REAP the Experts: Why Pruning Prevails for One-Shot MoE compression
Paper • 2510.13999 • Published • 19
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: TypeError
Message: Couldn't cast array of type string to null
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 295, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2281, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2233, in cast_table_to_schema
cast_array_to_feature(
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 1804, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2095, in cast_array_to_feature
return array_cast(
^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 1806, in wrapper
return func(array, *args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 1957, in array_cast
raise TypeError(f"Couldn't cast array of type {_short_str(array.type)} to {_short_str(pa_type)}")
TypeError: Couldn't cast array of type string to null
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1347, in compute_config_parquet_and_info_response
parquet_operations = convert_to_parquet(builder)
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 980, in convert_to_parquet
builder.download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 884, in download_and_prepare
self._download_and_prepare(
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 947, in _download_and_prepare
self._prepare_split(split_generator, **prepare_split_kwargs)
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1739, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1925, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id string | repo_id string | source_id string | split string | subset null | text string |
|---|---|---|---|---|---|
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000000:40a88b4e6975f710 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Document:
Sarah Johnson is a Software Engineer at level 2 in the Engineering department, where she has been employed for 4.5 years. Her performance over recent review cycles reflects consistent excellence and steady professional growth. In the Q1 2023 performance review, she received a rating of 4.2 out of 5.0,... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000001:b2e941b65fd9b286 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Problem:
Convert the unstructured text into the specified structured format.
The Silver Jubilee Anniversary Dinner is a formal celebration scheduled to take place on June 15, 2025, with the official start time set at 7:00 PM (19:00:00 in 24-hour format). The event will be held at The Grand Meridian Hall, locat... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000002:589e65ca52249256 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Instrutions:
Fit the document to the given format.
# Document:
This workplace diversity report corresponds to Organization ID ORG789 and covers the reporting year 2023. The organization employs a total of 850 individuals across various departments. A key metric, the diversity index, measures the variety of ide... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000003:79993a1f15bc4fb8 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Instrutions:
Ensure your output validates against the given JSON schema.
# Document:
Water is essential for maintaining optimal human health and plays a multifaceted role in bodily functions. It is critical for cellular function, enabling metabolic processes, maintaining cell structure, and facilitating the tr... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000004:f6d02f6f8e703417 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Format your response as a JSON object adhering to:
- Schema structure: {"title": "WardrobeOrganizationPlan", "type": "object", "additionalProperties": false, "required": ["seasonalStorage", "clothingCategories", "outfitTemplates", "accessoryHierarchy", "colorPalette", "sustainabilityRating", "purchaseIntent", "fi... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000005:52a18d94ed722058 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Instrutions:
Fit the document to the given format.
# Document:
The ocean exploration mission, designated OM-202307 and titled "Deep Atlantic Hydrothermal Vent Survey," was conducted to investigate hydrothermal vent ecosystems along the Mid-Atlantic Ridge. Led by the Woods Hole Oceanographic Institution (WHOI) ... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000006:3835dc4e6cd65b68 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Problem:
Parse the document and populate the following data model.
The PlayStation 5, developed by Sony, a multinational technology company headquartered in Tokyo, Japan, represents the latest evolution in home gaming consoles. Sony, founded in 1946, continues its legacy in innovation with the PS5, officially ... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000007:6d951a0423d0e857 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Instrutions:
Ensure your output validates against the given JSON schema.
# Document:
The individual, Maria Thompson, was born on March 15, 1952. She maintains a consistent and active lifestyle, engaging in physical activity five times per week. Her primary form of exercise is walking, which she performs for an... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000008:3b08d43f09ee6f4a | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Document:
The traveler is preparing for a leisure trip to Kyoto, Japan, a city known for its rich cultural heritage, historic temples, and temperate climate. The journey will span 7 days and 6 nights, taking place during the autumn season—a particularly scenic time of year when the foliage transforms into vibra... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000009:f2fd4eaf341376a8 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Response Formatting Schema: {"type": "json_schema", "json_schema": {"name": "scene_description", "schema": {"$id": "https://example.com/FamilyHealthAndNutritionDiscussionSchema", "type": "object", "additionalProperties": false, "required": ["familyMember", "dailyNutritionLog", "parentingApproach"], "properties": ... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000010:15d132f05dda0fd1 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Response Formatting Schema: {"type": "object", "required": ["child", "schoolDay", "subjects", "homeworkStatus", "extracurriculars", "emotions", "parentNotes"], "properties": {"child": {"type": "object", "required": ["name", "age", "grade"], "properties": {"name": {"type": "string"}, "age": {"type": "integer", "mi... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000011:c6d15723c592ae16 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Instrutions:
Fit the document to the given format.
# Document:
The dog park visit took place on August 17, 2023, a sunny Thursday morning ideal for outdoor activities. The outing was assigned a unique identifier—DPV-2023-08-17-001—which serves as the official visit ID for tracking and record-keeping purposes. ... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000012:cff8bfa065dd0a27 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Response Formatting Schema: {"type": "object", "required": ["reportId", "consumer", "creditScore", "scoreDetails", "creditHistory", "inquiries", "publicRecords", "recommendations", "reportGeneratedAt"], "properties": {"reportId": {"type": "string", "format": "uuid"}, "consumer": {"type": "object", "required": ["f... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000013:e670804cd7a0098d | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Response Formatting Schema: {"type": "json_schema", "json_schema": {"name": "scene_description", "schema": {"type": "object", "title": "TravelSafetyAssessment", "required": ["destination", "travelerInfo", "safetyChecklist"], "additionalProperties": false, "properties": {"destination": {"type": "object", "required... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000014:67b22c06176106cc | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Document:
The housing scenario described pertains to renting a home, which represents a short-term, flexible housing solution with distinct financial and lifestyle implications. Renting involves lower initial and recurring financial obligations compared to homeownership, but it does not contribute to long-term ... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000015:874d56167bd22435 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Structure your response according to the following JSON schema specification: {"type": "json_schema", "json_schema": {"name": "scene_description", "schema": {"type": "object", "properties": {"strategyName": {"type": "string"}, "monthlyIncome": {"type": "number"}, "expenseCategories": {"type": "object", "propertie... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000016:6e496f304c224ac4 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Create a structured JSON response that:
1. Implements proper data typing
2. Includes all required fields
3. Handles special characters appropriately
4. Is unindented JSON format
5. Validates against schema constraints
6. Provides appropriate array formatting
7. Uses consistent formatting and escaping
8. Maintains... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000017:d3e094ec1119944d | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Information
The community clean-up event is scheduled for Saturday, June 15, 2024, and will take place from 9:00 AM to 12:00 PM at Westside Community Park, located at 123 Main Street. This initiative is organized by the Westside Neighborhood Association in collaboration with the city's Department of Public Work... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000018:a09beb7e3134df3f | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
# Information
The cryptocurrency featured in the portfolio summary is Ethereum, identified by the trading symbol ETH. It is classified as a "coin," which signifies that it operates on its own independent, dedicated blockchain network. As a native digital currency within the Ethereum ecosystem, ETH is not built on... |
structured_json_schema:nvidia/Nemotron-RL-instruction_following-structured_outputs:000019:b15773fbe9a20138 | nvidia/Nemotron-RL-instruction_following-structured_outputs | structured_json_schema | train | null | User:
Structure your response according to the following JSON schema specification: {'type': 'json_schema', 'json_schema': {'name': 'scene_description', 'schema': {'title': 'InvestmentPortfolioAnalysis', 'type': 'object', 'properties': {'version': {'type': 'string'}}, 'required': ['version'], 'additionalProperties': Fa... |
End of preview.
Support this work: donate.sybilsolutions.ai
REAP surfaces: GLM | MiniMax | Qwen | Gemma | Paper | Code | PR17 | Cerebras Collection
structured-outputs-calibration-v1
Structured-output calibration set for REAP observer runs, focused on preserving:
- strict JSON generation
- schema-conditioned JSON responses
- Mermaid diagram generation
- fenced Mermaid block formatting
Contents
data.jsonl: normalized calibration recordssummary.json: build summary with source countsconfig.yaml: source mix used to produce the dataset
Source mix
nvidia/Nemotron-RL-instruction_following-structured_outputs— 160Arun63/sharegpt-structured-output-json— 30Celiadraw/text-to-mermaid— 120sts07142/mermaid_samples_13k(wrapped) — 120
Total samples: 430
Record format
Each JSONL row contains:
idsource_idrepo_idsubsetsplittext
text is normalized into prompt/response-style plain text suitable for calibration / observation workloads.
- Downloads last month
- 81