inference.py

import torch
from transformers import AutoTokenizer, AutoModelForSequenceClassification, ModernBertConfig
from typing import Dict, Any
import yaml
import os
from models import ModernBertForSentiment

class SentimentInference:
    def __init__(self, config_path: str = "config.yaml"):
        """Load configuration and initialize model and tokenizer from local checkpoint or Hugging Face Hub."""
        print(f"--- Debug: SentimentInference __init__ received config_path: {config_path} ---") # Add this
        with open(config_path, 'r') as f:
            config_data = yaml.safe_load(f)
        print(f"--- Debug: SentimentInference loaded config_data: {config_data} ---") # Add this
        
        model_yaml_cfg = config_data.get('model', {})
        inference_yaml_cfg = config_data.get('inference', {})
        
        model_hf_repo_id = model_yaml_cfg.get('name_or_path')
        tokenizer_hf_repo_id = model_yaml_cfg.get('tokenizer_name_or_path', model_hf_repo_id)
        local_model_weights_path = inference_yaml_cfg.get('model_path') # Path for local .pt file

        print(f"--- Debug: model_hf_repo_id: {model_hf_repo_id} ---") # Add this
        print(f"--- Debug: local_model_weights_path: {local_model_weights_path} ---") # Add this

        self.max_length = inference_yaml_cfg.get('max_length', model_yaml_cfg.get('max_length', 512))

        # --- Tokenizer Loading (always from Hub for now, or could be made conditional) ---
        if not tokenizer_hf_repo_id and not model_hf_repo_id:
            raise ValueError("Either model.tokenizer_name_or_path or model.name_or_path (as fallback for tokenizer) must be specified in config.yaml")
        effective_tokenizer_repo_id = tokenizer_hf_repo_id or model_hf_repo_id
        print(f"Loading tokenizer from: {effective_tokenizer_repo_id}")
        self.tokenizer = AutoTokenizer.from_pretrained(effective_tokenizer_repo_id)

        # --- Model Loading --- #
        # Determine if we are loading from a local .pt file or from Hugging Face Hub
        load_from_local_pt = False
        if local_model_weights_path and os.path.isfile(local_model_weights_path):
            print(f"Found local model weights path: {local_model_weights_path}")
            print(f"--- Debug: Found local model weights path: {local_model_weights_path} ---") # Add this
            load_from_local_pt = True
        elif not model_hf_repo_id:
            raise ValueError("No local model_path found and model.name_or_path (for Hub) is not specified in config.yaml")

        print(f"--- Debug: load_from_local_pt is: {load_from_local_pt} ---") # Add this

        if load_from_local_pt:
            print("Attempting to load model from local .pt checkpoint...")
            print("--- Debug: Entering LOCAL .pt loading path ---") # Add this
            # Base BERT config must still be loaded, usually from a Hub ID (e.g., original base model)
            # This base_model_for_config_id is crucial for building the correct ModernBertForSentiment structure.
            base_model_for_config_id = model_yaml_cfg.get('base_model_for_config', model_hf_repo_id or tokenizer_hf_repo_id)
            print(f"--- Debug: base_model_for_config_id (for local .pt): {base_model_for_config_id} ---") # Add this
            if not base_model_for_config_id:
                 raise ValueError("For local .pt loading, model.base_model_for_config must be specified in config.yaml (e.g., 'answerdotai/ModernBERT-base') to build the model structure.")
            
            print(f"Loading ModernBertConfig for structure from: {base_model_for_config_id}")
            bert_config = ModernBertConfig.from_pretrained(base_model_for_config_id)
            
            # Augment config with parameters from model_yaml_cfg
            bert_config.pooling_strategy = model_yaml_cfg.get('pooling_strategy', 'mean')
            bert_config.num_weighted_layers = model_yaml_cfg.get('num_weighted_layers', 4)
            bert_config.classifier_dropout = model_yaml_cfg.get('dropout')
            bert_config.num_labels = model_yaml_cfg.get('num_labels', 1)
            # bert_config.loss_function = model_yaml_cfg.get('loss_function') # If needed by __init__

            print("Instantiating ModernBertForSentiment model structure...")
            self.model = ModernBertForSentiment(bert_config)
            
            print(f"Loading model weights from local checkpoint: {local_model_weights_path}")
            checkpoint = torch.load(local_model_weights_path, map_location=torch.device('cpu'))
            if isinstance(checkpoint, dict) and 'model_state_dict' in checkpoint:
                model_state_to_load = checkpoint['model_state_dict']
            else:
                model_state_to_load = checkpoint # Assume it's the state_dict itself
            self.model.load_state_dict(model_state_to_load)
            print(f"Model loaded successfully from local checkpoint: {local_model_weights_path}.")

        else: # Load from Hugging Face Hub
            print(f"Attempting to load model from Hugging Face Hub: {model_hf_repo_id}...")
            print(f"--- Debug: Entering HUGGING FACE HUB loading path ---") # Add this
            print(f"--- Debug: model_hf_repo_id (for Hub loading): {model_hf_repo_id} ---") # Add this
            if not model_hf_repo_id:
                raise ValueError("model.name_or_path must be specified in config.yaml for Hub loading.")

            print(f"Loading base ModernBertConfig from: {model_hf_repo_id}")
            loaded_config = ModernBertConfig.from_pretrained(model_hf_repo_id)
            
            # Augment loaded_config
            loaded_config.pooling_strategy = model_yaml_cfg.get('pooling_strategy', 'mean')
            loaded_config.num_weighted_layers = model_yaml_cfg.get('num_weighted_layers', 6) # Default to 6 now
            loaded_config.classifier_dropout = model_yaml_cfg.get('dropout')
            loaded_config.num_labels = model_yaml_cfg.get('num_labels', 1)

            print(f"Instantiating and loading model weights for {model_hf_repo_id}...")
            self.model = AutoModelForSequenceClassification.from_pretrained(
                model_hf_repo_id,
                config=loaded_config,
                trust_remote_code=True,
                force_download=True  # <--- TEMPORARY - remove when everything is working
            )
            print(f"Model {model_hf_repo_id} loaded successfully from Hugging Face Hub.")
        
        self.model.eval()
        
    def predict(self, text: str) -> Dict[str, Any]:
        inputs = self.tokenizer(text, return_tensors="pt", truncation=True, max_length=self.max_length, padding=True)
        with torch.no_grad():
            outputs = self.model(input_ids=inputs['input_ids'], attention_mask=inputs['attention_mask'])
        logits = outputs.get("logits") # Use .get for safety
        if logits is None:
            raise ValueError("Model output did not contain 'logits'. Check model's forward pass.")
        prob = torch.sigmoid(logits).item()
        return {"sentiment": "positive" if prob > 0.5 else "negative", "confidence": prob}