Add files using upload-large-folder tool

LICENSES.md

@@ -0,0 +1,37 @@
+# License and provenance notice
+
+This compilation uses `license: other`. No single blanket license is asserted
+over all upstream questions, images, or benchmark annotations. The release
+schema, cleaning script, and newly authored metadata follow the repository's
+project license, but upstream content remains governed by its original terms.
+
+## General text
+
+Sources include QANTA, MedQA, GPQA, FinQA, PubMedQA, SuperGLUE, FinanceBench,
+PIQA, HellaSwag, SocialIQA, Minerva/Hendrycks mathematics, MMLU-Pro, MathQA,
+TACO, APPS, and the recorded long-context sources. The final dataset card must
+link each redistributed source to its exact version and license. The historical
+`wustar/*` mirrors recorded in the build config are not anonymously resolvable;
+they are provenance identifiers, not evidence of a redistribution license.
+
+## Multimodal
+
+MMR-Bench V2 aggregates multiple upstream benchmarks and does not assert a
+single license over their questions or images. Preserve the source project's
+`LICENSES.md` and original benchmark terms. The formal test sources include
+AI2D, ScienceQA, ScienceQA-IMG, DocVQA, OCR-VQA, and VQAv2/COCO-derived data.
+MMR-4000 additionally includes MMStar, RealWorldQA, SEEDBench2 Plus, SimpleVQA,
+OCRBench, InfoVQA, ChartQAPro, CharXiv, MathVista, MathVision, MathVerse,
+LogicVista, WeMath, MMMU-Pro, BLINK, ERQA, VStarBench, and HallusionBench.
+The inherited MMR notice marks binary redistribution as unverified.
+
+## Chinese
+
+Sources include AlignBench, C-Eval, CIF-Bench, CLUE-C3, CLUE-CHID,
+CLUE-CMRC2018, CLUE-CSL, CLUE-OCNLI, CMB-Clin, ChineseSimpleQA, FinEval,
+FollowBench-zh, GAOKAO-Bench, LawBench, and Chinese LongBench tasks. Preserve
+the original per-dataset licenses and attribution notices.
+
+Redistribution must be reviewed against each upstream license. When an upstream
+license does not permit redistribution, publish only IDs, derived routing
+labels, embeddings where permitted, and a source resolver.

README.md

@@ -0,0 +1,65 @@
+---
+license: other
+pretty_name: RobustRouteBench
+configs:
+- config_name: text
+  data_files:
+  - split: train
+    path: data/text/train-*.parquet
+  - split: validation
+    path: data/text/validation-*.parquet
+  - split: test
+    path: data/text/test-*.parquet
+- config_name: multimodal
+  data_files:
+  - split: train
+    path: data/multimodal/train-*.parquet
+  - split: validation
+    path: data/multimodal/validation-*.parquet
+  - split: test
+    path: data/multimodal/test-*.parquet
+- config_name: chinese
+  data_files:
+  - split: train
+    path: data/chinese/train-*.parquet
+  - split: validation
+    path: data/chinese/validation-*.parquet
+  - split: test
+    path: data/chinese/test-*.parquet
+---
+
+# RobustRouteBench
+
+RobustRouteBench is a unified public benchmark for general-text, multimodal,
+and Chinese-language model routing.
+
+```python
+from datasets import load_dataset
+
+text = load_dataset("SinapisAI/RobustRouteBench", "text")
+multimodal = load_dataset("SinapisAI/RobustRouteBench", "multimodal")
+chinese = load_dataset("SinapisAI/RobustRouteBench", "chinese")
+```
+
+Each scenario exposes separate `train`, `validation`, and `test` splits. The
+500-row validation split is a convenience overlay containing 400 standard and
+100 robust queries. Those rows intentionally remain in the complete test split,
+so validation and test overlap by design.
+
+The test split contains both standard and robustness evaluation rows. Use
+`evaluation_type` (`standard` or `robust`) and `variant_type` to report slices.
+
+Each row is query-centric. `performance`, `cost_cny`, and `latency_ms` are
+aligned to `model_ids`. All costs use CNY; historical USD proxies use the frozen
+benchmark conversion `1 USD = 7.2 CNY`.
+
+Source outcomes that were not successful have performance zero. No source
+diagnostic payloads or raw model responses are published. Official BGE-M3
+embeddings are included as baseline features but alternative encoders are
+allowed.
+
+The repository is self-contained. All MMR-4000 training images and all standard
+and robustness evaluation media are stored under `media/multimodal/`; no other
+dataset repository is required to train or evaluate a router. See `LICENSES.md`
+for upstream attribution and terms. This compilation uses `license: other`; no
+blanket license is asserted over all upstream questions or images.

data/chinese/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:427a198c55d8a323d4f525506cc380d0c3bf45ef6f483c56411be8119e521984
+size 103944022

data/chinese/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:01a4fb52c9de99989e99e4bcf68eeeadced8c36b400f8e41bf9589f9955c80a1
+size 56291524

data/chinese/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:273dcc0ac995cfc5077c0347b75e027dce518611cc78afaa9c34009dde6fcfc1
+size 4617464

data/multimodal/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c1a237329ae71eaec37e80682d5e004d770a8751f1df3610182ee331b0ea652c
+size 38316997

data/multimodal/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e05164bfbe35337bc768fb29928c065c085e9aa2b7c433aa083c8df67e8ee5b
+size 21911118

data/multimodal/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a1c4d7b08a64ae34f91649f471fccbd4076be8635bee11bae728e237ff16aa0
+size 2815950

data/text/test-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cd351857a30a9f94f261e4df8ba6b433a8b75d484dd88b411c95b4a576f64ab2
+size 287326444

data/text/train-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c29e24b06a0b821e173e13af395532c9d209545b36a391e47e0da22ec9794213
+size 23676804

data/text/validation-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:11d9104e32f4d396f170fda70c7fe1d9f0e1e1b66ca1fbe3baf4e9bcea5343f8
+size 8600178

manifest.json

@@ -0,0 +1,169 @@
+{
+  "schema_version": "robust-route-bench-v1.0.0",
+  "version": "1.0.0",
+  "generated_at": "2026-10-07T06:54:26.586691+00:00",
+  "repository": "SinapisAI/RobustRouteBench",
+  "fully_public": true,
+  "self_contained": true,
+  "external_data_dependencies": false,
+  "validation_overlaps_test": true,
+  "validation_policy": "400 standard + 100 robust per scenario",
+  "cost_policy": {
+    "currency": "CNY",
+    "usd_to_cny_rate": 7.2,
+    "rate_is_frozen_benchmark_constant": true
+  },
+  "contains_source_diagnostics": false,
+  "contains_raw_model_responses": false,
+  "media_archives": {
+    "train": {
+      "path": "media/multimodal/mmr-4000-train.zip",
+      "bytes": 533630327,
+      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
+    },
+    "standard": {
+      "path": "media/multimodal/benchmark-qa-5000.zip",
+      "bytes": 359143226,
+      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
+    },
+    "robust": {
+      "path": "media/multimodal/robustness-benchmark-2000.zip",
+      "bytes": 454277129,
+      "sha256": "a39935c9c704e5976074eae8ed8f60d395fb629e2f6776203d5bc9ae7f460935"
+    }
+  },
+  "scenarios": {
+    "text": {
+      "unique_queries": 10188,
+      "models": 63,
+      "published_rows": {
+        "train": 3000,
+        "validation": 500,
+        "test": 7188
+      },
+      "source_rows": {
+        "train": 3000,
+        "standard": 4188,
+        "robust": 3000
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      }
+    },
+    "multimodal": {
+      "unique_queries": 11000,
+      "models": 45,
+      "published_rows": {
+        "train": 4000,
+        "validation": 500,
+        "test": 7000
+      },
+      "source_rows": {
+        "train": 4000,
+        "standard": 5000,
+        "robust": 2000
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      }
+    },
+    "chinese": {
+      "unique_queries": 14090,
+      "models": 10,
+      "published_rows": {
+        "train": 7590,
+        "validation": 500,
+        "test": 6500
+      },
+      "source_rows": {
+        "train": 7590,
+        "standard": 4400,
+        "robust": 2100
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      }
+    }
+  },
+  "files": {
+    "LICENSES.md": {
+      "bytes": 1951,
+      "sha256": "25a37064f44d101f433d64cf449100abed88dfd11bfe9e20fcb87573f2ad193b"
+    },
+    "README.md": {
+      "bytes": 2325,
+      "sha256": "b71b8537d1a20c4a4198b53014b99016f485ec33d02fca657e1f7af7cd4b85b0"
+    },
+    "data/chinese/test-00000-of-00001.parquet": {
+      "bytes": 103944022,
+      "sha256": "427a198c55d8a323d4f525506cc380d0c3bf45ef6f483c56411be8119e521984"
+    },
+    "data/chinese/train-00000-of-00001.parquet": {
+      "bytes": 56291524,
+      "sha256": "01a4fb52c9de99989e99e4bcf68eeeadced8c36b400f8e41bf9589f9955c80a1"
+    },
+    "data/chinese/validation-00000-of-00001.parquet": {
+      "bytes": 4617464,
+      "sha256": "273dcc0ac995cfc5077c0347b75e027dce518611cc78afaa9c34009dde6fcfc1"
+    },
+    "data/multimodal/test-00000-of-00001.parquet": {
+      "bytes": 38316997,
+      "sha256": "c1a237329ae71eaec37e80682d5e004d770a8751f1df3610182ee331b0ea652c"
+    },
+    "data/multimodal/train-00000-of-00001.parquet": {
+      "bytes": 21911118,
+      "sha256": "8e05164bfbe35337bc768fb29928c065c085e9aa2b7c433aa083c8df67e8ee5b"
+    },
+    "data/multimodal/validation-00000-of-00001.parquet": {
+      "bytes": 2815950,
+      "sha256": "2a1c4d7b08a64ae34f91649f471fccbd4076be8635bee11bae728e237ff16aa0"
+    },
+    "data/text/test-00000-of-00001.parquet": {
+      "bytes": 287326444,
+      "sha256": "cd351857a30a9f94f261e4df8ba6b433a8b75d484dd88b411c95b4a576f64ab2"
+    },
+    "data/text/train-00000-of-00001.parquet": {
+      "bytes": 23676804,
+      "sha256": "c29e24b06a0b821e173e13af395532c9d209545b36a391e47e0da22ec9794213"
+    },
+    "data/text/validation-00000-of-00001.parquet": {
+      "bytes": 8600178,
+      "sha256": "11d9104e32f4d396f170fda70c7fe1d9f0e1e1b66ca1fbe3baf4e9bcea5343f8"
+    },
+    "media/multimodal/README.json": {
+      "bytes": 951,
+      "sha256": "2db0afa1865167c68f31596e762f7f5d523ccdc2e3918444a8d3a91e421abd0c"
+    },
+    "media/multimodal/benchmark-qa-5000.zip": {
+      "bytes": 359143226,
+      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
+    },
+    "media/multimodal/mmr-4000-train.zip": {
+      "bytes": 533630327,
+      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
+    },
+    "media/multimodal/robustness-benchmark-2000.zip": {
+      "bytes": 454277129,
+      "sha256": "a39935c9c704e5976074eae8ed8f60d395fb629e2f6776203d5bc9ae7f460935"
+    },
+    "metadata/chinese-models.json": {
+      "bytes": 409,
+      "sha256": "839923eb8b46c8cdebf8ac434d0a3c2632dd8743679aeafdf4faa006ec7474ae"
+    },
+    "metadata/multimodal-models.json": {
+      "bytes": 1313,
+      "sha256": "23c5f96870b0a38036a3ad47276dae1abbd7f15cc3e3f82e9c20824df2cd570c"
+    },
+    "metadata/text-models.json": {
+      "bytes": 2087,
+      "sha256": "a667c93e09ef001481690a64e691e8637c0fc25bc9537e1de19e3a71cec9723e"
+    },
+    "normalized-schema.json": {
+      "bytes": 1792,
+      "sha256": "4350e6160af5b05a264e5d6730ebed213792af70524f70c8535946929fd8b2f4"
+    }
+  }
+}

media/multimodal/README.json

@@ -0,0 +1,24 @@
+{
+  "self_contained": true,
+  "external_download_required": false,
+  "archives": {
+    "train": {
+      "path": "media/multimodal/mmr-4000-train.zip",
+      "bytes": 533630327,
+      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
+    },
+    "standard": {
+      "path": "media/multimodal/benchmark-qa-5000.zip",
+      "bytes": 359143226,
+      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
+    },
+    "robust": {
+      "path": "media/multimodal/robustness-benchmark-2000.zip",
+      "bytes": 454277129,
+      "sha256": "a39935c9c704e5976074eae8ed8f60d395fb629e2f6776203d5bc9ae7f460935"
+    }
+  },
+  "train_archive_contents": "MMR-4000 instances, selection metadata, provenance, validation metadata, and all 4,285 referenced image files",
+  "standard_archive_contents": "Vision QA 5,000 instances and media",
+  "robust_archive_contents": "Robustness 2,000 instances, media, and references"
+}

media/multimodal/benchmark-qa-5000.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef
+size 359143226

media/multimodal/mmr-4000-train.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66
+size 533630327

media/multimodal/robustness-benchmark-2000.zip

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a39935c9c704e5976074eae8ed8f60d395fb629e2f6776203d5bc9ae7f460935
+size 454277129

metadata/chinese-models.json

@@ -0,0 +1,17 @@
+{
+  "scenario": "chinese",
+  "model_count": 10,
+  "model_ids": [
+    "deepseek/deepseek-v4-flash",
+    "deepseek/deepseek-v4-pro",
+    "minimax/minimax-m2.7",
+    "minimax/minimax-m3",
+    "moonshot/kimi-k3",
+    "qwen/qwen3.7-max",
+    "qwen/qwen3.7-plus",
+    "qwen/qwen3.8-max",
+    "z-ai/glm-5.2",
+    "z-ai/glm-5.3"
+  ],
+  "array_alignment": "performance, cost_cny, and latency_ms use model_ids order"
+}

metadata/multimodal-models.json

@@ -0,0 +1,52 @@
+{
+  "scenario": "multimodal",
+  "model_count": 45,
+  "model_ids": [
+    "GLM-4.1V-9B-Thinking",
+    "Gemma3-4B",
+    "Gemma4-12B-it",
+    "InternVL3-14B-Instruct",
+    "InternVL3-2B-Instruct",
+    "InternVL3-78B",
+    "InternVL3-8B-Instruct",
+    "InternVL3_5-14B-Instruct",
+    "InternVL3_5-2B-Instruct",
+    "InternVL3_5-30B-A3B",
+    "InternVL3_5-4B-Instruct",
+    "InternVL3_5-8B-Instruct",
+    "InternVL3_5-GPT-OSS-20B-A4B-Preview",
+    "Keye-VL-2.0-30B-A3B",
+    "MiMo-VL-7B-RL",
+    "MiniCPM-V-4.6",
+    "Ovis2-16B",
+    "Ovis2-1B",
+    "Ovis2-2B",
+    "Ovis2-34B",
+    "Ovis2-4B",
+    "Ovis2-8B",
+    "Ovis2.5-2B",
+    "Ovis2.5-9B",
+    "Ovis2.6-30B-A3B",
+    "Qwen2.5-VL-32B-Instruct",
+    "Qwen2.5-VL-3B-Instruct",
+    "Qwen2.5-VL-72B-Instruct",
+    "Qwen2.5-VL-7B-Instruct",
+    "Qwen3-VL-2B-Instruct",
+    "Qwen3-VL-4B-Instruct",
+    "anthropic/claude-opus-4-7",
+    "google/gemini-3.5-flash",
+    "moonshot/kimi-k3",
+    "openai/gpt-5",
+    "openai/gpt-5-6-sol",
+    "openai/gpt-5.5",
+    "qwen/qwen3-vl-30b-a3b-instruct",
+    "qwen/qwen3-vl-32b-instruct",
+    "qwen/qwen3-vl-8b-instruct",
+    "qwen/qwen3.5-27b",
+    "qwen/qwen3.5-9b",
+    "qwen/qwen3.6-27b",
+    "qwen/qwen3.6-35b-a3b",
+    "qwen/qwen3.8-max"
+  ],
+  "array_alignment": "performance, cost_cny, and latency_ms use model_ids order"
+}

metadata/text-models.json

@@ -0,0 +1,70 @@
+{
+  "scenario": "text",
+  "model_count": 63,
+  "model_ids": [
+    "openai-gpt-5-6-sol",
+    "openai-gpt-5-6-terra",
+    "openai-gpt-5-6-luna",
+    "openai-gpt-5.4-mini",
+    "openai-gpt-5.4-nano",
+    "openai-gpt-5-mini",
+    "openai-gpt-4.1",
+    "openai-gpt-4.1-mini",
+    "openai-o4-mini",
+    "openai-gpt-oss-20b",
+    "openai-gpt-oss-120b",
+    "moonshot-kimi-k3",
+    "minimax-minimax-m3",
+    "minimax-minimax-m2.7",
+    "z-ai-glm-5.2",
+    "z-ai-glm-4-32b-0414",
+    "z-ai-glm-4.5-air",
+    "z-ai-glm-4.7",
+    "deepseek-deepseek-v4-pro",
+    "deepseek-deepseek-v4-flash",
+    "deepseek-deepseek-r1-distill-qwen-1.5b",
+    "deepseek-deepseek-r1-distill-qwen-7b",
+    "deepseek-deepseek-r1-distill-qwen-14b",
+    "deepseek-deepseek-r1-distill-qwen-32b",
+    "deepseek-deepseek-r1",
+    "qwen-qwen3.8-max",
+    "qwen-qwen3.7-max",
+    "qwen-qwen3.7-plus",
+    "qwen-qwen3-14b",
+    "qwen-qwen3-32b",
+    "meta-llama-meta-llama-3.1-8b-instruct",
+    "qwen-qwen3-30b-a3b",
+    "z-ai-glm-5.3",
+    "google-gemini-3.1-flash-lite",
+    "google-gemini-3.5-flash",
+    "anthropic-claude-fable-5",
+    "anthropic-claude-opus-5",
+    "qwen-qwen3-0.6b",
+    "qwen-qwen3-1.7b",
+    "qwen-qwen3-4b",
+    "qwen-qwen3-8b",
+    "qwen-qwen2.5-coder-0.5b",
+    "qwen-qwen2.5-coder-3b",
+    "qwen-qwen2.5-coder-7b",
+    "google-gemma-4-e2b",
+    "google-gemma-3-4b-it",
+    "microsoft-phi-3.5-mini-instruct",
+    "microsoft-phi-4-mini-instruct",
+    "qwen-qwen2.5-coder-14b-instruct",
+    "qwen-qwen2.5-vl-3b-instruct",
+    "qwen-qwen2.5-vl-7b-instruct",
+    "meta-llama-llama-3.2-1b-instruct",
+    "meta-llama-llama-3.2-3b-instruct",
+    "google-gemma-2-9b-it",
+    "opengvlab-internvl3-1b",
+    "opengvlab-internvl3-2b",
+    "opengvlab-internvl3-8b",
+    "opengvlab-internvl3-9b",
+    "mistralai-ministral-3-3b-instruct-2512",
+    "mistralai-ministral-3-8b-instruct-2512",
+    "mistralai-ministral-3-8b-reasoning-2512",
+    "mistralai-ministral-3-14b-instruct-2512",
+    "microsoft-phi-4-multimodal-instruct"
+  ],
+  "array_alignment": "performance, cost_cny, and latency_ms use model_ids order"
+}

normalized-schema.json

@@ -0,0 +1,60 @@
+{
+  "schema_version": "robust-route-bench-v1.0.0",
+  "layout": {
+    "configs": [
+      "text",
+      "multimodal",
+      "chinese"
+    ],
+    "splits": [
+      "train",
+      "validation",
+      "test"
+    ],
+    "validation_overlaps_test": true,
+    "test_slices": [
+      "standard",
+      "robust"
+    ]
+  },
+  "row_format": "one query with candidate-model-aligned arrays",
+  "fields": {
+    "scenario": "string:not_null",
+    "published_split": "string:not_null",
+    "source_split": "string:not_null",
+    "evaluation_type": "string:not_null",
+    "query_id": "string:not_null",
+    "base_query_id": "string:nullable",
+    "is_validation": "bool:not_null",
+    "domain": "string:not_null",
+    "source_dataset": "string:nullable",
+    "source_sample_id": "string:nullable",
+    "variant_type": "string:not_null",
+    "query_text": "large_string:nullable",
+    "input_json": "large_string:not_null",
+    "media_json": "large_string:not_null",
+    "provenance_json": "large_string:not_null",
+    "embedding_model": "string:not_null",
+    "embedding": "fixed_size_list<float32,1024>:not_null",
+    "model_ids": "list<string>:not_null",
+    "performance": "list<float32>:not_null",
+    "cost_cny": "list<float64>:not_null",
+    "latency_ms": "list<float64>:not_null"
+  },
+  "array_alignment": "performance[i], cost_cny[i], and latency_ms[i] describe model_ids[i]",
+  "cleaning_policy": {
+    "non_successful_result": "performance is set to zero",
+    "source_diagnostics_published": false,
+    "raw_model_responses_published": false
+  },
+  "cost_policy": {
+    "currency": "CNY",
+    "usd_to_cny_rate": 7.2,
+    "rate_is_frozen_benchmark_constant": true
+  },
+  "media_policy": {
+    "self_contained": true,
+    "external_download_required": false,
+    "location": "media/multimodal/"
+  }
+}