Quickstart

Python API

import ecal

result = ecal.estimate(
    model_type="MLP",
    model_params={"num_layers": 3, "din": 10, "dout": 2},
    num_samples=1000,
    num_epochs=50,
    hardware="apple_m2",
)

print(f"Total energy: {result['total']:.4f} J")
print(f"eCAL: {result['ecal_j_per_bit']:.2e} J/bit")

See ecal.api.estimate() for the full parameter reference.

CLI

# Estimate energy for an MLP
ecal estimate --model MLP --layers 3 --epochs 50 --hardware apple_m2

# JSON output
ecal estimate --model Transformer --layers 6 --hardware nvidia_h100_sxm --json

# List available hardware profiles
ecal profiles

# Version
ecal --version

See CLI reference for the full list of flags.

Supported Models

Model	FLOP Calculator	Key Parameters
MLP	MLPCalculator	num_layers, din, dout
CNN	CNNCalculator	num_cnv_layers, num_pool_layers, i_r, k_r
KAN	KANCalculator	num_layers, grid_size, din, dout
Transformer	TransformerCalculator	context_length, embedding_size, num_heads, num_decoder_blocks