Dataset Viewer
Auto-converted to Parquet Duplicate
The dataset viewer is not available for this split.
Job manager crashed while running this job (missing heartbeats).
Error code:   JobManagerCrashedError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Uchen vs Umê — splits, benchmark, Parquet, and inference

Model: openpecha/uchen-ume-classifier

Parquet dataset

Hub: openpecha/uchen-ume-classification-benchmark

File Rows Columns
train.parquet 9,110 id, image_bytes, script
validation.parquet 1,000 same
test.parquet 851 same
benchmark.parquet 60 same
from datasets import load_dataset

repo = "openpecha/uchen-ume-classification-benchmark"
train = load_dataset(repo, split="train")
bench = load_dataset(repo, split="benchmark")

Training preprocess (verified in train.py)

Split with_preprocess without_preprocess
train center crop none
val center crop none
test none (full page) none

Benchmark inference should use --preprocess none for both models when comparing to the test split. Center-cropping at inference on with_preprocess inflates benchmark scores (~98% acc) because it matches val, not test.

Benchmark results (60 images, full pages, preprocess none)

Model Accuracy Macro-F1 AUC
without_preprocess 85.0% 0.848 0.970
with_preprocess 68.3% 0.648 0.953

Test split (867 full pages): without_preprocess 80.7% / 0.708; with_preprocess 56.1% / 0.506.

Inference

pip install -r requirements-inference.txt

Benchmark — recommended (full pages)

python inference_uchen_ume.py \
  --benchmark-dir benchmark \
  --model-repo openpecha/uchen-ume-classifier \
  --weights without_preprocess/final_model.pt \
  --preprocess none \
  --output benchmark_predictions.json

Benchmark — with_preprocess (same as training test: no crop)

python inference_uchen_ume.py \
  --benchmark-dir benchmark \
  --weights with_preprocess/final_model.pt \
  --preprocess none

Benchmark from JSON + URLs

python inference_uchen_ume.py \
  --benchmark-json benchmark/benchmark_holdout.json \
  --fetch-urls \
  --preprocess none

Single image

python inference_uchen_ume.py --image page.jpg --weights without_preprocess/final_model.pt --preprocess none

JSON manifests

  • splits/train_val_test_splits.json — pages + image_url
  • benchmark/benchmark_holdout.json — 60 holdout pages + image_url
Downloads last month
9,527

Models trained or fine-tuned on openpecha/uchen-ume-classification-benchmark