upload data

README.md

@@ -1,4 +1,76 @@
 ---
 license: apache-2.0
-
+task_categories:
+- text-generation
+language:
+- id
+- en
+tags:
+- math
+- reasoning
+- cot
+- tool-use
+- code-interpreter
+- engineering
+- quantitative-finance
+- olympic-math
+size_categories:
+- 100K<n<1M
+configs:
+  - config_name: default
+    data_files:
+      - split: train
+        path: data/train-*
+      - split: validation
+        path: data/validation-*
 ---
+
+# Lumina-Math-Repository (Lumina-Math-Foundation-V2)
+
+Lumina-Math-Foundation-V2 is a massive, high-density mathematical and engineering reasoning dataset comprising **500,000 training instances** and **5,100 validation instances** uniformly spanning **255 specialized domain batches**.
+
+Designed specifically for **Modular Tool-Integrated Reasoning (Code-Interpreter)**, each instance cleanly isolates conceptual mathematical reasoning from executable Python verification code, execution output, and final formal solutions.
+
+## Key Highlights
+- **100% Non-Trivial Ground Truth:** Every mathematical derivation and industrial design equation adheres to real engineering and academic standards (ASME, AISC, SNI, Uniswap v3 / Curve invariants, Calabi-Yau 3-folds, and GPU Roofline performance models). No toy formulas or dummy math.
+- **Modular Tool-Integrated Reasoning (Code-Interpreter):**
+  - `thought`: Pure conceptual mathematical thinking and planning before code execution. Free of artificial phase markers or sycophancy.
+  - `code`: Deterministic, executable Python verification script (clean code without decorative ASCII boxes or curriculum fluff).
+  - `code_output`: Deterministic runtime / assertion confirmation.
+  - `thought_2`: Post-execution synthesis evaluating computational verification against analytical derivation.
+  - `solution`: Clean formal mathematical solution with unnumbered, standard headings (`Langkah Penyelesaian` & `Jawaban Akhir`).
+- **Multi-Mode Reasoning Velocity:**
+  - **Short (~25%):** High-velocity derivation directly yielding the target answer.
+  - **Medium (~50%):** Standard balanced Chain-of-Thought (CoT) diagnostic and step-by-step resolution.
+  - **Long (~25%):** Comprehensive axiomatic derivation exploring boundary cases and formal verification.
+- **Balanced 255-Batch Coverage:**
+  - Training set: ~1,960 - 1,961 records per batch across all 255 batches.
+  - Validation set: Exactly 20 records per batch across all 255 batches ($255 \times 20 = 5,100$). Disjoint seed space guarantees zero data contamination.
+- **Direct Parquet Format:** Compact, columnar, and immediately queryable through the Hugging Face Dataset Viewer and `datasets.load_dataset`.
+
+## Dataset Structure
+
+### Features
+| Column | Type | Description |
+|---|---|---|
+| `id` | string | Unique deterministic record identifier |
+| `domain` | string | Specialized domain classification |
+| `topic` | string | Specific mathematical/engineering topic |
+| `mode` | string | Reasoning mode (`short`, `medium`, `long`) |
+| `problem` | string | Clean problem statement / prompt |
+| `thought` | string | Conceptual mathematical reasoning before code execution |
+| `code` | string | Deterministic executable Python verification script |
+| `code_output` | string | Execution output / assertion verification |
+| `thought_2` | string | Post-execution synthesis reasoning |
+| `solution` | string | Clean formal solution dossier (`Langkah Penyelesaian` & `Jawaban Akhir`) |
+| `final_answer` | string | Extracted scalar/expression answer |
+| `boxed_answer` | string | LaTeX boxed representation `\boxed{...}` |
+
+## Usage
+```python
+from datasets import load_dataset
+
+ds = load_dataset("cloverx-id/Lumina-Math-Repository")
+print(ds["train"][0])
+print(ds["validation"][0])
+```

data/{train-00000-of-00010.parquet → train-00000-of-00016.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:33b0c72c52498c4bd6f77fbdea5846dd3064ea9151a0a6dbe856b4513981707c
-size 316190721
+oid sha256:74bbcc2228a14468d4346c798044cd471d9120f199149f4c241812221ea9a8b6
+size 123377165

data/{train-00001-of-00010.parquet → train-00001-of-00016.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb6679eb9769af29c3b54da124afe1942296cdde1379104bbb536eec8355fc18
-size 315341610
+oid sha256:f5b13a7907bcef4380958af547ce5b9068d7406998391e62c2314b36f2feea56
+size 122860750

data/{train-00002-of-00010.parquet → train-00002-of-00016.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:944245ab86b7429b9d028043f024e9e16202a4a67761be942832bfb1dc682303
-size 314148050
+oid sha256:53add329b2779d3ed8b48465e4ceba9c8272da74009b01199902df724402111a
+size 122855465

data/{train-00003-of-00010.parquet → train-00003-of-00016.parquet}

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a12a86726f3c50820ba125cda102bb9c2434f54e13d261a07b001663d42f3a0
-size 314840230
+oid sha256:92d54b525101ae371ad27df6fb34bdebb025645f2c55f8ef6ae8a335357a520d
+size 122684482

data/train-00004-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:1c596c3442c1382877273bb776124433bb12a961f8781739791e3ed670a3b847
-size 315170195

data/train-00004-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03eb3bab2edc2c5b0f26ef9638040e5100e9d1d8b8494712fef00a558e8239ba
+size 122690978

data/train-00005-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:71145a1126e0394d3433e9f84cee45c80f1e8ba7a11e3c08ab410e065c8820e8
-size 314926420

data/train-00005-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f0637e598b65be2d0bd295c51de5b893997c79cdec0a3f3e5e2ede3508105525
+size 123424334

data/train-00006-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:cbf326c5ff79a1a64bc170d0947195ccfe8b760b07485f0b67b983b6c961c331
-size 314729568

data/train-00006-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ca97efbfce513dcc6037c7e62aef6dbbe99a3b03f4265751c40f19048c0e0c99
+size 122153589

data/train-00007-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fd52e4ac0b072fcb6671fdb04db220a450317a4833575dce5e2a9465fe94d465
-size 315729828

data/train-00007-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b2e02b5f4f353a640fc51b7f1c60ed9a66afad43fd8933d353bc739de1b99f1
+size 123105792

data/train-00008-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:fe3caaa666dc959f2c2c2055f16075a613ce36b66431ac6db6e1f3add99903bb
-size 314539826

data/train-00008-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0f0d8a5b8555b738cdb37761609b4e5f1e1deafc8910a95d2dfd915939262655
+size 122698802

data/train-00009-of-00010.parquet

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ef97663736e905e075381a621d491d968ac70a6c9b8bffa1a58adc85a91f7f9c
-size 315409755

data/train-00009-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f984dc3cb801f50d079588e9128481c004060485d1c637c02d2c6b530950007b
+size 122890099

data/train-00010-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfbcd086dd6f3cf5cd703fddf9348bb99e0f47a3b6ab513922af1976865061d0
+size 122828436

data/train-00011-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f9422623a387564a825e70ab0fa91a0eba1290aaa416c50a1c889b52ade30eae
+size 123780033

data/train-00012-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4ee22b5097a2f7483c4cb38f90aa18ebd1c8d0e10625bb92fa8fb4a99be2c6ed
+size 122633967

data/train-00013-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d6d32f0ee2e6d10030b0f7140fbd59be3d4eec8781a16cc51d762dc9d889724a
+size 123123855

data/train-00014-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16af8d7e2035d7d488f7b4aeb3fc39cebf11691cbcf6127fb82d7275e55d55c3
+size 122368205

data/train-00015-of-00016.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87d5276eaf8882d866cd705f4a2f518b9731cdd6386bc4d6d511b871bce96995
+size 122962184

data/validation-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3417adcfdcf52b8b77f511a817395b7e50545c37f674f4a2e1078dd0f888475c
-size 32310321
+oid sha256:c88e13c7b7e8c2764dbddbfcb1fc74ee58b3a93d7dd2c5c926ac2fec90d9f962
+size 19678730