wandb / app.py
tcapelle's picture
add slider and link
be81fab
raw
history blame
1.58 kB
import os, random
import wandb
import streamlit as st
import streamlit.components.v1 as components
from utils import train, WORDS
project = "st"
entity = "capecape"
HEIGHT = 720
def get_project(api, name, entity=None):
return api.project(name, entity=entity).to_html(height=HEIGHT)
st.title("The wandb Dashboard πŸ‘‡")
run_name = "-".join(random.choices(WORDS, k=2)) + f"-{random.randint(0,100)}"
# Sidebar
sb = st.sidebar
sb.title("Train your model")
# wandb_token = sb.text_input("paste your wandb Api key if you want: https://wandb.ai/authorize", type="password")
# wandb.login(key=wandb_token)
wandb.login(anonymous="must")
api = wandb.Api()
st.success(f"You should see a new run named **{run_name}**, it\'ll have a green circle while it\'s still active")
# render wandb dashboard
components.html(get_project(api, project, entity), height=HEIGHT)
# run params
runs = 1
epochs = sb.slider('Number of epochs:', min_value=100, max_value=500, value=100)
pseudo_code = """
We will execute a simple training loop
```python
wandb.init(project="st", ...)
for i in range(epochs):
acc = 1 - 2 ** -i - random()
loss = 2 ** -i + random()
wandb.log({"acc": acc,
"loss": loss})
```
"""
sb.write(pseudo_code)
sb.write("Click πŸ‘‡ to start logging")
# train model
if sb.button("Run Example"):
print("Running training")
for i in range(runs):
my_bar = sb.progress(0)
train(name=run_name, project=project, entity=entity, epochs=epochs, bar=my_bar)
st.subheader("Check our πŸ”₯ [Pytorch Intro colab](https://wandb.me/intro) πŸ”₯")