Downloads · 30 days
0
hw2-classical-automl-3d-print
srivathsanb14/hw2-classical-automl-3d-print
hw2-classical-automl-3d-print is a tabular classification model from srivathsanb14. Use it for the tabular classification task on the model card, and read the license before you ship it in a product. It is set up for autogluon. The card lists the license as mit.
Predict whether a Bambu-sliced part is a long print (= 6 hours) from slicer settings. Teaching artifact only.
Access
Public
Updated Sep 21, 2026
Repo size
815 KB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.zip815 KB · 99%
At a glance
- Task
- Tabular Classification
- Library
- autogluon
- License
- mit
- Access
- Public
- Created
- Sep 21, 2026
- Updated
- Sep 21, 2026
- SHA
- c200d352
Datasets
Model card
From the Hugging Face model README
HW2: 3D-print long-job classifier (AutoGluon Tabular)
Purpose
Predict whether a Bambu-sliced part is a long print (>= 6 hours) from slicer settings. Teaching artifact only.
Task and inputs
Target long_print: 0 = short, 1 = long. Positive class is 1.
| Feature | Type |
|---|---|
| filament_g | numeric, grams |
| infill_pct | numeric, 0-100 |
| layer_height_mm | numeric, mm |
| nozzle_mm | categorical hardware |
| use_category | categorical (functional / novelty) |
Excluded: print_time_min, print_time_hms (the target is a threshold of these), plus IDs and augmentation metadata.
Data and splits
Source: yennik16/3d-print-jobs-tabular (classmate HW1; yennik16).
| Split | Rows | Contents |
|---|---|---|
| train | 321 | 21 real parts + synthetic children |
| validation | 4 | 4 unaugmented originals |
| test | 5 | 5 unaugmented originals |
Training setup
| Setting | Value |
|---|---|
| Framework | AutoGluon Tabular 1.6.1 |
| Preset | medium_quality |
| Time budget | 300s (wall 28.5s) |
| Selection metric | balanced_accuracy on published validation |
| Seed | 24679 |
| Bagging folds | 0 |
| Stack levels | 0 |
Best model: WeightedEnsemble_L2.
Metrics
Test on 5 original parts.
| Metric | Model | Majority baseline |
|---|---|---|
| Accuracy | 1.0000 (5/5) | 0.6000 |
| Balanced accuracy | 1.0000 | 0.5000 |
| Weighted F1 | 1.0000 | — |
| Macro F1 | 1.0000 | — |
One flipped prediction moves accuracy by 20 points.
Limitations and ethics
Do not schedule paid print jobs with this model.
License
MIT for this repo. Dataset is CC BY 4.0 (attribute yennik16).
Hardware / compute
Google Colab, 300s AutoML budget, seed 24679. Python 3.13.15.
Load
from pathlib import Path
import zipfile
from huggingface_hub import hf_hub_download
from autogluon.tabular import TabularPredictor
archive = hf_hub_download("srivathsanb14/hw2-classical-automl-3d-print", "autogluon_predictor_dir.zip")
out = Path("ag_3dprint"); out.mkdir(exist_ok=True)
with zipfile.ZipFile(archive) as zf:
zf.extractall(out)
predictor = TabularPredictor.load(str(next(out.rglob("predictor.pkl")).parent))
AI usage disclosure
I chose the classmate dataset, the leakage rule, and reviewed the leaderboard before upload. Cursor helped scaffold this notebook.
- Task
- Tabular Classification
- Library
- autogluon
- License
- mit
- Languages
- en
- Created
- Sep 21, 2026
- Updated
- Sep 21, 2026