Update app.py

app.py

@@ -1,20 +1,15 @@
 import os
-import re
-import pickle
-import zipfile
 import numpy as np
 import tensorflow as tf
 from fastapi import FastAPI
 from pydantic import BaseModel
 from fastapi.middleware.cors import CORSMiddleware
 
-# Optimasi RAM (Wajib untuk Hugging Face Free Tier)
+# Optimasi RAM server HF
 os.environ["TF_CPP_MIN_LOG_LEVEL"] = "3"
 os.environ["CUDA_VISIBLE_DEVICES"] = "-1"
-os.environ["TF_NUM_INTEROP_THREADS"] = "1"
-os.environ["TF_NUM_INTRAOP_THREADS"] = "1"
 
-app = FastAPI(title="NER & Classification API Service")
+app = FastAPI(title="AI Classification Service")
 app.add_middleware(
     CORSMiddleware,
     allow_origins=["*"],
@@ -23,174 +18,45 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
-# Global Variables
-ner_model = None
-vocab = None
 clf_model = None
 
-MAX_SEQ_LEN = 64
-IDX2LABEL = {0: "<PAD>", 1: "O", 2: "B-MERCHANT", 3: "I-MERCHANT", 4: "B-ITEM", 5: "I-ITEM", 6: "B-TOTAL", 7: "I-TOTAL"}
-UNK_TOKEN = "<UNK>"
-NUM_TOKEN = "<NUM>"
-PRICE_TOKEN = "<PRICE>"
-
-class TextRequest(BaseModel):
-    text: str
-
-# =========================================================
-# JURUS 1: BYPASS NER (Menghindari bug quantization_config)
-# =========================================================
-def build_manual_ner_model(vocab_size=7929, max_len=64, num_classes=8):
-    inputs = tf.keras.Input(shape=(max_len,), dtype='int32', name='tokens')
-    x = tf.keras.layers.Embedding(input_dim=vocab_size, output_dim=64, mask_zero=True, name='embedding')(inputs)
-    x = tf.keras.layers.SpatialDropout1D(0.3, name='spatial_dropout1d')(x)
-    
-    lstm1_f = tf.keras.layers.LSTM(128, return_sequences=True, dropout=0.4, name='forward_lstm')
-    lstm1_b = tf.keras.layers.LSTM(128, return_sequences=True, go_backwards=True, dropout=0.4, name='backward_lstm')
-    x = tf.keras.layers.Bidirectional(lstm1_f, backward_layer=lstm1_b, name='bilstm_1')(x)
-    
-    lstm2_f = tf.keras.layers.LSTM(64, return_sequences=True, dropout=0.4, name='forward_lstm_1')
-    lstm2_b = tf.keras.layers.LSTM(64, return_sequences=True, go_backwards=True, dropout=0.4, name='backward_lstm_1')
-    x = tf.keras.layers.Bidirectional(lstm2_f, backward_layer=lstm2_b, name='bilstm_2')(x)
-    
-    dense1 = tf.keras.layers.Dense(64, activation='relu', name='dense')
-    x = tf.keras.layers.TimeDistributed(dense1, name='proj')(x)
-    x = tf.keras.layers.Dropout(0.4, name='dropout')(x)
-    
-    dense2 = tf.keras.layers.Dense(num_classes, activation='linear', name='dense_1')
-    x = tf.keras.layers.TimeDistributed(dense2, name='logits')(x)
-    probs = tf.keras.layers.Activation('softmax', name='probabilities')(x)
-    
-    return tf.keras.Model(inputs=inputs, outputs=probs, name='BiLSTM_NER')
-
-def get_ner_model():
-    global ner_model, vocab
-    if ner_model is None:
-        with open("vocab.pkl", "rb") as f:
-            vocab = pickle.load(f)
-        ner_model = build_manual_ner_model()
-        ner_model.load_weights("receipt_ner_model.keras")
-    return ner_model, vocab
+# Schema request dari Space A
+class ItemRequest(BaseModel):
+    items_text: str
 
 # =========================================================
-# JURUS 2: BYPASS KLASIFIKASI + PAKSA BANGUN KAMUS 
+# LOAD MODEL SECARA NORMAL
 # =========================================================
-def build_manual_clf_model():
-    inputs = tf.keras.Input(shape=(1,), dtype=tf.string, name='input_layer')
-    x = tf.keras.layers.TextVectorization(
-        max_tokens=5000, 
-        output_sequence_length=50, 
-        name='text_vectorization'
-    )(inputs)
-    x = tf.keras.layers.Embedding(input_dim=5000, output_dim=64, name='embedding')(x)
-    x = tf.keras.layers.GlobalAveragePooling1D(name='global_average_pooling1d')(x)
-    x = tf.keras.layers.Dense(64, activation='relu', name='dense')(x)
-    x = tf.keras.layers.Dropout(0.2, name='dropout')(x)
-    outputs = tf.keras.layers.Dense(7, activation='softmax', name='dense_1')(x)
-    
-    return tf.keras.Model(inputs=inputs, outputs=outputs, name='functional')
-
-import zipfile # Pastikan import ini ada di bagian atas file
-
 def get_clf_model():
     global clf_model
     if clf_model is None:
-        clf_model = build_manual_clf_model()
-        clf_model.load_weights("model_strukly.keras")
-        
-        # --- JURUS HACKER: EKSTRAK KAMUS LANGSUNG DARI FILE .KERAS (ZIP) ---
-        tv_layer = clf_model.get_layer('text_vectorization')
-        try:
-            # Buka model .keras sebagai file ZIP
-            with zipfile.ZipFile("model_strukly.keras", 'r') as z:
-                for name in z.namelist():
-                    # Cari file teks di dalam folder assets
-                    if 'assets/' in name and name.endswith('.txt'):
-                        with z.open(name) as f:
-                            # Baca isi file, pisahkan per baris, dan jadikan kamus
-                            vocab_list = f.read().decode('utf-8').splitlines()
-                            tv_layer.set_vocabulary(vocab_list)
-                            print(f"Kamus berhasil disuntikkan secara paksa dari: {name}")
-                            break
-        except Exception as e:
-            print(f"Gagal mengekstrak kamus dari dalam ZIP: {e}")
-        # -------------------------------------------------------------------
-        
+        print("Memuat model klasifikasi secara normal...")
+        # Load langsung file .keras bawaan Anda
+        clf_model = tf.keras.models.load_model("model_strukly.keras", compile=False)
+        print("Model berhasil dimuat!")
     return clf_model
 
-# =========================================================
-# FUNGSI UTAMA (Inference API)
-# =========================================================
-def normalize_token(token: str):
-    token = token.strip()
-    if re.fullmatch(r"\d{1,3}([.,]\d{3})*", token): return PRICE_TOKEN
-    if re.fullmatch(r"\d+", token): return NUM_TOKEN
-    token = re.sub(r"^[^a-zA-Z0-9']+|[^a-zA-Z0-9']+$", "", token)
-    return token.upper() if token else UNK_TOKEN
-
-@app.post("/predict-ner/")
-def predict_ner(request: TextRequest):
+@app.post("/classify/")
+def classify_items(request: ItemRequest):
     try:
-        raw_text = request.text
-        entities = {"MERCHANT": [], "ITEM": [], "TOTAL": []}
-        class_id = None
+        raw_items = request.items_text
         
-        if not raw_text.strip():
-            return {"success": True, "entities": entities, "class_id": class_id}
-
-        # --- 1. PROSES NER ---
-        model, vcb = get_ner_model()
-        tokens = [t for line in raw_text.split("\n") for t in line.strip().split() if t]
+        # Jika Space A mengirim string kosong (tidak ada item terdeteksi)
+        if not raw_items.strip():
+             return {"success": True, "class_id": None}
+             
+        model = get_clf_model()
+        input_data = np.array([raw_items])
         
-        if tokens:
-            norm_tokens = [normalize_token(t) for t in tokens]
-            X = np.zeros((1, MAX_SEQ_LEN), dtype=np.int32)
-            length = min(len(norm_tokens), MAX_SEQ_LEN)
-            
-            for i in range(length):
-                X[0, i] = vcb.get(norm_tokens[i], vcb.get(UNK_TOKEN, 0))
-                
-            probs = model(X, training=False).numpy()
-            labels = [IDX2LABEL[int(np.argmax(probs[0, i]))] for i in range(length)]
-            
-            current_type = None
-            span = []
-            for tok, lbl in zip(tokens[:length], labels):
-                if lbl.startswith("B-"):
-                    if current_type: entities[current_type].append(" ".join(span))
-                    current_type = lbl[2:]
-                    span = [tok]
-                elif lbl.startswith("I-") and current_type == lbl[2:]:
-                    span.append(tok)
-                else:
-                    if current_type: entities[current_type].append(" ".join(span))
-                    current_type = None
-                    span = []
-            if current_type: entities[current_type].append(" ".join(span))
-
-        # --- 2. PROSES KLASIFIKASI ---
-        items_list = entities.get("ITEM", [])
-        teks_untuk_klasifikasi = " ".join(items_list)
+        # Prediksi dengan model AI
+        hasil_prediksi = model(input_data, training=False).numpy()
         
-        if teks_untuk_klasifikasi:
-            try:
-                c_model = get_clf_model()
-                input_data = np.array([teks_untuk_klasifikasi])
-                
-                hasil_prediksi = c_model(input_data, training=False).numpy()
-                
-                # Langsung mengembalikan index klasifikasi tertinggi (0 - 6)
-                class_id = int(np.argmax(hasil_prediksi[0]))
-            except Exception as e:
-                print(f"Error Klasifikasi internal: {e}")
-
-        return {
-            "success": True, 
-            "entities": entities, 
-            "class_id": class_id
-        }
+        # Ambil indeks dengan probabilitas tertinggi
+        class_id = int(np.argmax(hasil_prediksi[0]))
+        
+        return {"success": True, "class_id": class_id}
         
     except Exception as e:
         import traceback
-        print(f"Error NER/CLF detail: {traceback.format_exc()}")
-        return {"success": False, "error": str(e), "entities": {"MERCHANT": [], "ITEM": [], "TOTAL": []}, "class_id": None}
+        print(f"Error AI Klasifikasi: {traceback.format_exc()}")
+        return {"success": False, "error": str(e), "class_id": None}