trackio/pr_310/en/transformers_integration.md

Transformers Integration

Trackio integrates natively with Transformers so you can log metrics with minimal setup. Ensure you have the latest version of transformers installed (version 4.54.0 or higher).

import numpy as np
from datasets import Dataset
from transformers import Trainer, AutoModelForCausalLM, TrainingArguments

# Create a fake dataset
data = np.random.randint(0, 1000, (8192, 64)).tolist()
dataset = Dataset.from_dict({"input_ids": data, "labels": data})

# Train a model using the Trainer API
trainer = Trainer(
    model=AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-0.6B"),
    args=TrainingArguments(run_name="Qwen3-0.6B-training", report_to="trackio"),
    train_dataset=dataset,
)
trainer.train()

Configuring Project and Space

You can specify your Trackio project name and space ID using environment variables:

export TRACKIO_PROJECT_NAME="my-project"
export TRACKIO_SPACE_ID="username/space_id"

Or set them directly in Python:

import os

os.environ["TRACKIO_PROJECT_NAME"] = "my-project"
os.environ["TRACKIO_SPACE_ID"] = "username/space_id"

# rest of your code...