# Uncomment the following line to install necessary libraries
# !pip install transformers datasets torch torchvision torchaudio python-dotenv numpy==1.26.4 accelerate -q

import os
from dotenv import load_dotenv
from huggingface_hub import login

# Load environment variables from .env file
load_dotenv()
hf_token = os.getenv('HUGGINGFACE_ACCESS_TOKEN')

if not hf_token:
    raise ValueError('HUGGINGFACE_ACCESS_TOKEN not found in .env file')

login(token=hf_token)

from transformers import pipeline

sentiment_analyzer = pipeline("sentiment-analysis", model="distilbert-base-uncased-finetuned-sst-2-english")
result = sentiment_analyzer("Hugging Face is awesome!")
print(result)

from transformers import pipeline
import warnings
warnings.filterwarnings('ignore')  # Suppress warnings
import logging
logging.getLogger('transformers').setLevel(logging.ERROR)  # Suppress transformers logs

text_generator = pipeline("text-generation")
result = text_generator("In a world where AI is becoming more prevalent,", max_length=50)

# Only print the generated text
if isinstance(result, list) and len(result) > 0 and 'generated_text' in result[0]:
    print(result[0]['generated_text'])
else:
    print(result)

ner_pipeline = pipeline("ner", grouped_entities=True)
result = ner_pipeline("My name is John Doe and I live in New York City.")
print(result)

from transformers import AutoTokenizer, AutoModelForSequenceClassification

# Load the tokenizer and model
model_name = "distilbert-base-uncased-finetuned-sst-2-english"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)

# Prepare the input
text = "Hugging Face is a great tool for NLP."
inputs = tokenizer(text, return_tensors="pt")

# Make the prediction
outputs = model(**inputs)
logits = outputs.logits
predicted_class = logits.argmax().item()

print(f"Predicted class: {predicted_class}")

# Example with negative sentiment (should predict class 0)
text_neg = "This is a terrible experience. I am very disappointed."
inputs_neg = tokenizer(text_neg, return_tensors="pt")
outputs_neg = model(**inputs_neg)
logits_neg = outputs_neg.logits
predicted_class_neg = logits_neg.argmax().item()
print(f"Predicted class: {predicted_class_neg}")

import warnings
warnings.filterwarnings('ignore')  # Suppress all warnings
import logging
logging.getLogger('transformers').setLevel(logging.ERROR)  # Suppress transformers logs
logging.getLogger('datasets').setLevel(logging.ERROR)  # Suppress datasets logs
logging.getLogger('filelock').setLevel(logging.ERROR)  # Suppress filelock logs

from datasets import load_dataset
from transformers import AutoTokenizer, AutoModelForSequenceClassification, Trainer, TrainingArguments

# Load a dataset
dataset = load_dataset("imdb", split="train[:1%]")

# Load a tokenizer and model
model_name = "distilbert-base-uncased"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name, num_labels=2)

# Preprocess the dataset
def tokenize_function(examples):
    return tokenizer(examples["text"], padding="max_length", truncation=True)

tokenized_dataset = dataset.map(tokenize_function, batched=True)

# Set up the training arguments
training_args = TrainingArguments(
    output_dir="./results",
    num_train_epochs=1,
    per_device_train_batch_size=16,
    per_device_eval_batch_size=16,
    warmup_steps=500,
    weight_decay=0.01,
    logging_dir="./logs",
)

# Create the Trainer
trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized_dataset,
)

# Fine-tune the model
trainer.train()