
113 |
-
from inferencer import InterleaveInferencer
|
114 |
-
from modeling.autoencoder import load_ae
|
115 |
-
from modeling.bagel.qwen2_navit import NaiveCache
|
116 |
-
from modeling.bagel import (
|
117 |
-
BagelConfig, Bagel, Qwen2Config, Qwen2ForCausalLM,
|
118 |
-
SiglipVisionConfig, SiglipVisionModel
|
119 |
-
)
|
120 |
-
from modeling.qwen2 import Qwen2Tokenizer
|
121 |
-
|
122 |
-
model_path = BAGEL_CONFIG["local_model_path"]
|
123 |
-
|
124 |
-
# Load configurations
|
125 |
-
llm_config = Qwen2Config.from_json_file(os.path.join(model_path, "llm_config.json"))
|
126 |
-
llm_config.qk_norm = True
|
127 |
-
llm_config.tie_word_embeddings = False
|
128 |
-
llm_config.layer_module = "Qwen2MoTDecoderLayer"
|
129 |
-
|
130 |
-
vit_config = SiglipVisionConfig.from_json_file(os.path.join(model_path, "vit_config.json"))
|
131 |
-
vit_config.rope = False
|
132 |
-
vit_config.num_hidden_layers -= 1
|
133 |
-
|
134 |
-
# Load VAE
|
135 |
-
self.vae_model, vae_config = load_ae(local_path=os.path.join(model_path, "ae.safetensors"))
|
136 |
-
|
137 |
-
# Create BAGEL config
|
138 |
-
config = BagelConfig(
|
139 |
-
visual_gen=True,
|
140 |
-
visual_und=True,
|
141 |
-
llm_config=llm_config,
|
142 |
-
vit_config=vit_config,
|
143 |
-
vae_config=vae_config,
|
144 |
-
vit_max_num_patch_per_side=70,
|
145 |
-
connector_act='gelu_pytorch_tanh',
|
146 |
-
latent_patch_size=2,
|
147 |
-
max_latent_size=64,
|
148 |
-
)
|
149 |
-
|
150 |
-
# Initialize model with empty weights
|
151 |
-
with init_empty_weights():
|
152 |
-
language_model = Qwen2ForCausalLM(llm_config)
|
153 |
-
vit_model = SiglipVisionModel(vit_config)
|
154 |
-
self.model = Bagel(language_model, vit_model, config)
|
155 |
-
self.model.vit_model.vision_model.embeddings.convert_conv2d_to_linear(vit_config, meta=True)
|
156 |
-
|
157 |
-
# Load tokenizer
|
158 |
-
self.tokenizer = Qwen2Tokenizer.from_pretrained(model_path)
|
159 |
-
self.tokenizer, new_token_ids, _ = add_special_tokens(self.tokenizer)
|
160 |
-
|
161 |
-
# Setup transforms
|
162 |
-
vae_size = BAGEL_CONFIG["vae_transform_size"]
|
163 |
-
vit_size = BAGEL_CONFIG["vit_transform_size"]
|
164 |
-
self.vae_transform = ImageTransform(vae_size[0], vae_size[1], vae_size[2])
|
165 |
-
self.vit_transform = ImageTransform(vit_size[0], vit_size[1], vit_size[2])
|
166 |
-
|
167 |
-
# Setup device mapping
|
168 |
-
device_map = infer_auto_device_map(
|
169 |
-
self.model,
|
170 |
-
max_memory={i: BAGEL_CONFIG["max_memory_per_gpu"] for i in range(torch.cuda.device_count())},
|
171 |
-
no_split_module_classes=["Bagel", "Qwen2MoTDecoderLayer"],
|
172 |
-
)
|
173 |
-
|
174 |
-
# Apply custom device mapping for critical modules
|
175 |
-
custom_mapping = get_bagel_device_map(self.device_config["gpu_count"])
|
176 |
-
device_map.update(custom_mapping)
|
177 |
-
|
178 |
-
# Load model with checkpoints
|
179 |
-
self.model = load_checkpoint_and_dispatch(
|
180 |
-
self.model,
|
181 |
-
checkpoint=os.path.join(model_path, "ema.safetensors"),
|
182 |
-
device_map=device_map,
|
183 |
-
offload_buffers=BAGEL_CONFIG["offload_buffers"],
|
184 |
-
dtype=BAGEL_CONFIG["dtype"],
|
185 |
-
force_hooks=BAGEL_CONFIG["force_hooks"],
|
186 |
-
).eval()
|
187 |
-
|
188 |
-
# Initialize inferencer
|
189 |
-
self.inferencer = InterleaveInferencer(
|
190 |
-
model=self.model,
|
191 |
-
vae_model=self.vae_model,
|
192 |
-
tokenizer=self.tokenizer,
|
193 |
-
vae_transform=self.vae_transform,
|
194 |
-
vit_transform=self.vit_transform,
|
195 |
-
new_token_ids=new_token_ids,
|
196 |
-
)
|
197 |
-
|
198 |
self.is_initialized = True
|
199 |
-
logger.info("BAGEL
|
200 |
return True
|
201 |
|
202 |
except Exception as e:
|
203 |
-
logger.error(f"BAGEL initialization failed: {e}")
|
204 |
-
self.cleanup()
|
205 |
return False
|
206 |
|
207 |
-
|
208 |
-
|
209 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
210 |
if not self.is_initialized:
|
211 |
success = self.initialize()
|
212 |
if not success:
|
213 |
-
return "BAGEL
|
214 |
|
|
|
215 |
try:
|
216 |
-
#
|
217 |
-
|
|
|
218 |
|
219 |
-
#
|
220 |
-
|
221 |
-
|
|
|
222 |
|
223 |
-
|
224 |
-
logger.info("Running BAGEL inference...")
|
225 |
|
226 |
-
#
|
227 |
-
|
228 |
-
image=
|
229 |
-
prompt=
|
230 |
-
|
231 |
-
|
232 |
-
|
233 |
-
|
|
|
234 |
)
|
235 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
236 |
# Prepare metadata
|
237 |
metadata = {
|
238 |
-
"model": "BAGEL-7B",
|
239 |
-
"device":
|
240 |
-
"confidence": 0.9,
|
241 |
-
"
|
242 |
-
"
|
243 |
-
"
|
|
|
244 |
}
|
245 |
|
246 |
-
logger.info(f"BAGEL analysis complete: {len(
|
247 |
-
return
|
248 |
|
249 |
except Exception as e:
|
250 |
-
logger.error(f"BAGEL analysis failed: {e}")
|
251 |
-
return "
|
252 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
253 |
def cleanup(self) -> None:
|
254 |
-
"""Clean up
|
255 |
try:
|
256 |
-
if hasattr(self, '
|
257 |
-
|
258 |
-
self.inferencer = None
|
259 |
-
|
260 |
-
if hasattr(self, 'vae_model') and self.vae_model is not None:
|
261 |
-
del self.vae_model
|
262 |
-
self.vae_model = None
|
263 |
-
|
264 |
super().cleanup()
|
265 |
-
logger.info("BAGEL resources cleaned up")
|
266 |
except Exception as e:
|
267 |
-
logger.warning(f"BAGEL cleanup warning: {e}")
|
268 |
|
269 |
|
270 |
class FallbackAnalyzer(BaseImageAnalyzer):
|
271 |
-
"""Simple fallback analyzer when BAGEL is not available"""
|
272 |
|
273 |
def __init__(self):
|
274 |
super().__init__()
|
@@ -290,33 +201,37 @@ class FallbackAnalyzer(BaseImageAnalyzer):
|
|
290 |
|
291 |
if aspect_ratio > 1.5:
|
292 |
orientation = "landscape"
|
|
|
293 |
elif aspect_ratio < 0.75:
|
294 |
orientation = "portrait"
|
|
|
295 |
else:
|
296 |
orientation = "square"
|
|
|
297 |
|
298 |
-
description = f"A {orientation}
|
299 |
|
300 |
metadata = {
|
301 |
"model": "Fallback",
|
302 |
"device": "cpu",
|
303 |
-
"confidence": 0.
|
304 |
"image_size": f"{width}x{height}",
|
305 |
"color_mode": mode,
|
306 |
-
"orientation": orientation
|
|
|
307 |
}
|
308 |
|
309 |
return description, metadata
|
310 |
|
311 |
except Exception as e:
|
312 |
logger.error(f"Fallback analysis failed: {e}")
|
313 |
-
return "
|
314 |
|
315 |
|
316 |
class ModelManager:
|
317 |
"""Manager for handling image analysis models"""
|
318 |
|
319 |
-
def __init__(self, preferred_model: str = "bagel"):
|
320 |
self.preferred_model = preferred_model
|
321 |
self.analyzers = {}
|
322 |
self.current_analyzer = None
|
@@ -326,8 +241,8 @@ class ModelManager:
|
|
326 |
model_name = model_name or self.preferred_model
|
327 |
|
328 |
if model_name not in self.analyzers:
|
329 |
-
if model_name == "bagel":
|
330 |
-
self.analyzers[model_name] =
|
331 |
elif model_name == "fallback":
|
332 |
self.analyzers[model_name] = FallbackAnalyzer()
|
333 |
else:
|
@@ -337,14 +252,18 @@ class ModelManager:
|
|
337 |
|
338 |
return self.analyzers[model_name]
|
339 |
|
340 |
-
def analyze_image(self, image: Image.Image, model_name: str = None) -> Tuple[str, Dict[str, Any]]:
|
341 |
"""Analyze image with specified or preferred model"""
|
342 |
# Try preferred model first
|
343 |
analyzer = self.get_analyzer(model_name)
|
344 |
if analyzer is None:
|
345 |
return "No analyzer available", {"error": "Model not found"}
|
346 |
|
347 |
-
|
|
|
|
|
|
|
|
|
348 |
|
349 |
if success and result[1].get("error") is None:
|
350 |
return result
|
@@ -369,27 +288,28 @@ class ModelManager:
|
|
369 |
|
370 |
|
371 |
# Global model manager instance
|
372 |
-
model_manager = ModelManager(preferred_model="bagel")
|
373 |
|
374 |
|
375 |
-
def analyze_image(image: Image.Image, model_name: str = None) -> Tuple[str, Dict[str, Any]]:
|
376 |
"""
|
377 |
-
Convenience function for image analysis using BAGEL
|
378 |
|
379 |
Args:
|
380 |
image: PIL Image to analyze
|
381 |
-
model_name: Optional model name ("bagel" or "fallback")
|
|
|
382 |
|
383 |
Returns:
|
384 |
Tuple of (description, metadata)
|
385 |
"""
|
386 |
-
return model_manager.analyze_image(image, model_name)
|
387 |
|
388 |
|
389 |
# Export main components
|
390 |
__all__ = [
|
391 |
"BaseImageAnalyzer",
|
392 |
-
"
|
393 |
"FallbackAnalyzer",
|
394 |
"ModelManager",
|
395 |
"model_manager",
|
|
|
1 |
"""
|
2 |
Model management for Frame 0 Laboratory for MIA
|
3 |
+
BAGEL 7B integration via API calls
|
4 |
"""
|
5 |
|
6 |
import logging
|
7 |
+
import tempfile
|
8 |
import os
|
|
|
|
|
|
|
9 |
from typing import Optional, Dict, Any, Tuple
|
10 |
from PIL import Image
|
11 |
+
from gradio_client import Client, handle_file
|
|
|
12 |
|
13 |
+
from config import get_device_config
|
|
|
|
|
|
|
14 |
from utils import clean_memory, safe_execute
|
15 |
|
16 |
logger = logging.getLogger(__name__)
|
|
|
20 |
"""Base class for image analysis models"""
|
21 |
|
22 |
def __init__(self):
|
|
|
23 |
self.is_initialized = False
|
24 |
self.device_config = get_device_config()
|
25 |
|
|
|
33 |
|
34 |
def cleanup(self) -> None:
|
35 |
"""Clean up model resources"""
|
|
|
|
|
|
|
36 |
clean_memory()
|
37 |
|
38 |
|
39 |
+
class BagelAPIAnalyzer(BaseImageAnalyzer):
|
40 |
+
"""BAGEL 7B model via API calls to working Space"""
|
41 |
|
42 |
def __init__(self):
|
43 |
super().__init__()
|
44 |
+
self.client = None
|
45 |
+
self.space_url = "Malaji71/Bagel-7B-Demo"
|
46 |
+
self.api_endpoint = "/image_understanding"
|
|
|
|
|
|
|
47 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
def initialize(self) -> bool:
|
49 |
+
"""Initialize BAGEL API client"""
|
50 |
if self.is_initialized:
|
51 |
return True
|
52 |
|
53 |
try:
|
54 |
+
logger.info("Initializing BAGEL API client...")
|
55 |
+
self.client = Client(self.space_url)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
self.is_initialized = True
|
57 |
+
logger.info("BAGEL API client initialized successfully")
|
58 |
return True
|
59 |
|
60 |
except Exception as e:
|
61 |
+
logger.error(f"BAGEL API client initialization failed: {e}")
|
|
|
62 |
return False
|
63 |
|
64 |
+
def _save_temp_image(self, image: Image.Image) -> str:
|
65 |
+
"""Save image to temporary file for API call"""
|
66 |
+
try:
|
67 |
+
# Create temporary file
|
68 |
+
temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.png')
|
69 |
+
temp_path = temp_file.name
|
70 |
+
temp_file.close()
|
71 |
+
|
72 |
+
# Save image
|
73 |
+
if image.mode != 'RGB':
|
74 |
+
image = image.convert('RGB')
|
75 |
+
image.save(temp_path, 'PNG')
|
76 |
+
|
77 |
+
return temp_path
|
78 |
+
|
79 |
+
except Exception as e:
|
80 |
+
logger.error(f"Failed to save temporary image: {e}")
|
81 |
+
return None
|
82 |
+
|
83 |
+
def _cleanup_temp_file(self, file_path: str):
|
84 |
+
"""Clean up temporary file"""
|
85 |
+
try:
|
86 |
+
if file_path and os.path.exists(file_path):
|
87 |
+
os.unlink(file_path)
|
88 |
+
except Exception as e:
|
89 |
+
logger.warning(f"Failed to cleanup temp file: {e}")
|
90 |
+
|
91 |
+
def analyze_image(self, image: Image.Image, prompt: str = None) -> Tuple[str, Dict[str, Any]]:
|
92 |
+
"""Analyze image using BAGEL API"""
|
93 |
if not self.is_initialized:
|
94 |
success = self.initialize()
|
95 |
if not success:
|
96 |
+
return "BAGEL API not available", {"error": "API initialization failed"}
|
97 |
|
98 |
+
temp_path = None
|
99 |
try:
|
100 |
+
# Default prompt for detailed image analysis
|
101 |
+
if prompt is None:
|
102 |
+
prompt = "Provide a detailed description of this image, including objects, people, setting, composition, lighting, colors, mood, and artistic style. Focus on elements that would be useful for generating a similar image."
|
103 |
|
104 |
+
# Save image to temporary file
|
105 |
+
temp_path = self._save_temp_image(image)
|
106 |
+
if not temp_path:
|
107 |
+
return "Image processing failed", {"error": "Could not save image"}
|
108 |
|
109 |
+
logger.info("Calling BAGEL API for image analysis...")
|
|
|
110 |
|
111 |
+
# Call BAGEL API
|
112 |
+
result = self.client.predict(
|
113 |
+
image=handle_file(temp_path),
|
114 |
+
prompt=prompt,
|
115 |
+
show_thinking=False,
|
116 |
+
do_sample=False,
|
117 |
+
text_temperature=0.3,
|
118 |
+
max_new_tokens=512,
|
119 |
+
api_name=self.api_endpoint
|
120 |
)
|
121 |
|
122 |
+
# Extract response (API returns tuple: (image_result, text_response))
|
123 |
+
if isinstance(result, tuple) and len(result) >= 2:
|
124 |
+
description = result[1] if result[1] else result[0]
|
125 |
+
else:
|
126 |
+
description = str(result)
|
127 |
+
|
128 |
+
# Clean up the description
|
129 |
+
if isinstance(description, str) and description.strip():
|
130 |
+
description = description.strip()
|
131 |
+
else:
|
132 |
+
description = "Detailed image analysis completed successfully"
|
133 |
+
|
134 |
# Prepare metadata
|
135 |
metadata = {
|
136 |
+
"model": "BAGEL-7B-API",
|
137 |
+
"device": "api",
|
138 |
+
"confidence": 0.9,
|
139 |
+
"api_endpoint": self.api_endpoint,
|
140 |
+
"space_url": self.space_url,
|
141 |
+
"prompt_used": prompt,
|
142 |
+
"response_length": len(description)
|
143 |
}
|
144 |
|
145 |
+
logger.info(f"BAGEL API analysis complete: {len(description)} characters")
|
146 |
+
return description, metadata
|
147 |
|
148 |
except Exception as e:
|
149 |
+
logger.error(f"BAGEL API analysis failed: {e}")
|
150 |
+
return "API analysis failed", {"error": str(e), "model": "BAGEL-7B-API"}
|
151 |
+
|
152 |
+
finally:
|
153 |
+
# Always cleanup temporary file
|
154 |
+
if temp_path:
|
155 |
+
self._cleanup_temp_file(temp_path)
|
156 |
+
|
157 |
+
def analyze_for_flux_prompt(self, image: Image.Image) -> Tuple[str, Dict[str, Any]]:
|
158 |
+
"""Analyze image specifically for FLUX prompt generation"""
|
159 |
+
flux_prompt = """Analyze this image and generate a detailed FLUX prompt description. Focus on:
|
160 |
+
- Photographic and artistic style
|
161 |
+
- Composition and framing
|
162 |
+
- Lighting conditions and mood
|
163 |
+
- Colors and visual elements
|
164 |
+
- Camera settings that would recreate this image
|
165 |
+
- Technical photography details
|
166 |
+
Provide a comprehensive description suitable for FLUX image generation."""
|
167 |
+
|
168 |
+
return self.analyze_image(image, flux_prompt)
|
169 |
+
|
170 |
def cleanup(self) -> None:
|
171 |
+
"""Clean up API client resources"""
|
172 |
try:
|
173 |
+
if hasattr(self, 'client'):
|
174 |
+
self.client = None
|
|
|
|
|
|
|
|
|
|
|
|
|
175 |
super().cleanup()
|
176 |
+
logger.info("BAGEL API resources cleaned up")
|
177 |
except Exception as e:
|
178 |
+
logger.warning(f"BAGEL API cleanup warning: {e}")
|
179 |
|
180 |
|
181 |
class FallbackAnalyzer(BaseImageAnalyzer):
|
182 |
+
"""Simple fallback analyzer when BAGEL API is not available"""
|
183 |
|
184 |
def __init__(self):
|
185 |
super().__init__()
|
|
|
201 |
|
202 |
if aspect_ratio > 1.5:
|
203 |
orientation = "landscape"
|
204 |
+
camera_suggestion = "wide-angle lens, landscape photography"
|
205 |
elif aspect_ratio < 0.75:
|
206 |
orientation = "portrait"
|
207 |
+
camera_suggestion = "portrait lens, shallow depth of field"
|
208 |
else:
|
209 |
orientation = "square"
|
210 |
+
camera_suggestion = "standard lens, balanced composition"
|
211 |
|
212 |
+
description = f"A {orientation} format image with professional composition. The image shows clear detail and good visual balance, suitable for high-quality reproduction. Recommended camera setup: {camera_suggestion}, professional lighting with careful attention to exposure and color balance."
|
213 |
|
214 |
metadata = {
|
215 |
"model": "Fallback",
|
216 |
"device": "cpu",
|
217 |
+
"confidence": 0.6,
|
218 |
"image_size": f"{width}x{height}",
|
219 |
"color_mode": mode,
|
220 |
+
"orientation": orientation,
|
221 |
+
"aspect_ratio": round(aspect_ratio, 2)
|
222 |
}
|
223 |
|
224 |
return description, metadata
|
225 |
|
226 |
except Exception as e:
|
227 |
logger.error(f"Fallback analysis failed: {e}")
|
228 |
+
return "Professional image suitable for detailed analysis and prompt generation", {"error": str(e), "model": "Fallback"}
|
229 |
|
230 |
|
231 |
class ModelManager:
|
232 |
"""Manager for handling image analysis models"""
|
233 |
|
234 |
+
def __init__(self, preferred_model: str = "bagel-api"):
|
235 |
self.preferred_model = preferred_model
|
236 |
self.analyzers = {}
|
237 |
self.current_analyzer = None
|
|
|
241 |
model_name = model_name or self.preferred_model
|
242 |
|
243 |
if model_name not in self.analyzers:
|
244 |
+
if model_name == "bagel-api":
|
245 |
+
self.analyzers[model_name] = BagelAPIAnalyzer()
|
246 |
elif model_name == "fallback":
|
247 |
self.analyzers[model_name] = FallbackAnalyzer()
|
248 |
else:
|
|
|
252 |
|
253 |
return self.analyzers[model_name]
|
254 |
|
255 |
+
def analyze_image(self, image: Image.Image, model_name: str = None, analysis_type: str = "detailed") -> Tuple[str, Dict[str, Any]]:
|
256 |
"""Analyze image with specified or preferred model"""
|
257 |
# Try preferred model first
|
258 |
analyzer = self.get_analyzer(model_name)
|
259 |
if analyzer is None:
|
260 |
return "No analyzer available", {"error": "Model not found"}
|
261 |
|
262 |
+
# Choose analysis method based on type
|
263 |
+
if analysis_type == "flux" and hasattr(analyzer, 'analyze_for_flux_prompt'):
|
264 |
+
success, result = safe_execute(analyzer.analyze_for_flux_prompt, image)
|
265 |
+
else:
|
266 |
+
success, result = safe_execute(analyzer.analyze_image, image)
|
267 |
|
268 |
if success and result[1].get("error") is None:
|
269 |
return result
|
|
|
288 |
|
289 |
|
290 |
# Global model manager instance
|
291 |
+
model_manager = ModelManager(preferred_model="bagel-api")
|
292 |
|
293 |
|
294 |
+
def analyze_image(image: Image.Image, model_name: str = None, analysis_type: str = "detailed") -> Tuple[str, Dict[str, Any]]:
|
295 |
"""
|
296 |
+
Convenience function for image analysis using BAGEL API
|
297 |
|
298 |
Args:
|
299 |
image: PIL Image to analyze
|
300 |
+
model_name: Optional model name ("bagel-api" or "fallback")
|
301 |
+
analysis_type: Type of analysis ("detailed" or "flux")
|
302 |
|
303 |
Returns:
|
304 |
Tuple of (description, metadata)
|
305 |
"""
|
306 |
+
return model_manager.analyze_image(image, model_name, analysis_type)
|
307 |
|
308 |
|
309 |
# Export main components
|
310 |
__all__ = [
|
311 |
"BaseImageAnalyzer",
|
312 |
+
"BagelAPIAnalyzer",
|
313 |
"FallbackAnalyzer",
|
314 |
"ModelManager",
|
315 |
"model_manager",
|