Views
No views yet
| Seed | AUC | MRR | NDCG@5 | NDCG@10 |
|---|---|---|---|---|
| 123 | 0.6764 | 0.3237 | 0.3593 | 0.4212 |
| 42 * | 0.6825 | 0.3285 | 0.3631 | 0.4262 |
| 456 | 0.6802 | 0.3296 | 0.3621 | 0.4245 |
| mean ± std | 0.6797±0.0025 | 0.3273±0.0026 | 0.3615±0.0016 | 0.4240±0.0021 |
1model:
2 name: caum
3 architecture:
4 news_encoder:
5 type: multi_head_self_attention
6 num_heads: 12
7 head_dim: 20
8 attention_hidden_dim: 200
9 entity_embedding_dim: 100
10 entity_num_heads: 4
11 entity_head_dim: 40
12 category_embedding_dim: 100
13 user_encoder:
14 type: candidate_aware
15 candi_selfatt:
16 num_heads: 20
17 head_dim: 20
18 candi_cnn:
19 half_window: 1
20 candi_att:
21 hidden_dim: 400
22 mid_dim: 256
23 click_predictor:
24 type: dot_product
25 embedding:
26 size: 300
27 trainable: true
28 news_dim: 400
29 use_entity: true
30 use_category: true
31 dropout_rate: 0.2
32 seed: 42
33inputs:
34 title:
35 max_length: 30
36 history:
37 max_length: 50
38 impressions:
39 max_length: 5
40 max_entities: 5
41 process_title: true
42 process_abstract: false
43 process_category: true
44 process_subcategory: false
45 process_user_id: false
46 process_entities: true
47training:
48 loss:
49 name: categorical_crossentropy
50 from_logits: true
51 reduction: sum_over_batch_size
52 label_smoothing: 0.0
53 optimizer: adam
54 learning_rate: 5.0e-05
55 batch_size: 64
56 num_epochs: 10
57 gradient_clip_val: 1.0
58 grad_accum_steps: 1
59 early_stopping:
60 patience: 3
61 min_improvement: 0.01
62 negative_sampling:
63 strategy: random
64 candidates: 4
65evaluation:
66 mode: fast
67 evaluator: caum
68 metrics:
69 - auc
70 - mrr
71 - ndcg@5
72 - ndcg@10
73 batch_size: 256newsrex/CAUM-JAX-MIND-small-bert-random/
├── 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.md1git clone https://github.com/igor17400/NewsReX.git
2cd NewsReX && uv sync
3
4# Run evaluation with best seed weights
5uv run python src/eval.py \
6 experiment=mind/glove/caum \
7 framework=jax \
8 weights=hf://newsrex/CAUM-JAX-MIND-small-bert-random/model.safetensors
9
10# Run evaluation with a specific seed
11uv run python src/eval.py \
12 experiment=mind/glove/caum \
13 framework=jax \
14 weights=hf://newsrex/CAUM-JAX-MIND-small-bert-random/seed_42/model.safetensors1@misc{newsrex2026,
2 title={NewsReX: An Open-Source Multi-Framework for Neural News Recommendation},
3 author={Igor L. R. Azevedo and Toyotaro Suzumura and Yuichiro Yasui},
4 year={2025},
5 eprint={2508.21572},
6 archivePrefix={arXiv},
7 primaryClass={cs.IR},
8 url={https://arxiv.org/abs/2508.21572},
9}