Create app.py

app.py

@@ -0,0 +1,54 @@
+# Import necessary libraries
+from transformers import AutoModelForCausalLM, AutoTokenizer, Trainer, TrainingArguments
+from datasets import load_dataset
+
+# Step 1: Load pre-trained model and tokenizer
+MODEL_NAME = "deepseek-ai/DeepSeek-V3-0324"  # Pre-trained model from Hugging Face
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+model = AutoModelForCausalLM.from_pretrained(MODEL_NAME)
+
+# Step 2: Load your custom dataset from Hugging Face
+dataset = load_dataset("epicDev123/deepseek")  # Replace with your dataset name if different
+
+# Step 3: Tokenization function (tokenize the text data for model input)
+def tokenize_function(examples):
+    return tokenizer(examples["text"], padding="max_length", truncation=True)
+
+# Tokenize the dataset
+tokenized_datasets = dataset.map(tokenize_function, batched=True)
+
+# Step 4: Set up training arguments
+training_args = TrainingArguments(
+    output_dir="./results",              # Output directory to save the fine-tuned model
+    num_train_epochs=3,                  # Number of training epochs
+    per_device_train_batch_size=8,       # Batch size for training
+    per_device_eval_batch_size=8,        # Batch size for evaluation
+    warmup_steps=500,                    # Number of warmup steps
+    weight_decay=0.01,                   # Weight decay for regularization
+    logging_dir="./logs",                # Directory for logs
+    logging_steps=10,                    # Log training every 10 steps
+    save_steps=500,                      # Save model checkpoints every 500 steps
+    evaluation_strategy="epoch",         # Evaluate model after each epoch
+    save_total_limit=2,                  # Limit the number of saved checkpoints
+)
+
+# Step 5: Initialize the Trainer
+trainer = Trainer(
+    model=model,                         # Model to train
+    args=training_args,                  # Training arguments
+    train_dataset=tokenized_datasets["train"],  # Training dataset
+    eval_dataset=tokenized_datasets["validation"],  # Validation dataset (optional)
+)
+
+# Step 6: Fine-tune the model
+trainer.train()
+
+# Step 7: Save the fine-tuned model
+model.save_pretrained("./fine_tuned_deepseek")   # Save the model to a directory
+tokenizer.save_pretrained("./fine_tuned_deepseek")  # Save the tokenizer
+
+# Step 8: Optionally, you can push the model to Hugging Face Model Hub (after logging into Hugging Face)
+model.push_to_hub("your-username/fine-tuned-deepseek")  # Replace with your username and desired model name
+tokenizer.push_to_hub("your-username/fine-tuned-deepseek")
+
+print("Fine-tuning complete! Your model has been saved.")