| import json,hashlib,re,unicodedata,collections,time |
| from pathlib import Path |
| from datasets import load_dataset,Dataset |
| from huggingface_hub import HfApi |
| api=HfApi();source='Helsinki-NLP/opus-100';revision=api.dataset_info(source).sha |
| out=Path('/tmp/AstheralDataset');(out/'data').mkdir(parents=True,exist_ok=True) |
| langs={'de':'German','fr':'French','es':'Spanish','it':'Italian','pt':'Portuguese'} |
| seen_values=set();ids=set();splits={'train':[],'validation':[],'test':[]};rejected=collections.Counter() |
| def normalize(s): |
| s=unicodedata.normalize('NFKC',s).casefold() |
| return ' '.join(''.join(c if c.isalnum() else ' ' for c in s).split()) |
| for split,n in [('test',50),('validation',50),('train',900)]: |
| for lang,name in langs.items(): |
| config='-'.join(sorted(['en',lang]));ds=load_dataset(source,config,split=split,revision=revision).shuffle(seed=20261113) |
| added=0 |
| for index,row in enumerate(ds): |
| a=row['translation']['en'].strip();b=row['translation'][lang].strip() |
| if not(20<=len(a)<=500 and 12<=len(b)<=600) or len(a.split())<4 or len(b.split())<3: |
| rejected['length']+=1;continue |
| if any(re.search(r'lorem ipsum|placeholder|\bTODO\b|https?://|<[^>]+>|\{\{|\ufffd',s,re.I) for s in [a,b]): |
| rejected['non_sentence_content']+=1;continue |
| x,y=normalize(a),normalize(b) |
| if x==y or x in seen_values or y in seen_values: |
| rejected['duplicate_or_identical_text']+=1;continue |
| uid=hashlib.sha256((lang+'\\0'+x+'\\0'+y).encode()).hexdigest() |
| assert uid not in ids |
| ids.add(uid);seen_values.update([x,y]) |
| record={'id':uid,'source_language':'en','target_language':lang,'source_text':a,'target_text':b,'instruction':'Translate from English to '+name+'. Return only the translation.','source_dataset':source,'source_config':config,'source_split':split,'source_revision':revision,'source_row_index':int(ds._indices.column(0)[index].as_py()) if ds._indices is not None else index} |
| splits[split].append(record);added+=1 |
| if added==n:break |
| assert added==n,(lang,split,added) |
| print('COLLECTED',split,lang,added,flush=True) |
| for split,rows in splits.items(): |
| Dataset.from_list(rows).shuffle(seed=42).to_parquet(str(out/'data'/f'{split}.parquet')) |
| total=sum(map(len,splits.values()));assert total==5000 and len(ids)==5000 and len(seen_values)==10000 |
| counter={split:dict(collections.Counter(r['target_language'] for r in rows)) for split,rows in splits.items()} |
| manifest={'name':'AstheralDataset','version':'1.0','total_examples':total,'unique_example_ids':len(ids),'unique_normalized_text_values':len(seen_values),'split_counts':{k:len(v) for k,v in splits.items()},'language_counts':counter,'source_dataset':source,'source_revision':revision,'normalization':'Unicode NFKC, casefold, punctuation removed, whitespace collapsed','deduplication':'Global source/target text uniqueness across all languages and splits; identical source-target pairs excluded','rejected':dict(rejected),'file_sha256':{p.name:hashlib.sha256(p.read_bytes()).hexdigest() for p in (out/'data').glob('*.parquet')}} |
| (out/'manifest.json').write_text(json.dumps(manifest,indent=2)) |
| (out/'build_dataset.py').write_text(Path(__file__).read_text()) |
| readme='''--- |
| pretty_name: AstheralDataset |
| language: |
| - en |
| - de |
| - fr |
| - es |
| - it |
| - pt |
| task_categories: |
| - translation |
| size_categories: |
| - 1K<n<10K |
| license: other |
| license_name: source-corpus-terms |
| license_link: https://huggingface.co/datasets/Helsinki-NLP/opus-100 |
| configs: |
| - config_name: default |
| data_files: |
| - split: train |
| path: data/train.parquet |
| - split: validation |
| path: data/validation.parquet |
| - split: test |
| path: data/test.parquet |
| --- |
| # AstheralDataset |
| |
| **5,000 real English translation pairs**, curated by SEE7 for German, French, Spanish, Italian and Portuguese. |
| |
| ## Contents |
| |
| | Target | Train | Validation | Test | Total | |
| | --- | ---: | ---: | ---: | ---: | |
| | German | 900 | 50 | 50 | 1,000 | |
| | French | 900 | 50 | 50 | 1,000 | |
| | Spanish | 900 | 50 | 50 | 1,000 | |
| | Italian | 900 | 50 | 50 | 1,000 | |
| | Portuguese | 900 | 50 | 50 | 1,000 | |
| | **Total** | **4,500** | **250** | **250** | **5,000** | |
| |
| ## Record fields |
| |
| Each row contains a unique SHA-256 ID, English source text, translated target text, source and target language codes, a translation instruction, and the original dataset/config/split/revision/row index. |
| |
| ## Curation |
| |
| Text is drawn from real parallel sentences in [OPUS-100](https://huggingface.co/datasets/Helsinki-NLP/opus-100). Selection preserves the original train, validation and test split membership. All 10,000 text values are globally unique after Unicode normalization, case folding, punctuation removal and whitespace normalization. Source-target identity, short fragments, markup and placeholder patterns are filtered. No model-generated translations are included. |
| |
| The complete build script, exact source revision, split counts, rejection counts and Parquet checksums are published alongside the data. |
| |
| ## Load |
| |
| ```python |
| from datasets import load_dataset |
| |
| dataset = load_dataset("SEE7/AstheralDataset") |
| example = dataset["train"][0] |
| print(example["source_text"]) |
| print(example["target_text"]) |
| ``` |
| |
| ## Provenance and attribution |
| |
| Source: Helsinki-NLP/opus-100, revision '''+revision+'''. |
| OPUS-100: Zhang et al. (2020), “Improving Massively Multilingual Neural Machine Translation and Zero-Shot Translation.” Original parallel text comes from the OPUS collection. The original source corpus terms remain applicable. |
| |
| Curated dataset repository: SEE7/AstheralDataset. |
| ''' |
| (out/'README.md').write_text(readme) |
| api.upload_folder(repo_id='SEE7/AstheralDataset',repo_type='dataset',folder_path=str(out),commit_message='Publish 5000 real globally deduplicated translation pairs with provenance and checksums') |
| |
| published=load_dataset('SEE7/AstheralDataset') |
| assert {k:len(v) for k,v in published.items()}=={'train':4500,'validation':250,'test':250} |
| values=[];row_ids=[] |
| for split in published: |
| for row in published[split]: |
| values.extend([normalize(row['source_text']),normalize(row['target_text'])]);row_ids.append(row['id']) |
| assert len(set(values))==10000 and len(set(row_ids))==5000 |
| print('PUBLISHED_VERIFIED',json.dumps(manifest),flush=True) |
|
|
