--- dataset_info: - config_name: gemma-2-9b-it features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 435304373 num_examples: 400000 - name: validation num_bytes: 10859001 num_examples: 10000 - name: test num_bytes: 10876719 num_examples: 10000 download_size: 445389269 dataset_size: 457040093 - config_name: gemma-3-1b-it features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 435159943 num_examples: 400000 - name: validation num_bytes: 10855995 num_examples: 10000 - name: test num_bytes: 10872911 num_examples: 10000 download_size: 445229863 dataset_size: 456888849 - config_name: llama-3.1-8b-instruct features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 433544887 num_examples: 400000 - name: validation num_bytes: 10805855 num_examples: 10000 - name: test num_bytes: 10831419 num_examples: 10000 download_size: 443542903 dataset_size: 455182161 - config_name: llama-3.2-1b-instruct features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 433544887 num_examples: 400000 - name: validation num_bytes: 10805855 num_examples: 10000 - name: test num_bytes: 10831419 num_examples: 10000 download_size: 443542903 dataset_size: 455182161 - config_name: ministral-8b-instruct-2410 features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 407454567 num_examples: 400000 - name: validation num_bytes: 10166635 num_examples: 10000 - name: test num_bytes: 10173935 num_examples: 10000 download_size: 416074031 dataset_size: 427795137 - config_name: qwen2.5-1.5b-instruct features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 435953169 num_examples: 400000 - name: validation num_bytes: 10863717 num_examples: 10000 - name: test num_bytes: 10888885 num_examples: 10000 download_size: 446082255 dataset_size: 457705771 - config_name: qwen2.5-7b-instruct features: - name: _index dtype: int64 - name: _lad_clean_ids list: int64 - name: _lad_answer_mask list: bool - name: _lad_padding_mask list: bool - name: _lad_usable dtype: bool - name: _lad_empty_answer dtype: bool - name: _lad_truncated dtype: bool - name: source dtype: string - name: category dtype: string splits: - name: train num_bytes: 435953169 num_examples: 400000 - name: validation num_bytes: 10863717 num_examples: 10000 - name: test num_bytes: 10888885 num_examples: 10000 download_size: 446082255 dataset_size: 457705771 configs: - config_name: gemma-2-9b-it data_files: - split: train path: gemma-2-9b-it/train-* - split: validation path: gemma-2-9b-it/validation-* - split: test path: gemma-2-9b-it/test-* - config_name: gemma-3-1b-it data_files: - split: train path: gemma-3-1b-it/train-* - split: validation path: gemma-3-1b-it/validation-* - split: test path: gemma-3-1b-it/test-* - config_name: llama-3.1-8b-instruct data_files: - split: train path: llama-3.1-8b-instruct/train-* - split: validation path: llama-3.1-8b-instruct/validation-* - split: test path: llama-3.1-8b-instruct/test-* - config_name: llama-3.2-1b-instruct data_files: - split: train path: llama-3.2-1b-instruct/train-* - split: validation path: llama-3.2-1b-instruct/validation-* - split: test path: llama-3.2-1b-instruct/test-* - config_name: ministral-8b-instruct-2410 data_files: - split: train path: ministral-8b-instruct-2410/train-* - split: validation path: ministral-8b-instruct-2410/validation-* - split: test path: ministral-8b-instruct-2410/test-* - config_name: qwen2.5-1.5b-instruct data_files: - split: train path: qwen2.5-1.5b-instruct/train-* - split: validation path: qwen2.5-1.5b-instruct/validation-* - split: test path: qwen2.5-1.5b-instruct/test-* - config_name: qwen2.5-7b-instruct data_files: - split: train path: qwen2.5-7b-instruct/train-* - split: validation path: qwen2.5-7b-instruct/validation-* - split: test path: qwen2.5-7b-instruct/test-* license: other --- # BYOD: pretokenized 256-token training profiles Paper: [BYOD: Build Your Own DLM — Efficient Autoregressive-to-Diffusion Language Model Conversion](https://arxiv.org/abs/2610.41574) Clean token IDs and deterministic answer/padding masks for the standalone [BYOD implementation](https://github.com/RuurdKuiper/BYOD). Random masking is resampled online during training; these are not fixed noisy training inputs. Each configuration uses its own pinned tokenizer and native chat template. `prepared/CONFIG.json` records exact tokenizer/source revisions, recipe, row counts, source indices, seed, and training prefix. Use the matching configuration; token IDs are not interchangeable across models. The default publication uses the seed-42 prefix of up to 400,000 nonempty training examples and the nonempty validation/test splits for the standard 25,000-update recipe. Larger/custom training budgets require a compatible full-data publication or local preparation. The source is [Ruurd/LAD-training-1m-256](https://huggingface.co/datasets/Ruurd/LAD-training-1m-256) at revision `6041513b684d6ca627879ffdb547b8629bb8b8de`. `_index` identifies the original row in each source split; `source` and `category` retain row-level attribution. `_lad_clean_ids` and the two boolean masks are truncated exactly as in the training implementation. No model weights or tokenizer files are redistributed here. ## Terms and provenance This derivative retains the upstream mixture's restrictions and is **not** covered by the software repository's MIT license. In particular, Alpaca-derived rows carry non-commercial restrictions. See the [upstream provenance and terms](https://huggingface.co/datasets/Dovelove/BYOD-training-1m-256#upstream-repositories-and-terms). Comply with each source's terms and the applicable parent-model/tokenizer terms. Sources include Clean-Instruct-3M (no license declared on its source card), Tulu 3 (ODC-By), Alpaca-GPT4/Alpaca (CC BY-NC 4.0), MMLU/Orca-Math/GSM8K/OpenCoder (MIT), HellaSwag (MIT in its project), ARC (CC BY-SA 4.0), and MBPP (CC BY 4.0).