app.py

import streamlit as st
import pandas as pd
import os, csv
from huggingface_hub import hf_hub_download, HfApi

HF_TOKEN = os.getenv('HUGGING_FACE_HUB_TOKEN')

CACHED_FILE_PATH = hf_hub_download(repo_id="sasha/co2_submissions", filename="dynamic_emissions.csv", repo_type="dataset")

api = HfApi()

def write_to_csv(hardware, training_time, provider, carbon_intensity, dynamic_emissions):
    with open(CACHED_FILE_PATH,'a', newline='') as f:
        writer = csv.writer(f)
        writer.writerow([hardware, training_time, provider, carbon_intensity, dynamic_emissions])
    api.upload_file(
        path_or_fileobj=CACHED_FILE_PATH,
        path_in_repo="dynamic_emissions.csv",
        repo_id="sasha/co2_submissions",
        repo_type="dataset",
    )



st.set_page_config(
    page_title="AI Carbon Calculator",
    layout="wide",
)

tdp_url = "https://raw.githubusercontent.com/mlco2/impact/master/data/gpus.csv"
compute_url = "https://raw.githubusercontent.com/mlco2/impact/master/data/impact.csv"


electricity_url = "https://raw.githubusercontent.com/mlco2/impact/master/data/2021-10-27yearly_averages.csv"

server_sheet_id = "1DqYgQnEDLQVQm5acMAhLgHLD8xXCG9BIrk-_Nv6jF3k"
server_sheet_name = "Server%20Carbon%20Footprint"
server_url = f"https://docs.google.com/spreadsheets/d/{server_sheet_id}/gviz/tq?tqx=out:csv&sheet={server_sheet_name}"


embodied_gpu_sheet_id = "1DqYgQnEDLQVQm5acMAhLgHLD8xXCG9BIrk-_Nv6jF3k"
embodied_gpu_sheet_name = "Scope%203%20Ratios"
embodied_gpu_url = f"https://docs.google.com/spreadsheets/d/{embodied_gpu_sheet_id}/gviz/tq?tqx=out:csv&sheet={embodied_gpu_sheet_name}"

TDP =pd.read_csv(tdp_url)

instances = pd.read_csv(compute_url)
providers = [p.upper() for p in instances['provider'].unique().tolist()]
providers.append('Local/Private Infastructure')

kg_per_mile = 0.348

electricity = pd.read_csv(electricity_url)
servers = pd.read_csv(server_url)
embodied_gpu = pd.read_csv(embodied_gpu_url)

st.image('images/MIT_carbon_image_narrow.png') 
st.title("AI Carbon Calculator")

st.markdown('## Estimate your model\'s CO2 carbon footprint!')

st.markdown('##### You can use this tool to calculate different aspects of your model\'s carbon footprint.')

st.markdown('### Dynamic Emissions')
st.markdown('##### These are the emissions produced by generating the electricity needed to train your model.')
with st.expander("Calculate the emissions produced by energy consumption of model training"):
    col1, col2, col3, col4 = st.columns(4)
    with col1:
        hardware = st.selectbox('GPU used', TDP['name'].tolist())
        gpu_tdp = TDP['tdp_watts'][TDP['name'] == hardware].tolist()[0]
        st.markdown("Different GPUs have different TDP (Thermal Design Power), which impacts how much energy you use.")
    with col2:
       training_time = st.number_input('Total number of GPU hours')
       st.markdown('This is calculated by multiplying the number of GPUs you used by the training time: '
                   'i.e. if you used 100 GPUs for 10 hours, this is equal to 100x10 = 1,000 GPU hours.')
    with col3:
       provider = st.selectbox('Provider used', providers)
       st.markdown('If you can\'t find your provider here, select "Local/Private Infrastructure".')
    with col4:
        if provider != 'Local/Private Infastructure':
            provider_instances = instances['region'][instances['provider'] == provider.lower()].unique().tolist()
            region = st.selectbox('Provider used', provider_instances)
            carbon_intensity = instances['impact'][(instances['provider'] == provider.lower()) & (instances['region'] == region)].tolist()[0]

        else:
            carbon_intensity = st.number_input('Carbon intensity of your energy grid, in grams of CO2 per kWh')
            st.markdown('You can consult a resource like the [IEA](https://www.iea.org/countries) or '
                        ' [Electricity Map](https://app.electricitymaps.com/) to get this information.')
    dynamic_emissions = round(gpu_tdp * training_time * carbon_intensity/1000000)
    st.metric(label="Dynamic emissions", value=str(dynamic_emissions)+' kilograms of CO2eq')
    st.markdown('This is roughly equivalent to '+ str(round(dynamic_emissions/kg_per_mile,1)) + ' miles driven in an average US car'
    ' produced in 2021. [(Source: energy.gov)](https://www.energy.gov/eere/vehicles/articles/fotw-1223-january-31-2022-average-carbon-dioxide-emissions-2021-model-year)')
    st.button(label="Anonymously share my data", help="Share the data from your model anonymously for research purposes!",\
    on_click = lambda *args: write_to_csv(hardware, training_time, provider, carbon_intensity, dynamic_emissions))



st.markdown('### Idle Emissions')
st.markdown('##### These are the emissions produced by generating the electricity needed to power the rest of the infrastructure'
            'used for model training -- the datacenter, network, heating/cooling, storage, etc.')
st.markdown('Do you know what the PUE (Power Usage Effectiveness) of your infrastructure is?')



st.markdown('### Embodied Emissions')
st.markdown('Choose your hardware, runtime and cloud provider/physical infrastructure to estimate the carbon impact of your research.')



with st.expander("More information about our Methodology"):
    st.markdown('Building on the work of the [ML CO2 Calculator](https://mlco2.github.io/impact/), this tool allows you to consider'
                ' other aspects of your model\'s carbon footprint based on the LCA methodology.')


    st.image('images/LCA_CO2.png', caption='The LCA methodology - the parts in green are those we focus on.')