{
  "total": 225,
  "topics": [
    {
      "id": 685051,
      "title": "Congratulations and Thank You All!",
      "comment_count": 1,
      "views": 0,
      "votes": 19,
      "reused": true
    },
    {
      "id": 684963,
      "title": "Recap of Competition - Congratulations to the Winners!",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 663210,
      "title": "Welcome to the Deep Past Initiative Machine Translation Challenge!",
      "comment_count": 27,
      "views": 0,
      "votes": 20,
      "reused": true
    },
    {
      "id": 678899,
      "title": "A Stitch in Time Saves Nine",
      "comment_count": 71,
      "views": 0,
      "votes": 36,
      "reused": true
    },
    {
      "id": 665209,
      "title": "Two practical stumbling blocks in Akkadian → English MT (and how to address them)",
      "comment_count": 54,
      "views": 0,
      "votes": 75,
      "reused": true
    },
    {
      "id": 674136,
      "title": "Dataset Update - Mind the Gaps",
      "comment_count": 46,
      "views": 0,
      "votes": 26,
      "reused": true
    },
    {
      "id": 664177,
      "title": "Data Update",
      "comment_count": 33,
      "views": 0,
      "votes": 44,
      "reused": true
    },
    {
      "id": 663183,
      "title": "How to get started + Competition's Official Discord",
      "comment_count": 2,
      "views": 0,
      "votes": 7,
      "reused": true
    },
    {
      "id": 684381,
      "title": "[20th Place] ByT5-base, span corruption, synth & soup",
      "comment_count": 1,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 684345,
      "title": "[2nd Place] Data-Centric Akkadian NMT",
      "comment_count": 16,
      "views": 0,
      "votes": 42,
      "reused": true
    },
    {
      "id": 704840,
      "title": "Ensemble with Weighted MBR for Akkadian Translation",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 699983,
      "title": "What Actually Moved the Needle (and What Didn't) _NLLB-200 +LORA",
      "comment_count": 0,
      "views": 0,
      "votes": -1
    },
    {
      "id": 684847,
      "title": "[5th] Synthetic Data via Back-Translation",
      "comment_count": 4,
      "views": 0,
      "votes": 13,
      "reused": true
    },
    {
      "id": 684353,
      "title": "[DPC 1st] Data Quality Dictates Everything",
      "comment_count": 28,
      "views": 0,
      "votes": 93,
      "reused": true
    },
    {
      "id": 684910,
      "title": "[32nd] ByT5-base Fine-tuning | No LLMs, No Synthetic Data, No Ensembles",
      "comment_count": 5,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 686956,
      "title": "Deep Past Challenge | Top-13 Solution",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 686304,
      "title": "[59th Place, Silver] ByT5-base + LLM Data Augmentation + Cross-Model MBR Ensembl",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 685928,
      "title": "[9th] Place Solution for the Deep Past Challenge Competition",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 685925,
      "title": "Fourth Place Solution Writeup",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 684231,
      "title": "[6th] DPC solution - 15 models ensemble on diverse data sources",
      "comment_count": 12,
      "views": 0,
      "votes": 54,
      "reused": true
    },
    {
      "id": 684329,
      "title": "8th Place Solution: 2-Stage Fine-tuning + High Quality Data Extraction",
      "comment_count": 1,
      "views": 0,
      "votes": 25,
      "reused": true
    },
    {
      "id": 685251,
      "title": "160th Place Solution: +218 Shake Up",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 684425,
      "title": "[3rd] Synthetic Data to Teach OA Fundamentals",
      "comment_count": 11,
      "views": 0,
      "votes": 51,
      "reused": true
    },
    {
      "id": 684236,
      "title": "LLM + Online Learning is all you need",
      "comment_count": 13,
      "views": 0,
      "votes": 23,
      "reused": true
    },
    {
      "id": 684938,
      "title": "Could the final leaderboard scores be shown to 2–3 decimal places after the competition ends?",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 684866,
      "title": "[12th] NotebookLM & Single ByT5-Base Model",
      "comment_count": 1,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 684874,
      "title": "Akkadian Language & Its Link to the Assyrian Empire",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 684819,
      "title": "15th Place Gold Solution",
      "comment_count": 2,
      "views": 0,
      "votes": 16,
      "reused": true
    },
    {
      "id": 684537,
      "title": "Silver Medal Solution[29th]: ByT5-XL Custom CPT with Dictionary Data + KL-Anchor",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 684514,
      "title": "Silver Medal Solution[29th]: ByT5-XL Custom CPT with Dictionary Data + KL-Anchored LoRA",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 684429,
      "title": "11th Place Solution",
      "comment_count": 0,
      "views": 0,
      "votes": 13
    },
    {
      "id": 684422,
      "title": "One-Shot Silver Solution!",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 684419,
      "title": "ByT5-Large with word-level DAPT",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 684408,
      "title": "Final Submissions",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 684368,
      "title": "[25th Place] DPC Solution: Contextual ByT5-Large Ensemble, Chunking, and MBR",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 684332,
      "title": "17th Place Solution — Deep Past Challenge ",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 684215,
      "title": "Short 7th place note",
      "comment_count": 8,
      "views": 0,
      "votes": 34,
      "reused": true
    },
    {
      "id": 684211,
      "title": "10th Place Solution — Seq2Seq + CPT + Pseudo-Labeling",
      "comment_count": 2,
      "views": 0,
      "votes": 21,
      "reused": true
    },
    {
      "id": 684206,
      "title": "[22nd] CPT, GRPO and Extensive Work on the Data (No SFT)",
      "comment_count": 2,
      "views": 0,
      "votes": 23,
      "reused": true
    },
    {
      "id": 684278,
      "title": "89th Place - Solution",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 684274,
      "title": "68th Place Solution🥈",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 684213,
      "title": "#[27th] Solution: ByT5-base + Gemini-3.0 + Ensemble",
      "comment_count": 2,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 684270,
      "title": "[33rd] Single byt5-base model",
      "comment_count": 0,
      "views": 0,
      "votes": 11
    },
    {
      "id": 684267,
      "title": "[21th] ByT5-Base + Gemini Augmentation, No Ensemble",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 684265,
      "title": "36th Place Solution - LLM-Based Sentence Alignment Strategy + MBR Model Ensembli",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 684203,
      "title": "Multi-phase curriculum training ",
      "comment_count": 1,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 684249,
      "title": "AKT PDFs Processing using PaddleOCR-VL",
      "comment_count": 1,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 684242,
      "title": "Faster decoding",
      "comment_count": 2,
      "views": 0,
      "votes": 15,
      "reused": true
    },
    {
      "id": 684246,
      "title": "2-Stage CV, CPT & MBR Decoding",
      "comment_count": 0,
      "views": 0,
      "votes": 5
    },
    {
      "id": 684243,
      "title": "[50th] Solution: ByT5-base + TAPT + Ensemble",
      "comment_count": 0,
      "views": 0,
      "votes": 7
    },
    {
      "id": 684235,
      "title": "72th Place Solution",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 684187,
      "title": "Make the most of all available resources",
      "comment_count": 1,
      "views": 0,
      "votes": 13,
      "reused": true
    },
    {
      "id": 684189,
      "title": "[24th] Post-training Qwen2.5 32B and 72B with Gemini OCR/published texts pairs",
      "comment_count": 7,
      "views": 0,
      "votes": 36,
      "reused": true
    },
    {
      "id": 684116,
      "title": "Competition Metrics",
      "comment_count": 14,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 684202,
      "title": "Identical Submission Produced Different Public/Private Scores",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 684064,
      "title": "Notebooks Rerun status",
      "comment_count": 5,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 683666,
      "title": "Notebook Timeout Issue – Runtime Much Longer Than Expected?",
      "comment_count": 10,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 683786,
      "title": "Question about runtimes during private runs",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 683748,
      "title": "Question about Leaderboard precision and handling of tied scores",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 680420,
      "title": "Notebook Submission not throwing errors but also stuck in running",
      "comment_count": 11,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 682988,
      "title": "How much do you think the rankings on LB will shaking in this match?",
      "comment_count": 9,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 683713,
      "title": "Deal Garbled Text",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 683352,
      "title": "OA UnifiedGloss",
      "comment_count": 1,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 683536,
      "title": "can not upload submission file",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 683541,
      "title": "submission Error sudden appear",
      "comment_count": 4,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 683589,
      "title": "Guide to fix Accelerate error",
      "comment_count": 0,
      "views": 0,
      "votes": 6
    },
    {
      "id": 682467,
      "title": "unable to submit after Entry Deadline – Is a prior submission required?",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 683329,
      "title": "Issue regarding training speed 0.06 it/s",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 682707,
      "title": "Geophysics meets Akkadian- what we did. ",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 681923,
      "title": "Could anybody please share old variant of train.csv?",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 682321,
      "title": "is possible shake up?",
      "comment_count": 1,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 681514,
      "title": "How much data from the additional PDFs did you use?",
      "comment_count": 3,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 681565,
      "title": "Negative Learning Experience – Successful Failure",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 681734,
      "title": "dictionary-based pre-training",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 681603,
      "title": "[Doubt] -- Fork from Public Notebook",
      "comment_count": 3,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 681562,
      "title": "Submission- 9h scoring- fails",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 681486,
      "title": "original dataset?",
      "comment_count": 3,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 680041,
      "title": "0.78% Train Transliterations Found in eBL_Dictionary?",
      "comment_count": 4,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 680686,
      "title": "Regarding the open-source issue that is malignant and offers no nutritional value.",
      "comment_count": 13,
      "views": 0,
      "votes": 45,
      "reused": true
    },
    {
      "id": 681543,
      "title": "RAG is Needed?",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 681446,
      "title": "Why do the submitted tasks always time out?   internet out",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 680699,
      "title": "Long Submission Queue and Unexpected Timeouts",
      "comment_count": 5,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 679336,
      "title": "Massive bot attack",
      "comment_count": 18,
      "views": 0,
      "votes": 13,
      "reused": true
    },
    {
      "id": 681431,
      "title": "train.csv translation don't align with pdf version",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 681441,
      "title": "Why do the submitted tasks always time out?",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 681421,
      "title": "Cannot Publish The Notebook",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 680874,
      "title": "Local validation vs public LB mismatch after recent updates — clarification on normalization/eval details?",
      "comment_count": 3,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 681149,
      "title": "how to create a Dependency Installation Script",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 680889,
      "title": "https://www.deeppast.org/oare does not exist",
      "comment_count": 1,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 680612,
      "title": "[Operation issues] Why got stuck in Queued",
      "comment_count": 7,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 680446,
      "title": "(pdf)-Translation sometimes different from train.csv translation",
      "comment_count": 1,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 680256,
      "title": "eBL text divergence",
      "comment_count": 1,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 668621,
      "title": "Translate Gemma",
      "comment_count": 4,
      "views": 0,
      "votes": 8,
      "reused": true
    },
    {
      "id": 680173,
      "title": "Issue with submitting",
      "comment_count": 4,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 679878,
      "title": "Question about Score Discrepancies",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 673904,
      "title": "Insights from the Akkademia Codebase & PNAS Paper for the Deep Past Challenge",
      "comment_count": 5,
      "views": 0,
      "votes": 17,
      "reused": true
    },
    {
      "id": 679830,
      "title": "Question about the actual Team Merger Deadline: 3/10 or 3/16?",
      "comment_count": 4,
      "views": 0,
      "votes": 12,
      "reused": true
    },
    {
      "id": 679714,
      "title": "Low LB score",
      "comment_count": 2,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 679497,
      "title": "Lora on ByT5 large",
      "comment_count": 1,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 679867,
      "title": "BLEU and ChrF scores along with GM",
      "comment_count": 0,
      "views": 0,
      "votes": 3
    },
    {
      "id": 678974,
      "title": "Mid-competition updates have now become a trend in Kaggle competitions.",
      "comment_count": 6,
      "views": 0,
      "votes": 25,
      "reused": true
    },
    {
      "id": 679766,
      "title": "Old Assyrian Tokens",
      "comment_count": 1,
      "views": 0,
      "votes": 7,
      "reused": true
    },
    {
      "id": 679641,
      "title": "GPU for Local Training",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 663357,
      "title": "Other Public Data",
      "comment_count": 9,
      "views": 0,
      "votes": 31,
      "reused": true
    },
    {
      "id": 678829,
      "title": "Notebook Timeout",
      "comment_count": 5,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 666498,
      "title": "First competition",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 679190,
      "title": "competiton-aligned-data-cleaning nd baseline",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 679086,
      "title": "Failed Submissions Since Update",
      "comment_count": 3,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 679077,
      "title": "Clarification of Rules",
      "comment_count": 1,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 678836,
      "title": "Is this competition becoming a 'Regex Guessing Game'?",
      "comment_count": 6,
      "views": 0,
      "votes": 23,
      "reused": true
    },
    {
      "id": 674469,
      "title": "Why 300-350 people are stuck at 35.1 score?",
      "comment_count": 25,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 663650,
      "title": "Ground Truth of 4 examples in sample test?",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 674843,
      "title": "A small doubt related to private test set",
      "comment_count": 1,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 674944,
      "title": "Submission CSV Not Found anyone got this",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 668402,
      "title": "Compiled Discussions To Read (Avoid Bad Advice)",
      "comment_count": 22,
      "views": 0,
      "votes": 93,
      "reused": true
    },
    {
      "id": 674708,
      "title": "How should we tokenize?",
      "comment_count": 1,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 674468,
      "title": "line_start and line_end in test.csv",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 674620,
      "title": "Segmentation on train.csv",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 674342,
      "title": "Cleaning Translations in Training vs Test Data",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 674348,
      "title": "Cuneiform base model",
      "comment_count": 0,
      "views": 0,
      "votes": 14
    },
    {
      "id": 673932,
      "title": "Phonetic Transcription for Uppercase Logograms/Sumerograms",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 673853,
      "title": "Part I of book - Mesopotamia: The Old Assyrian Period (OBO 160/5, 2008) by Klaas R. Veenhof",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 673804,
      "title": "training data issues",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 673352,
      "title": "Why does everyone use the “final-byt5” model?",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 672511,
      "title": "Akkadian Translation Competition - Community Knowledge Synthesis",
      "comment_count": 14,
      "views": 0,
      "votes": 45,
      "reused": true
    },
    {
      "id": 664948,
      "title": "The translation column of the final test predictions into the corresponding <gap> and <big_gap> tags",
      "comment_count": 65,
      "views": 0,
      "votes": 55,
      "reused": true
    },
    {
      "id": 673577,
      "title": "Questions about publications.csv and published_texts.csv",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 673455,
      "title": "Public tool for easier OCR of PDF documents",
      "comment_count": 3,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 670040,
      "title": "Any luck with letters from Larsen PDF?",
      "comment_count": 19,
      "views": 0,
      "votes": 22,
      "reused": true
    },
    {
      "id": 668619,
      "title": "Probing the Best Leaderboard Public Models",
      "comment_count": 12,
      "views": 0,
      "votes": 42,
      "reused": true
    },
    {
      "id": 670393,
      "title": "Increase the Train dataset",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 666431,
      "title": "Use Sentences_Oare_FirstWord_LinNum to extract the sentence alignment",
      "comment_count": 5,
      "views": 0,
      "votes": 10,
      "reused": true
    },
    {
      "id": 672392,
      "title": "Seeking help",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 672129,
      "title": "Helper signals from the provided lexicon",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 672121,
      "title": "Month representation",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 664106,
      "title": "Could you please disclose the second or third decimal place?",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 671781,
      "title": "Strategy for Sentence Alignment and Normalization",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 665886,
      "title": "cv vs lb gap issues",
      "comment_count": 8,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 671453,
      "title": "About sentencealignment",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 671458,
      "title": "How are you handling Normalization and Sentence Alignment?",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 671402,
      "title": "Difficulty understanding additional data creation suggestion",
      "comment_count": 1,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 671381,
      "title": "Unable to Submit",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 671070,
      "title": "Question on using Meta NLLB models under the winner license terms",
      "comment_count": 0,
      "views": 0,
      "votes": 4
    },
    {
      "id": 670783,
      "title": "Clarification on I–IV and A–D Dataset Labels",
      "comment_count": 4,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 670874,
      "title": "Question about Modern Scribal Notations in the test set ground truth translation",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 670548,
      "title": "Importance of punctuation marks in translations",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 670489,
      "title": "There are missing matches between train.csv and Sentences_Oare_FirstWord_LinNum.csv.",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 670467,
      "title": "PDF Dataset",
      "comment_count": 1,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 670278,
      "title": "Thoughts on the test set. ",
      "comment_count": 11,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 670373,
      "title": "Big disconnect",
      "comment_count": 0,
      "views": 0,
      "votes": -2
    },
    {
      "id": 670084,
      "title": "How were the 34.5 LB baselines exactly trained?",
      "comment_count": 6,
      "views": 0,
      "votes": 15,
      "reused": true
    },
    {
      "id": 670130,
      "title": "Computation required",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 670046,
      "title": "English pre-processing ",
      "comment_count": 0,
      "views": 0,
      "votes": 2
    },
    {
      "id": 669544,
      "title": "Should we use pre-trained models or train them ourselves?",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 668975,
      "title": "Share the idea to retrain the tokenizer & model",
      "comment_count": 1,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 668812,
      "title": "Preprocessing",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 669407,
      "title": "[BUG] In the Data Tab",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 669062,
      "title": "Translation Alignment",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 667480,
      "title": "Floating Point Number Questions",
      "comment_count": 18,
      "views": 0,
      "votes": 18,
      "reused": true
    },
    {
      "id": 669163,
      "title": "Encoding mismatches in test.csv",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 669151,
      "title": "Training from scratch vs finetuning existing model",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 669055,
      "title": "Does this competition not require training models?",
      "comment_count": 3,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 668990,
      "title": "Tips for Fine-Tuning ByT5-Small Efficiently on Local GPU",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 664518,
      "title": "Clarification: Gap Format in Test Set (<gap> vs <big_gap>) - Did I understand correctly?",
      "comment_count": 11,
      "views": 0,
      "votes": 19,
      "reused": true
    },
    {
      "id": 668766,
      "title": "Ground truth examples needed",
      "comment_count": 8,
      "views": 0,
      "votes": -1,
      "reused": true
    },
    {
      "id": 668723,
      "title": "Processing asymmetric gaps in the ground truth",
      "comment_count": 2,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 667775,
      "title": "AICC_translation怎么使用？",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 663804,
      "title": "Some questions.",
      "comment_count": 7,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 668205,
      "title": "Test set lines and `text_id`",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 668327,
      "title": "Discrepancy: OA_Lexicon Standardized PNs vs. Training Set Ground Truth ",
      "comment_count": 4,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 668339,
      "title": "Calculate score without submission ",
      "comment_count": 4,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 667483,
      "title": "Significant number (up to 25%) of incorrectly truncated texts in train.csv",
      "comment_count": 5,
      "views": 0,
      "votes": 14,
      "reused": true
    },
    {
      "id": 667878,
      "title": "Write-up / extra data",
      "comment_count": 4,
      "views": 0,
      "votes": 7,
      "reused": true
    },
    {
      "id": 667542,
      "title": "What's the best way to use OA_Lexicon_eBL.csv?",
      "comment_count": 5,
      "views": 0,
      "votes": 7,
      "reused": true
    },
    {
      "id": 663945,
      "title": "Genesis Protocol ",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 667773,
      "title": "Can we drop accents?",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 667779,
      "title": "Asking if we can import notebook configured in jupyter notebook and submit",
      "comment_count": 1,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 667004,
      "title": "Compiling all strategies in the PyTorch only Training NB",
      "comment_count": 2,
      "views": 0,
      "votes": 10,
      "reused": true
    },
    {
      "id": 663839,
      "title": "How do quotations in the translations work?",
      "comment_count": 6,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 666798,
      "title": "Is test set's transliteration formatted the same as the train set",
      "comment_count": 6,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 667001,
      "title": "Lines with Apostrophes",
      "comment_count": 1,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 667065,
      "title": "Beginner's Guide & Baseline Approaches",
      "comment_count": 1,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 665928,
      "title": "Formatting suggestions",
      "comment_count": 5,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 666745,
      "title": "Improving Alignment Between Akkadian Text and English Translation",
      "comment_count": 8,
      "views": 0,
      "votes": 19,
      "reused": true
    },
    {
      "id": 666916,
      "title": "Missing personal names in OA_Lexicon_eBL.csv and possible fixes for 50% of them",
      "comment_count": 0,
      "views": 0,
      "votes": 8
    },
    {
      "id": 666443,
      "title": "How much can one use LLMs without violating the rule on Data Security?",
      "comment_count": 5,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 666690,
      "title": "How Does the competition scores ?",
      "comment_count": 2,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 666740,
      "title": "Scoring metrics",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 666727,
      "title": "metric issues",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 666507,
      "title": "How to treat blank data in eBL_Dictionary.csv",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 665949,
      "title": "How to handle dot h/H?",
      "comment_count": 9,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 663575,
      "title": "Discrepancies in logogram transliteration",
      "comment_count": 7,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 666260,
      "title": "Extra data set from Akkadian AI translation",
      "comment_count": 7,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 666426,
      "title": "Whats up with Sentences_Oare_FirstWord_LinNum.csv?",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 666425,
      "title": "Academic Landscape of Old Akkadian: Chicago vs Munich vs French schools",
      "comment_count": 0,
      "views": 0,
      "votes": 9
    },
    {
      "id": 663576,
      "title": "Another Public source of data (Michel Old Assyrian Merchant Letters)",
      "comment_count": 2,
      "views": 0,
      "votes": 11,
      "reused": true
    },
    {
      "id": 665696,
      "title": "[EDA] Critical Finding: Train-Test Transliteration Convention Mismatch",
      "comment_count": 10,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 665490,
      "title": "Hey I cannot submit my results ",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 663553,
      "title": "New to the Competition? Let’s Learn Akkadian Translation Together! (Beginner-Friendly Thread)",
      "comment_count": 5,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 665250,
      "title": "Discrepancy between Validation Loss and Metric when increasing Batch Size",
      "comment_count": 3,
      "views": 0,
      "votes": null,
      "reused": true
    },
    {
      "id": 665366,
      "title": "Technical Summary: Overcoming Mode Collapse in Akkadian Translation",
      "comment_count": 0,
      "views": 0,
      "votes": 1
    },
    {
      "id": 664985,
      "title": "Some more data",
      "comment_count": 4,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 665101,
      "title": "Unicode Fractions vs Decimals: LB Score Unchanged?",
      "comment_count": 8,
      "views": 0,
      "votes": 10,
      "reused": true
    },
    {
      "id": 664795,
      "title": "How To Handle These Examples",
      "comment_count": 7,
      "views": 0,
      "votes": 24,
      "reused": true
    },
    {
      "id": 665102,
      "title": "Finetuned model generates repeated gibberish",
      "comment_count": 3,
      "views": 0,
      "votes": 5,
      "reused": true
    },
    {
      "id": 665074,
      "title": "Model finetuning",
      "comment_count": 1,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 664981,
      "title": "BF16 vs FP32:Full Precision significantly outperforming BF16",
      "comment_count": 4,
      "views": 0,
      "votes": 12,
      "reused": true
    },
    {
      "id": 664411,
      "title": "Formatting of the Hidden Test Set",
      "comment_count": 17,
      "views": 0,
      "votes": 24,
      "reused": true
    },
    {
      "id": 664905,
      "title": "Question about OA_Lexicon_eBL.csv - Personal Name Spelling Inconsistency with Ground Truth",
      "comment_count": 3,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 664775,
      "title": "Mass processing of OCR texts using ChatGPT, Claude, etc.",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 664644,
      "title": "Beginner questions",
      "comment_count": 1,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 664487,
      "title": "My GPU quota has been exhausted",
      "comment_count": 2,
      "views": 0,
      "votes": 4,
      "reused": true
    },
    {
      "id": 664028,
      "title": "Is the oare_id in test data appears in the published_texts?",
      "comment_count": 2,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 663849,
      "title": "Incomplete translations",
      "comment_count": 2,
      "views": 0,
      "votes": 19,
      "reused": true
    },
    {
      "id": 663900,
      "title": "A nice podcast with Lex Fridman related to Cuneiforms",
      "comment_count": 1,
      "views": 0,
      "votes": 6,
      "reused": true
    },
    {
      "id": 663998,
      "title": "CDLI/ORACC overlap",
      "comment_count": 2,
      "views": 0,
      "votes": 2,
      "reused": true
    },
    {
      "id": 664079,
      "title": " Post-processing by LLM",
      "comment_count": 4,
      "views": 0,
      "votes": 21,
      "reused": true
    },
    {
      "id": 664195,
      "title": "Data Quality Issues (publications.csv)",
      "comment_count": 1,
      "views": 0,
      "votes": 1,
      "reused": true
    },
    {
      "id": 663388,
      "title": "Access to https://www.ebl.lmu.de/dictionary data",
      "comment_count": 8,
      "views": 0,
      "votes": 13,
      "reused": true
    },
    {
      "id": 663233,
      "title": "Akkadian to English: Cuneify tool, Fairseq toolkit, Python Package Akkademia and Gale-Church algorithm.",
      "comment_count": 7,
      "views": 0,
      "votes": 35,
      "reused": true
    },
    {
      "id": 663656,
      "title": "Usage of Synthetic Data",
      "comment_count": 2,
      "views": 0,
      "votes": 9,
      "reused": true
    },
    {
      "id": 663610,
      "title": "Advice Needed: Should an AI Beginner Join This Competition or Focus on Learning First?",
      "comment_count": 6,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 663856,
      "title": "Request to Enable TPU Support",
      "comment_count": 0,
      "views": 0,
      "votes": 0
    },
    {
      "id": 663444,
      "title": "Transliteration done implies that more than half of the work has been done ",
      "comment_count": 4,
      "views": 0,
      "votes": 3,
      "reused": true
    },
    {
      "id": 663505,
      "title": "Starter code and some tips",
      "comment_count": 3,
      "views": 0,
      "votes": 15,
      "reused": true
    }
  ],
  "errors": []
}