Downloads · 30 days
0
2026-24679-vegetable-mixed-classifier
CXu0630/2026-24679-vegetable-mixed-classifier
2026-24679-vegetable-mixed-classifier is a image classification model from CXu0630. Use it when you need a label for an image. It is set up for autogluon.
Binary image classifier predicting whether a photo shows a single vegetable (0) or a mixed group of several vegetable types (1).
Access
Public
Updated Sep 23, 2026
Repo size
31.6 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pkl16.5 MB · 52%
At a glance
- Task
- Image Classification
- Library
- autogluon
- Access
- Public
- Created
- Sep 23, 2026
- Updated
- Sep 23, 2026
- SHA
- 5d15166c
Datasets
Model card
From the Hugging Face model README
Vegetable Photo Mixed/Single Classifier
Binary image classifier predicting whether a photo shows a single vegetable (0) or a mixed group of several vegetable types (1).
Source data
Trained on yennik16/vegetable_dataset, a classmate-built
image dataset for 24-679 (Fall 2026), Carnegie Mellon University.
Training
Selected via an AutoGluon MultiModalPredictor AutoML search over backbone architecture
(mobilenetv3_small_100, resnet18, efficientnet_b0) and learning rate, using a Bayesian
searcher with the ASHA early-stopping scheduler (6 trials, 15-minute total budget). See
fit_summary() in the training notebook for the selected trial's exact configuration.
Test results (4 original photos)
| Metric | Value |
|---|---|
| Accuracy | 0.500 |
| Weighted F1 | 0.500 |
Intended use and limitations
Classroom exercise (24-679). A small, non-random photo sample; not validated beyond this course.
Load
import cloudpickle, huggingface_hub
path = huggingface_hub.hf_hub_download("CXu0630/2026-24679-vegetable-mixed-classifier", "autogluon_vegetable_predictor.pkl")
predictor = cloudpickle.load(open(path, "rb"))
- Task
- Image Classification
- Library
- autogluon
- Created
- Sep 23, 2026
- Updated
- Sep 23, 2026