Update app.py

app.py

@@ -1,42 +1,51 @@
 import gradio as gr
-from transformers import AutoTokenizer, AutoModelForTokenClassification, AutoConfig
+import torch
+from transformers import AutoTokenizer
 
-# Load configuration manually if it's missing in the repository
-try:
-    config = AutoConfig.from_pretrained("viv/UD_Greek-GUD")
-except OSError:
-    # Fallback: Create a configuration manually
-    config = AutoConfig.from_dict({
-        "model_type": "bert",  # Change this based on your model type
-        "hidden_size": 768,
-        "num_attention_heads": 12,
-        "num_hidden_layers": 12,
-        "vocab_size": 30000,
-        "max_position_embeddings": 512,
-        "type_vocab_size": 2,
-        "hidden_act": "gelu",
-        "layer_norm_eps": 1e-12,
-        "initializer_range": 0.02
-    })
+# Define model paths
+models_path = "./models"
+tokenizer_path = f"{models_path}/el_new_tokenizer.pt"
+lemmatizer_path = f"{models_path}/el_new_nocharlm_lemmatizer.pt"
+tagger_path = f"{models_path}/el_new_transformer_tagger.pt"
+parser_path = f"{models_path}/el_new_transformer_parser.pt"  # Updated parser model path
 
-# Load model and tokenizer with the provided or fallback config
-tokenizer = AutoTokenizer.from_pretrained("viv/UD_Greek-GUD")
-model = AutoModelForTokenClassification.from_pretrained("viv/UD_Greek-GUD", config=config)
+# Load models
+tokenizer = torch.load(tokenizer_path)  # Load tokenizer
+lemmatizer_model = torch.load(lemmatizer_path)  # Load lemmatizer
+tagger_model = torch.load(tagger_path)  # Load POS tagger
+parser_model = torch.load(parser_path)  # Load dependency parser
 
 # Prediction function
 def predict(text):
+    # Tokenize input
     inputs = tokenizer(text, return_tensors="pt")
-    outputs = model(**inputs)
-    # Process and return predictions
-    return outputs.logits.argmax(-1).tolist()
+    
+    # Perform lemmatization
+    lemma_outputs = lemmatizer_model(**inputs)
+    lemmas = lemma_outputs.logits.argmax(-1).tolist()  # Process lemmatizer output
+    
+    # Perform POS tagging
+    pos_outputs = tagger_model(**inputs)
+    pos_tags = pos_outputs.logits.argmax(-1).tolist()  # Process tagger output
+    
+    # Perform dependency parsing
+    dep_outputs = parser_model(**inputs)
+    dep_parse = dep_outputs.logits.argmax(-1).tolist()  # Process parser output
+    
+    # Return results
+    return {
+        "lemmas": lemmas,
+        "pos_tags": pos_tags,
+        "dep_parse": dep_parse,
+    }
 
 # Gradio Interface
 interface = gr.Interface(
     fn=predict,
     inputs="text",
-    outputs="text",
-    title="UD Greek GUD Model",
-    description="Analyze text using the UD Greek GUD model.",
+    outputs="json",
+    title="Greek NLP Pipeline",
+    description="Perform lemmatization, POS tagging, and dependency parsing for Greek text using custom models.",
 )
 
 # Launch interface