File size: 19,481 Bytes
6cee168
7b04190
f04ae31
 
 
7133210
 
f04ae31
 
 
7b04190
8173477
579d7f4
c90ce6e
3e40092
b1d5c20
 
c90ce6e
05ae2ff
7133210
b1d5c20
 
 
6cee168
7133210
c90ce6e
7133210
 
 
 
c90ce6e
 
7133210
 
 
 
 
b1d5c20
c90ce6e
6cee168
c90ce6e
7133210
05ae2ff
 
dbf6dc8
05ae2ff
 
 
7133210
c90ce6e
7133210
c90ce6e
 
 
6cee168
c90ce6e
a5a2a81
b1d5c20
 
c90ce6e
 
 
 
 
 
7133210
 
 
a5a2a81
6cee168
 
a5a2a81
7133210
a5a2a81
 
 
 
 
6cee168
 
7133210
05bcc3b
579d7f4
 
 
 
 
05bcc3b
56c38a0
7133210
f3e0682
b1d5c20
05ae2ff
6cee168
 
 
 
 
 
 
 
7133210
71f2456
56c38a0
a5a2a81
dbf6dc8
a5a2a81
56c38a0
2777fd6
7133210
 
 
 
 
2777fd6
b1d5c20
 
 
22fba5f
b1d5c20
 
 
 
7133210
b1d5c20
 
686dbcb
 
f04ae31
56c38a0
b1d5c20
f04ae31
56c38a0
b1d5c20
579d7f4
7133210
 
 
 
 
 
579d7f4
 
 
 
 
b1d5c20
579d7f4
7133210
579d7f4
b1d5c20
7133210
 
2777fd6
7133210
2777fd6
dbf6dc8
7133210
dbf6dc8
7133210
f04ae31
dbf6dc8
 
 
7133210
dbf6dc8
 
 
 
7133210
 
 
 
dbf6dc8
579d7f4
 
 
 
 
 
 
 
7133210
579d7f4
 
 
 
 
 
 
 
f3e0682
f04ae31
579d7f4
 
 
 
 
f04ae31
579d7f4
 
7133210
579d7f4
 
 
 
05ae2ff
579d7f4
7133210
579d7f4
6cee168
 
b1d5c20
 
 
 
 
6cee168
b1d5c20
 
 
 
f04ae31
 
dbf6dc8
05ae2ff
579d7f4
f04ae31
b1d5c20
05ae2ff
 
b1d5c20
 
f04ae31
 
dbf6dc8
7133210
f04ae31
7133210
b1d5c20
6cee168
 
f04ae31
7133210
6cee168
b1d5c20
6cee168
b1d5c20
f04ae31
7133210
 
 
 
 
 
 
 
 
 
 
 
f04ae31
7133210
b1d5c20
7133210
f04ae31
7133210
6cee168
7133210
 
 
 
 
46457a7
7133210
f04ae31
7133210
 
 
 
 
f04ae31
7133210
 
 
579d7f4
f04ae31
7133210
dbf6dc8
 
f04ae31
 
b1d5c20
91e44e7
56c38a0
dbf6dc8
 
 
 
91e44e7
dbf6dc8
f04ae31
7133210
 
 
 
f04ae31
 
91e44e7
 
 
05ae2ff
f04ae31
 
 
 
 
 
 
 
 
 
91e44e7
f04ae31
91e44e7
f04ae31
7133210
6cee168
05ae2ff
 
dbf6dc8
05ae2ff
 
7133210
 
b1d5c20
6cee168
b1d5c20
f04ae31
7133210
3a8fcc6
7133210
 
 
 
 
 
 
3a8fcc6
7133210
 
 
 
 
f04ae31
b1d5c20
6cee168
7133210
 
b1d5c20
7133210
6cee168
5dd7ad6
 
 
 
 
6cee168
b1d5c20
6cee168
22fba5f
6cee168
b1d5c20
 
f04ae31
686dbcb
 
f04ae31
 
 
dbf6dc8
 
f04ae31
dbf6dc8
 
b1d5c20
686dbcb
 
f04ae31
6cee168
b1d5c20
 
 
 
dbf6dc8
b1d5c20
dbf6dc8
b1d5c20
6cee168
 
b1d5c20
6cee168
 
5dd7ad6
6cee168
 
 
05ae2ff
5dd7ad6
6cee168
 
5dd7ad6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6cee168
 
b1d5c20
6cee168
 
b1d5c20
 
6cee168
a5a2a81
7133210
 
b1d5c20
 
 
 
 
7133210
dbf6dc8
b1d5c20
 
 
 
 
a5a2a81
22fba5f
c90ce6e
7133210
c90ce6e
a5a2a81
b1d5c20
12fd62f
56c38a0
12fd62f
 
f04ae31
6cee168
56c38a0
6cee168
 
56c38a0
f04ae31
56c38a0
22fba5f
 
 
 
56c38a0
2777fd6
a5a2a81
 
7133210
a5a2a81
 
7133210
 
17e4f3c
7133210
17e4f3c
 
12fd62f
f04ae31
17e4f3c
c35985e
6cee168
 
f04ae31
17e4f3c
 
f04ae31
17e4f3c
 
 
a5a2a81
f04ae31
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
# -*- coding: utf-8 -*-
"""
PULSE ECG Handler — Deterministik Versiyon
- Üretim ayarları: do_sample=False (Tutarlı çıktı), temperature/top_p etkisiz
- Stopping: Konuşma ayırıcıda (conv.sep/sep2) güvenli token-eşleşmeli kriter
- Görsel tensörü: .half() ve model cihazında
- Streamer: TextIteratorStreamer (demo gibi), thread ile generate
- Seed/deterministic KAPALI (do_sample=False ile determinizm sağlanır)
- STYLE_HINT: demo üslubuna yaklaşmak için
- Post-process: YALNIZCA whitespace/biçim normalizasyonu
"""
import os
import re
import json
import base64
import hashlib
import datetime
from io import BytesIO
from threading import Thread
from typing import Optional, Union
import torch
from PIL import Image
import requests

# ====== LLaVA & Transformers ======
try:
    from llava.constants import (
        IMAGE_TOKEN_INDEX,
        DEFAULT_IMAGE_TOKEN,
    )
    from llava.conversation import conv_templates, SeparatorStyle
    from llava.model.builder import load_pretrained_model
    from llava.mm_utils import (
        tokenizer_image_token,
        process_images,
        get_model_name_from_path,
    )
    from llava.utils import disable_torch_init
    LLAVA_AVAILABLE = True
except Exception as e:
    LLAVA_AVAILABLE = False
    print(f"[WARN] LLaVA not available: {e}")

try:
    from transformers import TextIteratorStreamer, StoppingCriteria
    TRANSFORMERS_AVAILABLE = True
except Exception as e:
    TRANSFORMERS_AVAILABLE = False
    print(f"[WARN] transformers not available: {e}")

# ====== HF Hub logging (opsiyonel) ======
try:
    from huggingface_hub import HfApi, login
    HF_HUB_AVAILABLE = True
except Exception:
    HF_HUB_AVAILABLE = False

api = None
repo_name = ""
if HF_HUB_AVAILABLE and "HF_TOKEN" in os.environ:
    try:
        login(token=os.environ["HF_TOKEN"], write_permission=True)
        api = HfApi()
        repo_name = os.environ.get("LOG_REPO", "")
    except Exception as e:
        print(f"[HF Hub] init failed: {e}")
        api = None
        repo_name = ""

LOGDIR = "./logs"
os.makedirs(LOGDIR, exist_ok=True)

# ====== Global State ======
tokenizer = None
model = None
image_processor = None
context_len = None
args = None
model_initialized = False

# ====== Style Hint (demo benzeri üslup) ======
STYLE_HINT = (
    "Write one concise narrative paragraph that covers rhythm, heart rate, cardiac axis, "
    "P waves and PR interval, QRS morphology and duration, ST segments, T waves, and QT/QTc. "
    "Use neutral, factual cardiology language. Avoid headings and bullet points. "
    "Finish with a single final line starting exactly with 'Structured clinical impression:' "
    "followed by a succinct, comma-separated summary of the key diagnoses."
)

# ===================== Utilities =====================

def _safe_upload(path: str):
    if api and repo_name and path and os.path.isfile(path):
        try:
            api.upload_file(
                path_or_fileobj=path,
                path_in_repo=path.replace("./logs/", ""),
                repo_id=repo_name,
                repo_type="dataset",
            )
        except Exception as e:
            print(f"[upload] failed for {path}: {e}")

def _conv_log_path() -> str:
    t = datetime.datetime.now()
    return os.path.join(LOGDIR, f"{t.year:04d}-{t.month:02d}-{t.day:02d}-user_conv.json")

def load_image_any(image_input: Union[str, dict]) -> Image.Image:
    """
    Desteklenen:
      - URL (http/https)
      - yerel dosya yolu
      - base64 (opsiyonel data URL prefix ile)
      - {"image": <base64|dataurl>}
    """
    if isinstance(image_input, str):
        s = image_input.strip()
        if s.startswith(("http://", "https://")):
            r = requests.get(s, timeout=(5, 20))
            r.raise_for_status()
            return Image.open(BytesIO(r.content)).convert("RGB")
        if os.path.exists(s):
            return Image.open(s).convert("RGB")
        # base64 (dataurl olabilir)
        if s.startswith("data:image"):
            s = s.split(",", 1)[1]
        raw = base64.b64decode(s)
        return Image.open(BytesIO(raw)).convert("RGB")
    
    if isinstance(image_input, dict) and "image" in image_input:
        return load_image_any(image_input["image"])
    
    raise ValueError("Unsupported image input format")

def _normalize_whitespace(text: str) -> str:
    """
    Gereksiz boşluk/boş satırları toparlar:
    - Satır başı/sonu boşluklarını siler
    - Birden çok boşluğu tek boşluğa indirger
    - 3+ boş satırı 1 boş satıra indirger
    """
    text = text.replace("\r\n", "\n").replace("\r", "\n")
    lines = [re.sub(r"[ \t]+", " ", ln.strip()) for ln in text.split("\n")]
    text = "\n".join(lines).strip()
    text = re.sub(r"\n{3,}", "\n\n", text)
    return text

def _postprocess_min(text: str) -> str:
    # Yalnızca whitespace/biçim temizliği
    return _normalize_whitespace(text)

# ====== Güvenli Stop Kriteri (conv separator) ======
class SafeKeywordsStoppingCriteria(StoppingCriteria):
    """
    conv.sep/sep2 bazlı token eşleşmesi; tensör → bool hatası yok.
    """
    def __init__(self, keyword: str, tokenizer):
        self.tokenizer = tokenizer
        tok = tokenizer(keyword, add_special_tokens=False, return_tensors="pt").input_ids[0]
        self.kw_ids = tok  # shape: (n,)
    
    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
        if input_ids is None or input_ids.shape[0] == 0:
            return False
        out = input_ids[0]  # assume bsz=1
        n = self.kw_ids.shape[0]
        if out.shape[0] < n:
            return False
        tail = out[-n:]
        kw = self.kw_ids.to(tail.device)
        return torch.equal(tail, kw)

# ===================== Core Generation =====================

class InferenceDemo:
    def __init__(self, args, model_path, tokenizer_, model_, image_processor_, context_len_):
        if not LLAVA_AVAILABLE:
            raise ImportError("LLaVA not available")
        disable_torch_init()
        self.tokenizer, self.model, self.image_processor, self.context_len = (
            tokenizer_, model_, image_processor_, context_len_
        )
        # Parite için sabit şablon
        self.conv_mode = "llava_v1"
        self.conversation = conv_templates[self.conv_mode].copy()
        self.num_frames = getattr(args, "num_frames", 16)

class ChatSessionManager:
    def __init__(self):
        self.chatbot = None
        self.args = None
        self.model_path = None
        
    def init_if_needed(self, args, model_path, tokenizer, model, image_processor, context_len):
        if self.chatbot is None:
            self.args = args
            self.model_path = model_path
            self.chatbot = InferenceDemo(args, model_path, tokenizer, model, image_processor, context_len)
            
    def get_chatbot(self, args, model_path, tokenizer, model, image_processor, context_len):
        self.init_if_needed(args, model_path, tokenizer, model, image_processor, context_len)
        # Her çağrıda taze template (demo gibi yeni tur)
        self.chatbot.conversation = conv_templates[self.chatbot.conv_mode].copy()
        return self.chatbot

chat_manager = ChatSessionManager()

def _build_prompt_and_ids(chatbot, user_text: str, device: torch.device):
    # DEMO PARİTE: sarım yok, tek görüntü için tek image token
    inp = f"{DEFAULT_IMAGE_TOKEN}\n{user_text}"
    chatbot.conversation.append_message(chatbot.conversation.roles[0], inp)
    chatbot.conversation.append_message(chatbot.conversation.roles[1], None)
    prompt = chatbot.conversation.get_prompt()
    input_ids = tokenizer_image_token(
        prompt, chatbot.tokenizer, IMAGE_TOKEN_INDEX, return_tensors="pt"
    ).unsqueeze(0).to(device)
    return prompt, input_ids

def generate_response(
    message_text: str,
    image_input,
    *,
    temperature: Optional[float] = None, # Deterministik modda yoksayılır
    top_p: Optional[float] = None,       # Deterministik modda yoksayılır
    max_new_tokens: Optional[int] = None,
    conv_mode_override: Optional[str] = None,
    repetition_penalty: Optional[float] = None,
    det_seed: Optional[int] = None,      # Deterministik modda yoksayılır
):
    if not (LLAVA_AVAILABLE and TRANSFORMERS_AVAILABLE):
        return {"error": "Required libraries not available (llava/transformers)"}
    if not message_text or image_input is None:
        return {"error": "Both 'message' and 'image' are required"}
    
    # Varsayılanlar
    if max_new_tokens is None: max_new_tokens = 4096
    if repetition_penalty is None: repetition_penalty = 1.0  # etkisiz
    
    # Chat session
    chatbot = chat_manager.get_chatbot(args, args.model_path, tokenizer, model, image_processor, context_len)
    if conv_mode_override and conv_mode_override in conv_templates:
        chatbot.conversation = conv_templates[conv_mode_override].copy()
        
    # Görüntü yükle
    try:
        pil_img = load_image_any(image_input)
    except Exception as e:
        return {"error": f"Failed to load image: {e}"}
        
    # Log için hash+path
    img_hash, img_path = "NA", None
    try:
        buf = BytesIO(); pil_img.save(buf, format="JPEG"); raw = buf.getvalue()
        img_hash = hashlib.md5(raw).hexdigest()
        t = datetime.datetime.now()
        img_path = os.path.join(LOGDIR, "serve_images", f"{t.year:04d}-{t.month:02d}-{t.day:02d}", f"{img_hash}.jpg")
        os.makedirs(os.path.dirname(img_path), exist_ok=True)
        if not os.path.isfile(img_path):
            pil_img.save(img_path)
    except Exception as e:
        print(f"[log] save image failed: {e}")
        
    # Cihaz/dtype
    device = next(chatbot.model.parameters()).device
    dtype = torch.float16  # demo: half
    
    # Görüntü ön-işleme → tensör
    try:
        processed = process_images([pil_img], chatbot.image_processor, chatbot.model.config)
        if isinstance(processed, (list, tuple)) and len(processed) > 0:
            image_tensor = processed[0]
        elif isinstance(processed, torch.Tensor):
            image_tensor = processed[0] if processed.ndim == 4 else processed
        else:
            return {"error": "Image processing returned empty"}
            
        if image_tensor.ndim == 3:
            image_tensor = image_tensor.unsqueeze(0)  # (1,C,H,W)
        image_tensor = image_tensor.to(device=device, dtype=dtype)  # demo: half + device
    except Exception as e:
        return {"error": f"Image processing failed: {e}"}
        
    # STYLE_HINT ekle ve prompt hazırla
    msg = (message_text or "").strip()
    msg = f"{msg}\n\n{STYLE_HINT}"
    _, input_ids = _build_prompt_and_ids(chatbot, msg, device)
    
    # Stop string (conv separator) → güvenli kriter
    stop_str = chatbot.conversation.sep if chatbot.conversation.sep_style != SeparatorStyle.TWO else chatbot.conversation.sep2
    stopping = SafeKeywordsStoppingCriteria(stop_str, chatbot.tokenizer)
    
    # Seed (do_sample=False olduğu için önemsiz, ancak kodda bırakılabilir)
    if det_seed is not None:
        try:
            s = int(det_seed)
            torch.manual_seed(s)
            if torch.cuda.is_available():
                torch.cuda.manual_seed(s)
                torch.cuda.manual_seed_all(s)
        except Exception:
            pass
            
    # Streamer (demo gibi)
    streamer = TextIteratorStreamer(
        chatbot.tokenizer, skip_prompt=True, skip_special_tokens=True
    )
    
    # Generate kwargs — Deterministik Ayarlar
    gen_kwargs = dict(
        inputs=input_ids,
        images=image_tensor,
        streamer=streamer,
        
        # 🟢 ÖNEMLİ DEĞİŞİKLİK: Deterministiği (Tutarlılığı) Aç
        do_sample=False,                     
        
        # temperature ve top_p ayarları artık yoksayılır
        # temperature=float(temperature), 
        # top_p=float(top_p),
        
        max_new_tokens=int(max_new_tokens), 
        repetition_penalty=float(repetition_penalty), 
        use_cache=False,
        stopping_criteria=[stopping],
    )
    
    # Üretim (arka thread) + akışı topla
    try:
        t = Thread(target=chatbot.model.generate, kwargs=gen_kwargs)
        t.start()
        chunks = []
        for piece in streamer:
            chunks.append(piece)
        text = "".join(chunks)
        text = _postprocess_min(text)   # yalnızca whitespace/format temizliği
        chatbot.conversation.messages[-1][-1] = text
    except Exception as e:
        return {"error": f"Generation failed: {e}"}
        
    # Log
    try:
        row = {
            "time": datetime.datetime.now().isoformat(),
            "type": "chat",
            "model": "PULSE-7B",
            "state": [(message_text, text)],
            "image_hash": img_hash,
            "image_path": img_path or "",
        }
        with open(_conv_log_path(), "a", encoding="utf-8") as f:
            f.write(json.dumps(row, ensure_ascii=False) + "\n")
        _safe_upload(_conv_log_path()); _safe_upload(img_path or "")
    except Exception as e:
        print(f"[log] failed: {e}")
        
    return {"status": "success", "response": text, "conversation_id": id(chatbot.conversation)}

# ===================== Public API =====================

def query(payload: dict):
    """HF Endpoint entry (demo-like)."""
    global model_initialized, tokenizer, model, image_processor, context_len, args

    # 🟢 Health check kısayolu: hem {"health_check": true} hem de {"message": "health_check"} desteklenir
    if payload.get("health_check") or payload.get("message") == "health_check":
        return health_check()
        
    if not model_initialized:
        if not initialize_model():
            return {"error": "Model initialization failed"}
        model_initialized = True
    try:
        message = payload.get("message") or payload.get("query") or payload.get("prompt") or payload.get("istem") or ""
        image   = payload.get("image") or payload.get("image_url") or payload.get("img") or None
        
        if not message.strip(): return {"error": "Missing 'message' text"}
        if image is None:       return {"error": "Missing 'image'. Use 'image', 'image_url', or 'img'."}
        
        # Deterministik modda temperature/top_p yoksayılır, ancak API uyumluluğu için tutulur
        temperature        = float(payload.get("temperature", 0.0))  # Default 0.0
        top_p              = float(payload.get("top_p", 1.0))
        max_new_tokens     = int(payload.get("max_output_tokens", payload.get("max_new_tokens", payload.get("max_tokens", 4096))))
        repetition_penalty = float(payload.get("repetition_penalty", 1.0)) 
        conv_mode_override = payload.get("conv_mode", None)
        det_seed           = payload.get("det_seed", None)
        if det_seed is not None:
            try: det_seed = int(det_seed)
            except Exception: det_seed = None
            
        return generate_response(
            message_text=message,
            image_input=image,
            temperature=temperature,
            top_p=top_p,
            max_new_tokens=max_new_tokens,
            conv_mode_override=conv_mode_override,
            repetition_penalty=repetition_penalty,
            det_seed=det_seed,
        )
    except Exception as e:
        return {"error": f"Query failed: {e}"}

def health_check():
    info = {
        "status": "healthy",
        "model_initialized": model_initialized,
        "llava_available": LLAVA_AVAILABLE,
        "transformers_available": TRANSFORMERS_AVAILABLE,
        "cuda_available": torch.cuda.is_available(),
    }

    if torch.cuda.is_available():
        try:
            device_index = torch.cuda.current_device()
            props = torch.cuda.get_device_properties(device_index)
            total_vram_gb = round(props.total_memory / (1024 ** 3), 2)
            used_vram_gb = round(torch.cuda.memory_allocated(device_index) / (1024 ** 3), 2)
            reserved_vram_gb = round(torch.cuda.memory_reserved(device_index) / (1024 ** 3), 2)

            info.update({
                "cuda_device_index": device_index,
                "cuda_name": props.name,
                "cuda_compute_capability": f"{props.major}.{props.minor}",
                "cuda_total_vram_gb": total_vram_gb,
                "cuda_used_vram_gb": used_vram_gb,
                "cuda_reserved_vram_gb": reserved_vram_gb,
                "torch_version": torch.__version__,
                "cuda_runtime_version": torch.version.cuda,
            })
        except Exception as e:
            info["cuda_error"] = str(e)

    return info


def get_model_info():
    if not model_initialized:
        return {"error": "Model not initialized"}
    return {
        "model_path": args.model_path if args else "Unknown",
        "context_len": context_len,
        "device": str(next(model.parameters()).device) if model else "Unknown",
    }

# ===================== Init & Session =====================

class _Args:
    def __init__(self):
        self.model_path = os.getenv("HF_MODEL_ID", "PULSE-ECG/PULSE-7B")
        self.model_base = None
        self.num_gpus   = int(os.getenv("NUM_GPUS", "1"))
        self.conv_mode  = "llava_v1"  # Parite için sabit
        self.max_new_tokens = int(os.getenv("MAX_NEW_TOKENS", "4096"))
        self.num_frames = 16
        self.load_8bit  = bool(int(os.getenv("LOAD_8BIT", "0")))
        self.load_4bit  = bool(int(os.getenv("LOAD_4BIT", "0")))
        self.debug      = bool(int(os.getenv("DEBUG", "0")))

def initialize_model():
    global tokenizer, model, image_processor, context_len, args
    if not LLAVA_AVAILABLE:
        print("[init] LLaVA not available; cannot init.")
        return False
    try:
        args = _Args()
        model_name = get_model_name_from_path(args.model_path)
        tokenizer_, model_, image_processor_, context_len_ = load_pretrained_model(
            args.model_path, args.model_base, model_name, args.load_8bit, args.load_4bit
        )
        # model'ı cuda’ya taşı
        try:
            _ = next(model_.parameters()).device
        except Exception:
            if torch.cuda.is_available():
                model_ = model_.to(torch.device("cuda"))
        
        model_.eval()
        globals()["tokenizer"] = tokenizer_
        globals()["model"] = model_
        globals()["image_processor"] = image_processor_
        globals()["context_len"] = context_len_
        chat_manager.init_if_needed(args, args.model_path, tokenizer_, model_, image_processor_, context_len_)
        print("[init] model/tokenizer/image_processor loaded.")
        return True
    except Exception as e:
        print(f"[init] failed: {e}")
        return False

# ===================== HF EndpointHandler =====================

class EndpointHandler:
    """Hugging Face Endpoint uyumlu sınıf"""
    def __init__(self, model_dir):
        self.model_dir = model_dir
        print(f"EndpointHandler initialized with model_dir: {model_dir}")
        
    def __call__(self, payload):
        if "inputs" in payload:
            return query(payload["inputs"])
        return query(payload)
        
    def health_check(self):
        return health_check()
        
    def get_model_info(self):
        return get_model_info()

if __name__ == "__main__":
    print("Handler ready (Deterministik Mode: do_sample=False). Use `EndpointHandler` or `query`.")