combined_no_america_with_metadata_1b_step4k

combined_no_america_with_metadata_1b_step4k

Released with the paper "MAPLE: Metadata Conditioned LLM Pretraining for Locale-Aware Question Answering" (EMNLP 2026).

Paper (EMNLP 2026) | arXiv | Project page | Code | LocalNewsQA

Summary

This repo contains the leave out america 1b step4k model exported from the 4k checkpoint for the metadata localization project. It was trained from scratch on the project corpus, using the Llama 3.2 tokenizer and vocabulary.

Variant Metadata

• Stage: pretrain
• Family: leave_one_out
• Size: 1b
• Metadata condition: with_metadata
• Checkpoint export: 4k
• Base model lineage: Trained from scratch; tokenizer/vocabulary from meta-llama/Llama-3.2-1B

Weights & Biases Provenance

• Run name: 20/12/2025_09:26:32_combined_no_america_with_metadata_1b
• Internal run URL: https://wandb.ai/iamshnoo/nanotron/runs/kane1qyw
• Note: the Weights & Biases workspace is private; public readers should use the summarized metrics and configuration below.
• State: finished
• Runtime: 114h 52m 10s

Run Summary

• KPI/train_lm_loss: 1.9947
• KPI/train_perplexity: 7.3496
• KPI/val_loss: 2.0776
• KPI/val_perplexity: 7.9855
• KPI/consumed_tokens/train: 41,943,040,000
• _step: 10,000

Training Configuration

• train_steps: 10,000
• sequence_length: 2,048
• micro_batch_size: 8
• batch_accumulation_per_replica: 64
• learning_rate: 0.003
• min_decay_lr: 0.0003
• checkpoint_interval: 1,000

Training Curves

Static plots below were exported from the private Weights & Biases run and embedded here for public access.

Train Loss

Validation Perplexity

Throughput

Project Context

This model is part of the metadata localization release. Related checkpoints and variants are grouped in the public Hugging Face collection Metadata Conditioned LLMs.

• Training data source: News on the Web (NOW) Corpus
• Project repository: https://github.com/iamshnoo/metadata_localization
• Paper: MAPLE, EMNLP 2026, also on arXiv at https://arxiv.org/abs/2601.15236
• Project page: https://iamshnoo.github.io/metadata_localization/

Citation

@inproceedings{mukherjee-etal-2026-maple,
    title = "{MAPLE}: Metadata Conditioned {LLM} Pretraining for Locale-Aware Question Answering",
    author = "Mukherjee, Anjishnu  and
      Zhu, Ziwei  and
      Anastasopoulos, Antonios",
    editor = "Lee, Hung-Yi  and
      Dev, Sunipa  and
      Li, Junyi Jessy  and
      Elliott, Desmond",
    booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2026",
    address = "Budapest, Hungary",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.emnlp-main.2053/",
    doi = "10.18653/2026.emnlp-main.2053",
    pages = "46746--46772",
    ISBN = "979-8-89176-459-0"
}

Last synced: 2026-04-02 14:45:35 UTC