newsrex/GLORY-JAX-MIND-small

NewsReX GLORY — JAX — MIND-small

GLORY news recommendation model trained on MIND-small using the NewsReX framework (JAX).

Test Results

Seed	AUC	MRR	NDCG@5	NDCG@10
123	0.6636	0.3148	0.3482	0.4116
42 *	0.6653	0.3201	0.3533	0.4168
456	0.6583	0.3107	0.3433	0.4072
mean ± std	0.6624±0.0030	0.3152±0.0038	0.3483±0.0041	0.4119±0.0040

* Best seed (weights at repo root)

Experiment Configuration

model:
  name: glory
  architecture:
    news_encoder:
      type: mha_pool
      head_num: 20
      head_dim: 20
      attention_hidden_dim: 200
    graph_encoder:
      type: gated_graph_conv
      gnn_num_layers: 3
      use_graph_type: 0
      directed: true
      k_hops: 2
      num_neighbors: 8
      entity_neighbors: 10
    click_predictor:
      type: dot_product
  embedding:
    size: 300
    trainable: true
  dropout_rate: 0.2
  seed: 42
  use_entity: true
  entity_emb_dim: 100
inputs:
  title:
    max_length: 30
  entity:
    max_length: 5
  history:
    max_length: 50
  impressions:
    max_length: 5
  process_title: true
  process_abstract: false
  process_category: true
  process_subcategory: true
  process_entities: true
  process_user_id: false
training:
  loss:
    name: categorical_crossentropy
    from_logits: true
    reduction: sum_over_batch_size
    label_smoothing: 0.0
  optimizer: adam
  learning_rate: 0.0002
  batch_size: 32
  grad_accum_steps: 1
  num_epochs: 20
  gradient_clip_val: 1.0
  early_stopping:
    patience: 5
    min_improvement: 0.01
  negative_sampling:
    strategy: random
    candidates: 4
evaluation:
  mode: fast
  evaluator: default
  metrics:
  - auc
  - mrr
  - ndcg@5
  - ndcg@10
  batch_size: 256

Repository Structure

newsrex/GLORY-JAX-MIND-small/
├── model.safetensors          ← best seed (42)
├── test_results.json
├── training_run_summary.json
├── seed_123/model.safetensors
├── seed_42/model.safetensors
├── seed_456/model.safetensors
└── README.md

Usage

git clone https://github.com/igor17400/NewsReX.git
cd NewsReX && uv sync

# Run evaluation with best seed weights
uv run python src/eval.py \
    experiment=mind/glory \
    framework=jax \
    weights=hf://newsrex/GLORY-JAX-MIND-small/model.safetensors

# Run evaluation with a specific seed
uv run python src/eval.py \
    experiment=mind/glory \
    framework=jax \
    weights=hf://newsrex/GLORY-JAX-MIND-small/seed_42/model.safetensors

Citation

@misc{newsrex2026,
  title={NewsReX: An Open-Source Multi-Framework for Neural News Recommendation},
  author={Igor L. R. Azevedo and Toyotaro Suzumura and Yuichiro Yasui},
  year={2025},
  eprint={2508.21572},
  archivePrefix={arXiv},
  primaryClass={cs.IR},
  url={https://arxiv.org/abs/2508.21572},
}