Upload folder using huggingface_hub

.gitattributes

@@ -1,59 +1,60 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.lz4 filter=lfs diff=lfs merge=lfs -text
-*.mds filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text
-# Audio files - uncompressed
-*.pcm filter=lfs diff=lfs merge=lfs -text
-*.sam filter=lfs diff=lfs merge=lfs -text
-*.raw filter=lfs diff=lfs merge=lfs -text
-# Audio files - compressed
-*.aac filter=lfs diff=lfs merge=lfs -text
-*.flac filter=lfs diff=lfs merge=lfs -text
-*.mp3 filter=lfs diff=lfs merge=lfs -text
-*.ogg filter=lfs diff=lfs merge=lfs -text
-*.wav filter=lfs diff=lfs merge=lfs -text
-# Image files - uncompressed
-*.bmp filter=lfs diff=lfs merge=lfs -text
-*.gif filter=lfs diff=lfs merge=lfs -text
-*.png filter=lfs diff=lfs merge=lfs -text
-*.tiff filter=lfs diff=lfs merge=lfs -text
-# Image files - compressed
-*.jpg filter=lfs diff=lfs merge=lfs -text
-*.jpeg filter=lfs diff=lfs merge=lfs -text
-*.webp filter=lfs diff=lfs merge=lfs -text
-# Video files - compressed
-*.mp4 filter=lfs diff=lfs merge=lfs -text
-*.webm filter=lfs diff=lfs merge=lfs -text
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mds filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
+# Video files - compressed
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.webm filter=lfs diff=lfs merge=lfs -text
+probes.jsonl filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,9 @@
+*.jsonl
+__pycache__/
+*.pyc
+.venv/
+.env
+.DS_Store
+.pypirc
+dist/
+!probes.jsonl

README.md

@@ -1,3 +1,64 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+---
+
+# Finesse Benchmark Database
+
+## Overview
+
+`finesse-benchmark-database` is a data generation factory for atomic probes in the Finesse benchmark. It generates `probes_atomic.jsonl` files from Wikimedia Wikipedia datasets, leveraging Hugging Face's `datasets` library, tokenizers from `transformers`, and optional PyTorch support.
+
+This tool is designed to create high-quality, language-specific probe datasets for benchmarking fine-grained understanding in NLP tasks.
+
+## Installation
+
+Install the package from PyPI:
+
+```bash
+pip install finesse-benchmark-database
+```
+
+Ensure you have Python 3.10+ installed.
+
+## Usage
+
+Here's a complete example of how to configure and generate a dataset:
+
+```python
+from finesse_benchmark_database.config import ProbeConfig
+from finesse_benchmark_database.main import generate_dataset
+
+# Define the configuration
+my_config = ProbeConfig(
+    languages=['en', 'ko'],  # Languages to generate probes for
+    samples_per_language=10,  # Number of samples per language (reduce for testing)
+    output_file='my_first_probes.jsonl',  # Output file path
+    seed=123  # Random seed for reproducibility
+)
+
+# Generate the dataset
+print(f"Generating '{my_config.output_file}'...")
+generate_dataset(my_config)
+print("Dataset generation completed!")
+```
+
+### Configuration Options
+
+- `languages`: List of language codes (e.g., ['en', 'ko', 'fr']).
+- `samples_per_language`: Number of probe samples to generate per language.
+- `output_file`: Path to the output JSONL file.
+- `seed`: Optional seed for deterministic results.
+
+## Requirements
+
+- `datasets`
+- `transformers`
+- `torch` (for tokenization)
+
+## License
+
+This project is licensed under the Apache-2.0 license.
+
+## Contributing
+
+Contributions are welcome! Please open issues or pull requests on the [GitHub repository](https://github.com/your-repo/finesse-benchmark-database).

licence

@@ -0,0 +1,202 @@
+
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

main.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python3
+"""
+Example script to demonstrate the usage of the finesse-benchmark-database package.
+This script creates a simple configuration and generates a probes_atomic.jsonl file.
+"""
+
+from finesse_benchmark_database.config import ProbeConfig
+from finesse_benchmark_database.main import generate_dataset
+
+LANGUAGES = [
+    'en',
+    'ko',
+    'es',
+    'ja',
+    'ru',
+    'zh',
+    'ar',
+    'id',
+    'de',
+    'vi',
+]
+
+if __name__ == "__main__":
+    # Define a simple configuration for demonstration
+    demo_config = ProbeConfig(
+        languages=LANGUAGES,  # Start with English for simplicity
+        samples_per_language=5000,  
+        output_file='probes.jsonl',  # Output file for the demo
+        seed=42  # Fixed seed for reproducibility
+    )
+
+    print(f"Generating demo dataset: '{demo_config.output_file}'")
+    print("This will create a small probes_atomic.jsonl file using Wikipedia data.")
+    
+    # Generate the dataset
+    generate_dataset(demo_config)
+    
+    print("Demo generation completed! Check 'demo_probes.jsonl' for the output.")
+    print("You can now inspect the file or integrate this into your Finesse benchmark workflow.")

probes.jsonl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ebd230cd02f2c010c3114310e1d93e71a189f22b4a6346a0067dd116834315b6
+size 849544261

pyproject.toml

@@ -0,0 +1,19 @@
+[tool.poetry]
+name = "finesse-benchmark-database"
+version = "0.1.12"
+description = "Data generation factory for atomic probes in Finesse benchmark. Generates probes_atomic.jsonl from Wikimedia Wikipedia."
+authors = ["winter.sci.dev <enzoescipy@gmail.com>"]
+readme = "README.md"
+packages = [{include = "finesse_benchmark_database", from = "src"}]
+
+[tool.poetry.dependencies]
+python = "^3.10"
+datasets = "^4.3.0"
+transformers = "^4.35.0"
+torch = "^2.1.0"  # For tokenizer if needed
+
+
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"

src/finesse_benchmark_database/__init__.py

@@ -0,0 +1,48 @@
+"""finesse-benchmark-database: Multilingual Atomic Probe Generator for Long-Context Evaluation
+
+This package provides a flexible, configurable library for generating high-quality, traceable datasets of 'strings of beads'—
+atomic 64-token text chunks sourced from multilingual Wikipedia articles. It serves as the foundational data generation
+pipeline for the Finesse long-context benchmarking framework, ensuring reproducibility, semantic diversity, and
+complete metadata tracking for advanced LLM evaluation.
+
+Core Principles:
+- Atomic Beads: Exact 64-token chunks (discard incompletes) to test pure memory granularity.
+- Traceable Origins: Each bead/string includes full metadata (dataset, article_id, lang) for debugging and verification.
+- Multilingual Balance: Supports 10+ languages with configurable quotas for fair coverage.
+- Library-First Design: Instantiable via ProbeConfig for custom experiments; no globals or hardcoding.
+- JSONL Output: Efficient streaming format for large-scale datasets.
+
+Key Components:
+- ProbeConfig: Dataclass for all settings (languages, samples, chunk size, etc.).
+- generate_all_strings_of_beads(config): Produces list of {'source': metadata, 'beads': [text_chunks]} dicts.
+- write_strings_to_probes_atomic(config, strings): Serializes to JSONL with auto-assigned string_ids.
+
+Example Usage:
+    from finesse_benchmark_database import ProbeConfig, generate_all_strings_of_beads, write_strings_to_probes_atomic
+    
+    config = ProbeConfig(
+        languages=['en', 'ko'],
+        samples_per_language=100,
+        chunk_token_size=64,
+        output_file='my_probes.jsonl',
+        seed=42
+    )
+    
+    beads_strings = generate_all_strings_of_beads(config)
+    write_strings_to_probes_atomic(config, beads_strings)
+    # Outputs my_probes.jsonl with ~200 traceable strings of beads.
+
+Installation:
+pip install finesse-benchmark-database
+# Or via Poetry: poetry add finesse-benchmark-database
+
+This package powers the creation of ~1M+ atomic probes for rigorous long-context memory testing.
+See main.py for a full pipeline example.
+"""
+
+__version__ = "0.1.0"
+
+from .config import ProbeConfig
+from .chunker import generate_all_strings_of_beads
+from .writer import write_strings_to_probes_atomic
+from .main import generate_dataset

src/finesse_benchmark_database/chunker.py

@@ -0,0 +1,154 @@
+"""Chunker module for Finesse Benchmark Database Generator
+
+This module implements the 'Gemstone Necklace Crafter' aka the atomic bead generator.
+It transforms raw Wikimedia Wikipedia documents into 'strings of beads': sequential chains
+of exactly 64-token atomic beads from each source article, preserving semantic flow within
+each string while ensuring atomicity and reproducibility across languages.
+
+Key Principles:
+- Uses official reference tokenizer: bert-base-multilingual-cased for universal fairness.
+- Exactly 64-token beads only; discard all incomplete final chunks (golden rule).
+- One 'string' per Wikipedia article, with beads in original sequential order.
+- Balanced collection: samples_per_language strings per language.
+- Streaming processing for efficiency on massive datasets.
+- Logged progress for transparency and debugging.
+
+Usage:
+from chunker import generate_all_strings_of_beads
+strings_of_beads = generate_all_strings_of_beads()
+# Result: List[Dict] where each inner dict contains 'source' and 'beads' keys.
+"""
+
+import logging
+import random
+from typing import List, Dict
+
+from .config import ProbeConfig
+from datasets import load_dataset
+from transformers import AutoTokenizer
+
+# Configure logging for progress tracking
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+def generate_all_strings_of_beads(config: ProbeConfig) -> List[Dict]:
+    """
+    Generate all 'strings of beads' across languages using the official tokenizer.
+    
+    Args:
+        config: ProbeConfig instance with all parameters (languages, samples_per_language, etc.).
+    
+    For each language:
+    - Stream Wikipedia articles.
+    - For each article (up to config.samples_per_language):
+      - Tokenize the full text.
+      - Sequentially slice into 64-token chunks.
+      - Decode only exact 64-token chunks to bead texts.
+      - Form a 'string' as [bead1, bead2, ...] if at least one bead exists.
+    - Collect all strings into a global 2D list.
+    
+    Returns:
+        List of dictionaries: Outer list by language/order, inner dicts contain 'source' and 'beads' keys.
+    """
+    # Set global seed for reproducibility
+    random.seed(config.seed)
+    
+    # Load the official reference tokenizer (our 'public scale')
+    logger.info(f"Loading official tokenizer: {config.tokenizer_name}")
+    tokenizer = AutoTokenizer.from_pretrained(config.tokenizer_name)
+    
+    all_strings_of_beads: List[Dict] = []
+    
+    if config.languages is None:
+        raise ValueError("config.languages must be set to a list of language codes.")
+    
+    for lang in config.languages:
+        logger.info(f"Starting processing for language: {lang} (target: {config.samples_per_language} strings)")
+        
+        # Load streaming dataset with the specified split
+        dataset = load_dataset(
+            "wikimedia/wikipedia",
+            f"20231101.{lang}",
+            streaming=True,
+            split="train"
+        )
+        
+        strings_for_lang: List[Dict] = []
+        article_count = 0
+        
+        for example in dataset:
+            if len(strings_for_lang) >= config.samples_per_language:
+                break
+            
+            # Extract and clean text
+            text = example.get("text", "").strip()
+            if not text:
+                continue
+            
+            # Extract article ID for metadata
+            article_id = example.get("id", "")
+            
+            # Tokenize the full article
+            tokens = tokenizer.encode(text, add_special_tokens=False)
+            
+            # Sequentially chunk into exact config.chunk_token_size-token beads
+            beads: List[str] = []
+            for i in range(0, len(tokens), config.chunk_token_size):
+                chunk_tokens = tokens[i:i + config.chunk_token_size]
+                
+                # Golden rule: Only accept exactly 64 tokens; discard incompletes
+                if len(chunk_tokens) == config.chunk_token_size:
+                    bead_text = tokenizer.decode(chunk_tokens, skip_special_tokens=True).strip()
+                    if bead_text:  # Ensure non-empty after decoding
+                        beads.append(bead_text)
+            
+            # Only add the string if it has at least one bead
+            if beads:
+                string_data = {
+                    "source": {
+                        "dataset": "wikimedia/wikipedia",
+                        "article_id": article_id,
+                        "lang": lang
+                    },
+                    "beads": beads
+                }
+                strings_for_lang.append(string_data)
+                article_count += 1
+                
+                # Log progress every 100 articles
+                if article_count % 100 == 0:
+                    logger.info(
+                        f"Language {lang}: Processed {article_count} articles, "
+                        f"collected {len(strings_for_lang)} strings so far"
+                    )
+        
+        # Add language's strings to the global collection
+        all_strings_of_beads.extend(strings_for_lang)
+        logger.info(f"Completed {lang}: {len(strings_for_lang)} strings generated "
+                    f"(from {article_count} articles)")
+    
+    total_strings = len(all_strings_of_beads)
+    logger.info(f"Generation complete: {total_strings} total strings of beads across {len(config.languages)} languages")
+    
+    return all_strings_of_beads
+
+if __name__ == "__main__":
+    # Example library usage: Create config and generate
+    from config import ProbeConfig
+    
+    # User creates and populates config
+    test_config = ProbeConfig(
+        languages=['en', 'ko'],  # Test with 2 languages
+        samples_per_language=5,  # Small sample for demo
+        chunk_token_size=64,
+        tokenizer_name="google-bert/bert-base-multilingual-cased",
+        output_file="probes_atomic.jsonl",
+        seed=42
+    )
+    
+    # Generate and print summary
+    strings = generate_all_strings_of_beads(test_config)
+    print(f"Generated {len(strings)} strings of beads.")
+    if strings:
+        avg_beads_per_string = sum(len(s['beads']) for s in strings) / len(strings)
+        print(f"Average beads per string: {avg_beads_per_string:.2f}")

src/finesse_benchmark_database/config.py

@@ -0,0 +1,79 @@
+"""Configuration module for Finesse Benchmark Database Generator
+
+This module holds all configurable parameters for generating atomic probes from Wikimedia Wikipedia datasets.
+Based on the 'Beads and String' model: 64-token atomic beads from diverse languages, ensuring semantic continuity within strings but independence across probes.
+
+Key Principles:
+- Fixed 64-token chunk size for atomic beads.
+- Balanced sampling across languages for global diversity.
+- Seeded randomness for perfect reproducibility.
+- Output in probes_atomic.jsonl format for dynamic assembly in evaluation.
+
+Usage:
+from config import tokenizer_name, languages, chunk_token_size, samples_per_language, output_file, seed
+"""
+
+from dataclasses import dataclass
+import random
+
+@dataclass
+class ProbeConfig:
+    """Central configuration for probe generation.
+    
+    This dataclass serves as a flexible template for library users.
+    Instantiate and populate it with desired values before passing to generate functions.
+    
+    Example:
+        config = ProbeConfig(
+            languages=['en', 'ko'],
+            samples_per_language=10,
+            chunk_token_size=64
+        )
+    """
+    
+    # Tokenizer for tokenization (default: multilingual BERT)
+    tokenizer_name: str = "google-bert/bert-base-multilingual-cased"
+    
+    # Languages for balanced multilingual coverage (must be set by user)
+    languages: list[str] = None
+    
+    # Atomic bead size (golden rule from design; override for custom experiments)
+    chunk_token_size: int = 64
+
+    # Number of 'strings of beads' (source documents) per language: The number of complete
+    # Wikipedia articles to process per language. Each document is chunked sequentially into
+    # multiple 64-token atomic beads, preserving original order and semantic flow within
+    # the string. This is NOT the total count of individual beads (which will be much higher,
+    # depending on document lengths), but the number of such connected 'necklaces' or
+    # 'strings' for balanced multilingual coverage.
+    samples_per_language: int = 10000
+    
+    # Output file for atomic probes
+    output_file: str = "probes_atomic.jsonl"
+    
+    # Fixed seed for reproducibility (immutable law; set to None for non-deterministic runs)
+    seed: int = 42
+
+def get_config() -> ProbeConfig:
+    """Instantiate and return the configuration with languages initialized."""
+    config = ProbeConfig()
+    config.languages = [
+        'en',  # English
+        'ko',  # Korean
+        'es',  # Spanish
+        'ja',  # Japanese
+        'ru',  # Russian
+        'zh',  # Chinese
+        'ar',  # Arabic
+        'id',  # Indonesian
+        'de',  # German
+        'vi',  # Vietnamese
+    ]
+    
+    # Set global seed for all randomness
+    random.seed(config.seed)
+    
+    return config
+
+# Default config instance
+CONFIG = get_config()

src/finesse_benchmark_database/main.py

@@ -0,0 +1,99 @@
+"""Main orchestration module for Finesse Benchmark Database Generator
+
+This script serves as the 'Conductor' that orchestrates the entire pipeline: from configuration setup
+through atomic bead generation to final JSONL serialization. It ties together config, chunker, and writer
+for end-to-end execution of our 'Gemstone Necklace' production process.
+
+Key Principles:
+- Ensures all modules use the shared CONFIG for consistency.
+- Executes chunking first (memory-intensive), then writing (I/O-focused).
+- Comprehensive logging for audit trail and debugging.
+- Designed for one-shot full generation; scale via external orchestration if needed.
+
+Usage:
+python main.py
+# Runs the full pipeline: generates ~100,000 strings of beads and writes to probes_atomic.jsonl
+"""
+
+import logging
+
+from .config import ProbeConfig
+from .chunker import generate_all_strings_of_beads
+from .writer import write_strings_to_probes_atomic
+
+# Configure logging for the entire pipeline
+logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
+logger = logging.getLogger(__name__)
+
+def generate_dataset(config: ProbeConfig):
+    logger.info(f"Config set: {len(config.languages)} languages, {config.chunk_token_size}-token beads, "
+                f"{config.samples_per_language} strings per language")
+    
+    # Step 2: Generate all structured 'strings of beads' using the chunker
+    logger.info("Starting bead generation with chunker...")
+    all_strings_of_beads = generate_all_strings_of_beads(config)
+    total_strings = len(all_strings_of_beads)
+    logger.info(f"Bead generation complete: {total_strings} total strings produced "
+                f"(~{total_strings / len(config.languages):.0f} per language)")
+    
+    # Step 3: Serialize to the atomic probes JSONL file using the writer
+    logger.info(f"Starting serialization to {config.output_file}...")
+    write_strings_to_probes_atomic(config, all_strings_of_beads)
+    logger.info(f"Serialization complete: All {total_strings} strings saved with metadata.")
+    
+    # Final summary
+    estimated_beads = sum(len(s['beads']) for s in all_strings_of_beads)
+    logger.info("=== Pipeline Summary (Test Run) ===")
+    logger.info(f"- Total strings: {total_strings}")
+    logger.info(f"- Total atomic beads: {estimated_beads}")
+    logger.info(f"- Coverage: {len(set(s['source']['lang'] for s in all_strings_of_beads))} languages")
+    logger.info(f"- Output file: {config.output_file}")
+    logger.info("For full production: Increase samples_per_language and add more languages in config.")
+    logger.info("=== Finesse Benchmark Database Generation Example Finished Successfully ===")
+
+
+def main() -> None:
+    """Execute the full Finesse benchmark database generation pipeline as a library example.
+    
+    This demonstrates how users would create a ProbeConfig, generate data, and write output.
+    For production, adjust config parameters as needed (e.g., more languages, larger samples).
+    """
+    logger.info("=== Finesse Benchmark Database Pipeline Started (Library Example) ===")
+    
+    # Step 1: User creates and configures ProbeConfig for this run
+    logger.info("Creating test configuration...")
+    test_config = ProbeConfig(
+        languages=['en'],  # Single language for quick demo
+        samples_per_language=5,  # Small sample for testing
+        chunk_token_size=64,
+        tokenizer_name="google-bert/bert-base-multilingual-cased",
+        output_file="test_probes_atomic.jsonl",
+        seed=42
+    )
+    logger.info(f"Config set: {len(test_config.languages)} languages, {test_config.chunk_token_size}-token beads, "
+                f"{test_config.samples_per_language} strings per language")
+    
+    # Step 2: Generate all structured 'strings of beads' using the chunker
+    logger.info("Starting bead generation with chunker...")
+    all_strings_of_beads = generate_all_strings_of_beads(test_config)
+    total_strings = len(all_strings_of_beads)
+    logger.info(f"Bead generation complete: {total_strings} total strings produced "
+                f"(~{total_strings / len(test_config.languages):.0f} per language)")
+    
+    # Step 3: Serialize to the atomic probes JSONL file using the writer
+    logger.info(f"Starting serialization to {test_config.output_file}...")
+    write_strings_to_probes_atomic(test_config, all_strings_of_beads)
+    logger.info(f"Serialization complete: All {total_strings} strings saved with metadata.")
+    
+    # Final summary
+    estimated_beads = sum(len(s['beads']) for s in all_strings_of_beads)
+    logger.info("=== Pipeline Summary (Test Run) ===")
+    logger.info(f"- Total strings: {total_strings}")
+    logger.info(f"- Total atomic beads: {estimated_beads}")
+    logger.info(f"- Coverage: {len(set(s['source']['lang'] for s in all_strings_of_beads))} languages")
+    logger.info(f"- Output file: {test_config.output_file}")
+    logger.info("For full production: Increase samples_per_language and add more languages in config.")
+    logger.info("=== Finesse Benchmark Database Generation Example Finished Successfully ===")
+
+if __name__ == "__main__":
+    main()

src/finesse_benchmark_database/writer.py

@@ -0,0 +1,83 @@
+"""Writer module for Finesse Benchmark Database Generator
+
+This module implements the 'Eternal Scribe': it takes the structured 'strings of beads' output from chunker.py
+and serializes them into probes_atomic.jsonl format. Each line is an independent JSON object with metadata
+for full traceability, ensuring the dataset is reproducible, queryable, and efficient for large-scale use.
+
+Key Principles:
+- JSONL format for streaming efficiency (one complete JSON object per line).
+- Assigns immutable 'string_id' for unique identification.
+- Preserves all source metadata (dataset, article_id, lang) for debugging and verification.
+- UTF-8 encoding to handle multilingual content without corruption.
+- Logged serialization progress for transparency.
+
+Usage:
+from writer import write_strings_to_probes_atomic
+from chunker import generate_all_strings_of_beads
+strings = generate_all_strings_of_beads()
+write_strings_to_probes_atomic(strings)
+# Results in probes_atomic.jsonl with all probes serialized.
+"""
+
+import json
+import logging
+from typing import List, Dict
+
+from .config import ProbeConfig
+
+# Configure logging for progress tracking
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+def write_strings_to_probes_atomic(config: ProbeConfig, all_strings_of_beads: List[Dict]) -> None:
+    """
+    Write the structured strings of beads to the atomic probes JSONL file.
+    
+    Args:
+        config: ProbeConfig instance specifying output_file and SEED.
+        all_strings_of_beads: List of dicts from chunker.py, each with 'source' and 'beads'.
+    
+    Side Effects:
+        Creates/appends to config.output_file.
+        Each string gets a sequential 'string_id' assigned.
+    """
+    output_path = config.output_file
+    logger.info(f"Starting serialization to {output_path}. Total strings: {len(all_strings_of_beads)}")
+    
+    with open(output_path, 'w', encoding='utf-8') as file:
+        for i, string_data in enumerate(all_strings_of_beads):
+            # Assign unique string_id (immutable global identifier)
+            string_data['string_id'] = i
+            
+            # Serialize to JSON line
+            json_line = json.dumps(string_data, ensure_ascii=False, separators=(',', ':'))
+            file.write(json_line + '\n')
+            
+            # Log progress every 1000 strings
+            if (i + 1) % 1000 == 0:
+                logger.info(f"Serialized {i + 1} strings to {output_path}")
+    
+    total_written = len(all_strings_of_beads)
+    logger.info(f"Serialization complete: {total_written} strings written to {output_path}")
+    logger.info(f"File structure: Each line is a JSON object with 'string_id', 'source' (metadata), and 'beads' ({config.chunk_token_size}-token chunks).")
+
+if __name__ == "__main__":
+    # Example library usage: Create config and run full pipeline
+    from config import ProbeConfig
+    from chunker import generate_all_strings_of_beads
+    
+    # User creates and populates config for a small test run
+    demo_config = ProbeConfig(
+        languages=['en'],  # Single language for quick demo
+        samples_per_language=3,  # Very small sample
+        chunk_token_size=64,
+        tokenizer_name="google-bert/bert-base-multilingual-cased",
+        output_file="demo_probes_atomic.jsonl",
+        seed=42
+    )
+    
+    print("Generating strings of beads...")
+    strings = generate_all_strings_of_beads(demo_config)
+    print(f"Generated {len(strings)} strings. Now serializing...")
+    write_strings_to_probes_atomic(demo_config, strings)
+    print("Pipeline complete. Check demo_probes_atomic.jsonl for output.")