Create train_orpheus.py

train_orpheus.py

@@ -0,0 +1,42 @@
+import os
+import torch
+from transformers import Trainer, TrainingArguments
+from datasets import load_dataset
+import subprocess
+
+# Install required packages
+subprocess.run("pip install git+https://github.com/canopyai/Orpheus-TTS.git", shell=True)
+subprocess.run("pip install orpheus-speech vllm==0.7.3", shell=True)
+
+# Load the dataset
+dataset = load_dataset("Emotional_Speech_Dataset_(ESD)")
+
+# Get the model
+from transformers import AutoModelForCausalLM, AutoTokenizer
+model = AutoModelForCausalLM.from_pretrained("canopylabs/orpheus-3b-0.1-pretrained")
+tokenizer = AutoTokenizer.from_pretrained("canopylabs/orpheus-3b-0.1-pretrained")
+
+# Setup training arguments
+training_args = TrainingArguments(
+    output_dir="./orpheus-finetuned",
+    per_device_train_batch_size=2,
+    gradient_accumulation_steps=4,
+    learning_rate=5e-5,
+    num_train_epochs=3,
+    save_strategy="steps",
+    save_steps=500,
+)
+
+# Start training
+trainer = Trainer(
+    model=model,
+    args=training_args,
+    train_dataset=dataset,
+    tokenizer=tokenizer,
+)
+
+trainer.train()
+
+# Save the model
+model.save_pretrained("./orpheus-finetuned-model")
+tokenizer.save_pretrained("./orpheus-finetuned-model")