Skip to main content

​
Overview

Lyceum Cloud provides various machine types optimized for different computational workloads. The platform offers both CPU and GPU configurations to match your specific requirements.

​
Execution Types

When starting an execution through the API, you can specify the execution_type parameter: 
{
  "code": "your_python_code",
  "execution_type": "cpu"  // or "gpu", "auto"
}

​
Available Options

  • cpu: Standard CPU execution environment
  • gpu: GPU-accelerated execution
  • auto: Automatic selection based on your code

​
Automatic Resource Selection

The platform can automatically select appropriate resources by analyzing your code: 
# The platform detects ML frameworks and selects GPU automatically
import torch
import tensorflow as tf

# Your deep learning code here
model = torch.nn.Sequential(...)
When using execution_type: "auto", the system will:
  • Detect imported packages and frameworks
  • Estimate resource requirements
  • Select the most appropriate machine type
  • Optimize for cost-effectiveness

​
Resource Monitoring

Monitor resource usage during execution to optimize your selection: 
import psutil

# Check CPU usage
cpu_percent = psutil.cpu_percent(interval=1)
print(f"CPU Usage: {cpu_percent}%")

# Check memory usage
memory = psutil.virtual_memory()
print(f"Memory Usage: {memory.percent}%")

# Check if GPU is available
try:
    import torch
    if torch.cuda.is_available():
        print(f"GPU Available: {torch.cuda.get_device_name()}")
        print(f"GPU Memory: {torch.cuda.mem_get_info()}")
except ImportError:
    print("PyTorch not installed")

​
Choosing the Right Type

​
Use CPU for:

  • Data processing and analysis with pandas/numpy
  • Web scraping and API calls
  • File processing and transformations
  • Statistical computing
  • General Python scripts

​
Use GPU for:

  • Deep learning model training
  • Computer vision tasks
  • Natural language processing with transformers
  • Parallel computing with CUDA
  • Large-scale matrix operations

​
Timeout Settings

Configure execution timeouts (1-600 seconds) to prevent runaway processes: 
{
  "code": "your_code",
  "timeout": 300,  // 5 minutes
  "execution_type": "cpu"
}

​
Environment Variables

Access execution environment information: 
import os

# Check execution type
execution_type = os.environ.get('LYCEUM_EXECUTION_TYPE', 'unknown')
print(f"Running on: {execution_type}")

# Check available resources
if os.environ.get('CUDA_VISIBLE_DEVICES'):
    print("GPU execution environment detected")

​
Storage Integration

All machine types have access to your cloud storage: 
# Files are available at the storage path
storage_path = os.environ.get('LYCEUM_STORAGE_PATH', '/lyceum/storage/')

# Access your uploaded files
import pandas as pd
df = pd.read_csv("{storage_path}")

# Save results
df.to_csv("{storage_path}")
For current pricing and detailed specifications, check the dashboard or contact support. Machine type availability may vary based on demand.