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: ArrowInvalid
Message: Length spanned by list offsets (112) larger than values array (length 111)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1816, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
for item in generator(*args, **kwargs):
~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2303, in cast_table_to_schema
cast_array_to_feature(
~~~~~~~~~~~~~~~~~~~~~^
table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
feature,
^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1852, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2116, in cast_array_to_feature
return pa.ListArray.from_arrays(array_offsets, casted_array_values)
~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "pyarrow/array.pxi", line 2755, in pyarrow.lib.ListArray.from_arrays
File "pyarrow/array.pxi", line 1854, in pyarrow.lib.Array.validate
check_status(self.ap.Validate())
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: Length spanned by list offsets (112) larger than values array (length 111)
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 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1869, 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.
responses_create_params unknown | verifier unknown | agent_ref dict | pass_rate float64 | pass_rate_passed int64 | pass_rate_total int64 | _ng_task_index int64 | _ng_rollout_index int64 | raw_rewards_and_lengths list | input_tokens/mean float64 | input_tokens/std float64 | input_tokens/max int64 | input_tokens/min int64 | output_tokens/mean float64 | output_tokens/std float64 | output_tokens/max int64 | output_tokens/min int64 | total_tokens/mean float64 | total_tokens/std float64 | total_tokens/max int64 | total_tokens/min int64 | dataset string | uuid string | metadata unknown | _hf_question_placeholder unknown | template_metadata unknown | exp_cal_state unknown | expected_action unknown | kwargs list | options list | sources list | id unknown | trajectory_id int64 | info dict | ref_patch null | ref_message dict | topic string | problem string | expected_answer string | subtopic string | difficulty_model_pass_rate float64 | difficulty_model_pass_at_n string | difficulty_model string | source string | question_format string | __duplicate__ bool | prompt string | _filled_prompt string | question_type string | raw_question string | question string | principle string | instruction_id_list list | reference_file string | verifier_type string | scenario null | num_unique_actions int64 | meta_info dict | qwen_235b_info dict | reward_profiles list | task_id int64 | rubric list | context string | answer string | domain string | schema_str string | schema_type string | problem_type string | schema_repr string | source_format null | num_turns int64 | source_record_id string | verifier_metadata dict | hash_id string | matched_sources list | adversarial_prompt string | attack_type string | prompt_without_jailbreak string | response_policy_mapped string | hazard_category string | labels list |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
{
"input": [
{
"role": "user",
"content": "Read the following document parts and answer the question. Cite every source part using [N].\n\n[Part 1]:\n# General Political Bureau of the Korean People's Army\n\nGeneral Political Bureau of the Korean People's ArmyNorth Korean Ministry of Defence departmen... | {
"type": "string_match",
"patterns": [
"\\[1\\]",
"\\[3\\]"
],
"expected_markers": [
"[1]",
"[3]"
]
} | {
"type": "responses_api_agents",
"name": "citation_format_simple_agent"
} | 0.375 | 3 | 8 | 338 | 7 | [
{
"_ng_rollout_index": 0,
"reward": 1,
"input_tokens": 665,
"output_tokens": 597,
"total_tokens": 1262
},
{
"_ng_rollout_index": 1,
"reward": 1,
"input_tokens": 665,
"output_tokens": 415,
"total_tokens": 1080
},
{
"_ng_rollout_index": 2,
"reward": 0,
"inpu... | 665 | 0 | 665 | 665 | 551.75 | 108.56536 | 772 | 415 | 1,216.75 | 108.56536 | 1,437 | 1,080 | amplified_dolphin_profile_ds3_citation | 2499b755-205f-445f-bbf0-b0a9736c03af | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{
"input": [
{
"role": "system",
"content": "You are a coding agent. You are expected to be precise, safe, and helpful.\n\nYour capabilities:\n\n- Receive user prompts and other context provided by the harness, such as files in the workspace.\n- Communicate with the user by streaming thinking & respon... | null | {
"type": "responses_api_agents",
"name": "swe_pivot_single_step_tool_use_with_argument_comparison_agent"
} | 0.625 | 5 | 8 | 2,305 | 5 | [
{
"_ng_rollout_index": 0,
"reward": 1,
"input_tokens": 33279,
"output_tokens": 61,
"total_tokens": 33340
},
{
"_ng_rollout_index": 1,
"reward": 1,
"input_tokens": 33279,
"output_tokens": 93,
"total_tokens": 33372
},
{
"_ng_rollout_index": 2,
"reward": 1,
"... | 33,279 | 0 | 33,279 | 33,279 | 83.25 | 21.258822 | 128 | 52 | 33,362.25 | 21.258822 | 33,407 | 33,331 | amplified_dolphin_profile_swe_pivot_len40k | 67441953-01f0-4e40-8e6a-6d8aa3c88a07 | {
"agent_cls": "codex",
"instance_id": "python-pillow__Pillow-e9c9f19c26c3a361c4fa17947b6b5d860e516c2c"
} | null | null | null | {
"type": "function_call",
"name": "shell_command",
"arguments": "{\"command\": \"cd /workspace/python-pillow__Pillow__ && grep -n \\\"def _write_multiple_frames\\\" src/PIL/PngImagePlugin.py\"}"
} | null | null | null | null | 6,222 | {
"turn": 1,
"step": 25,
"depth": 25
} | null | {
"role": "assistant",
"reasoning_content": "It finds the file but I need the line number. Let me get that and try again:",
"content": null,
"tool_calls": [
{
"id": "chatcmpl-tool-9e2bdc270e5cb8a7",
"type": "function",
"function": {
"name": "shell_command",
"arguments": "{\... | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{
"input": [
{
"role": "user",
"content": "<document>\nTitle: Le Bachut\n\n# Le Bachut\n\nLe Bachut**Le Bachut** [lə baʃy] is a quarter in the 8th arrondissement of Lyon.\n\nIt is bisected the avenue Berthelot and rue Marius Berliet. Served by the tram, it is connected to the Presqu'île, to the Hôpita... | {
"type": "regex",
"pattern_id": "kv_quoted",
"verify_regex": [
"\".+\":\\s*\".+\""
],
"verify_min_matches": 3
} | {
"type": "responses_api_agents",
"name": "freeform_formatting_simple_agent"
} | 0.625 | 5 | 8 | 2,152 | 7 | [
{
"_ng_rollout_index": 0,
"reward": 1,
"input_tokens": 614,
"output_tokens": 4586,
"total_tokens": 5200
},
{
"_ng_rollout_index": 1,
"reward": 0,
"input_tokens": 614,
"output_tokens": 330,
"total_tokens": 944
},
{
"_ng_rollout_index": 2,
"reward": 0,
"inpu... | 614 | 0 | 614 | 614 | 2,091.75 | 1,897.215509 | 5,579 | 317 | 2,705.75 | 1,897.215509 | 6,193 | 931 | amplified_dolphin_profile_ds2_freeform | 8b229b2b-dcbf-4090-b3e4-abbe100b5d73 | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{
"input": [
{
"role": "user",
"content": "Which of the following options is correct? Conclude your response by placing the letter of the correct option between double asterisks, as in **A**.\n\nIs the curl of current density (\\(\\nabla \\times \\mathbf{J}\\)) zero in a steady-state situation?A: \\\\... | null | {
"type": "responses_api_agents",
"name": "mcqa_simple_agent"
} | 0.875 | 7 | 8 | 589 | 2 | [
{
"_ng_rollout_index": 0,
"reward": 1,
"input_tokens": 166,
"output_tokens": 5324,
"total_tokens": 5490
},
{
"_ng_rollout_index": 1,
"reward": 1,
"input_tokens": 166,
"output_tokens": 11274,
"total_tokens": 11440
},
{
"_ng_rollout_index": 2,
"reward": 1,
"... | 166 | 0 | 166 | 166 | 7,542.625 | 3,726.769712 | 12,514 | 2,699 | 7,708.625 | 3,726.769712 | 12,680 | 2,865 | amplified_dolphin_profile_stem_mcqa_cot_rima_new | 42109022-63e8-4e09-9915-a196730035dc | null | null | {
"output_regex": "\\*\\*([A-Za-z])\\*\\*"
} | null | null | null | [
{
"A": "\\\\nabla_{\\\\vec{r}} × \\\\vec{J}(\\\\vec{r}') = \\\\vec{B}(\\\\vec{r})"
},
{
"B": "\\\\nabla_{\\\\vec{r}} × \\\\vec{J}(\\\\vec{r}') = \\\\nabla^2 B"
},
{
"C": "\\\\nabla_{\\\\vec{r}} × \\\\vec{J}(\\\\vec{r}) = 0"
},
{
"D": "\\\\nabla_{\\\\vec{r}} × \\\\vec{J}(\\\\vec{r}') = 0"... | null | "so--0055156" | null | null | null | null | Physics | Is the curl of current density (\(\nabla \times \mathbf{J}\)) zero in a steady-state situation?A: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = \\vec{B}(\\vec{r})
B: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = \\nabla^2 B
C: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}) = 0
D: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = 0 | D | Electromagnetism and Photonics | 0.666667 | 2 / 3 | Qwen3-30B-A3B | so | mcq-4 | false | Which of the following options is correct? Conclude your response by placing the letter of the correct option between double asterisks, as in **A**.
{problem} | Which of the following options is correct? Conclude your response by placing the letter of the correct option between double asterisks, as in **A**.
Is the curl of current density (\(\nabla \times \mathbf{J}\)) zero in a steady-state situation?A: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = \\vec{B}(\\vec{r})
B: \\nabla... | mcq | Is the curl of current density (\(\nabla \times \mathbf{J}\)) zero in a steady-state situation?A: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = \\vec{B}(\\vec{r})
B: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = \\nabla^2 B
C: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}) = 0
D: \\nabla_{\\vec{r}} × \\vec{J}(\\vec{r}') = 0 | Is the curl of current density (\(\nabla \times \mathbf{J}\)) zero in a steady-state situation? | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{
"input": [
{
"role": "user",
"content": "Give a Really Shortest Short Description Backstory about My Verse x Dragon Ball Super Crossover Timeline, Aiden Grayson meeting these characters, Goku (Chi-Chi’s Husband), Vegeta (Bulma’s Husband), Gohan (Goku and Chi-Chi’s Son), Videl, Bulma (Vegeta’s Wife),... | null | {
"type": "responses_api_agents",
"name": "genrm_simple_agent"
} | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | ultra_rlhf_blend_v1_len40k | c7e6cdbe-9c12-4579-b352-9d733f9e5c0b | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | You will be given one or more evaluation criteria (rubrics).
Evaluate both responses on EACH criterion individually first, then synthesize an overall judgment.
Criteria:
1. Please act as an impartial judge and evaluate the quality of the responses provided by two AI assistants to the user prompt. Begin your evaluation... | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{"input":[{"role":"system","content":"You are OpenCode, the best coding agent on the planet.\n\nYou (...TRUNCATED) | null | {"type":"responses_api_agents","name":"swe_pivot_single_step_tool_use_with_argument_comparison_agent(...TRUNCATED) | 0.125 | 1 | 8 | 1,253 | 7 | [{"_ng_rollout_index":0,"reward":0.0,"input_tokens":22966,"output_tokens":90,"total_tokens":23056},{(...TRUNCATED) | 22,966 | 0 | 22,966 | 22,966 | 102.75 | 15.904009 | 131 | 83 | 23,068.75 | 15.904009 | 23,097 | 23,049 | amplified_dolphin_profile_swe_pivot_len40k | f08e08d0-56ae-4cfb-a28c-2baa051601df | {
"agent_cls": "opencode",
"instance_id": "numpy__numpy-89dc3cfe0c1ad749b6b7e987883bb3bdddbb7792"
} | null | null | null | {"type":"function_call","name":"bash","arguments":"{\"command\": \"cd /workspace/numpy__numpy__ && p(...TRUNCATED) | null | null | null | null | 3,235 | {
"turn": 1,
"step": 13,
"depth": 13
} | null | {"role":"assistant","reasoning_content":"No existing tests specifically for recarray subclass, but t(...TRUNCATED) | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{"input":[{"role":"user","content":"Tu es un expert en l'éducation nationale française. Explique m(...TRUNCATED) | null | {
"type": "responses_api_agents",
"name": "instruction_following_simple_agent"
} | 0.25 | 2 | 8 | 442 | 4 | [{"_ng_rollout_index":0,"reward":0.0,"input_tokens":105,"output_tokens":1787,"total_tokens":1892},{"(...TRUNCATED) | 105 | 0 | 105 | 105 | 2,194.25 | 868.519106 | 3,327 | 1,112 | 2,299.25 | 868.519106 | 3,432 | 1,217 | amplified_dolphin_profile_instruction_following | d83f44d9-5b13-42cc-b57c-d0ee4709f9cd | null | null | null | null | null | [
{
"num_paragraphs": 2,
"nth_paragraph": 2,
"first_word": "homework"
},
{
"keywords": [
"consequence",
"wake"
]
}
] | null | null | 19470 | null | null | null | null | null | null | null | null | null | null | null | null | null | null | "Tu es un expert en l'éducation nationale française. Explique moi alors comment fonctionne l'épre(...TRUNCATED) | null | null | null | null | null | [
"length_constraints:nth_paragraph_first_word",
"keywords:existence"
] | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{
"input": [
{
"role": "user",
"content": ""
}
],
"temperature": 1,
"top_p": 1
} | null | {
"type": "responses_api_agents",
"name": "math_with_judge_simple_agent"
} | 0.25 | 2 | 8 | 21 | 7 | [{"_ng_rollout_index":0,"reward":0.0,"input_tokens":358,"output_tokens":38693,"total_tokens":39051},(...TRUNCATED) | 358 | 0 | 358 | 358 | 65,690.875 | 16,491.731636 | 87,784 | 38,693 | 66,048.875 | 16,491.731636 | 88,142 | 39,051 | amplified_dolphin_profile_math_cot | 8b7b723a-3c31-4974-9d5b-c0c12b968cbb | null | {"mode":"canonical","dataset":"Skywork/Skywork-OR1-RL-Data","split":"math","row":101820,"lead":"","t(...TRUNCATED) | null | null | null | null | null | [{"source_file":"math_tir/super_rl_with_pass_rate_tir_skywork_no_omni.jsonl","line_number":7106,"dat(...TRUNCATED) | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | super_rl_good_true_non_olympiad_all_available_plus_gpt_cleaned-all.jsonl | math_with_judge | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | ||
{"input":[{"role":"user","content":"Your response should satisfy these conditions:\n1. In your respo(...TRUNCATED) | null | {
"type": "responses_api_agents",
"name": "instruction_following_simple_agent"
} | 0.5 | 4 | 8 | 1,273 | 4 | [{"_ng_rollout_index":0,"reward":0.0,"input_tokens":88,"output_tokens":1728,"total_tokens":1816},{"_(...TRUNCATED) | 88 | 0 | 88 | 88 | 2,036 | 853.443027 | 3,442 | 952 | 2,124 | 853.443027 | 3,530 | 1,040 | amplified_dolphin_profile_instruction_following | 890ced64-cefc-4af4-88cb-2655d84e35bf | null | null | null | null | null | [{"capital_frequency":7,"capital_relation":"less than"},{"first_word":"success"},{"keyword":"height"(...TRUNCATED) | null | null | 32384 | null | null | null | null | null | null | null | null | null | null | null | null | null | null | "Your response should satisfy these conditions:\n1. In your response, words with all capital letters(...TRUNCATED) | null | null | null | null | null | [
"change_case:capital_word_frequency",
"first_word:first_word_sent",
"keywords:word_once"
] | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
{"input":[{"role":"user","content":"Follow these formatting instructions exactly:\nPlease answer usi(...TRUNCATED) | {"type":"regex","pattern_id":"mixed_numbered_heading","verify_regex":["^\\d+\\. \\*\\*.+\\*\\*"],"ve(...TRUNCATED) | {
"type": "responses_api_agents",
"name": "freeform_formatting_simple_agent"
} | 0.625 | 5 | 8 | 1,685 | 6 | [{"_ng_rollout_index":0,"reward":1.0,"input_tokens":2141,"output_tokens":3038,"total_tokens":5179},{(...TRUNCATED) | 2,141 | 0 | 2,141 | 2,141 | 6,572 | 8,421.180262 | 28,623 | 1,545 | 8,713 | 8,421.180262 | 30,764 | 3,686 | amplified_dolphin_profile_ds2_freeform | a095923f-4fe8-4259-a2f9-2ab9dccead75 | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null | null |
End of preview.
Dataset Description:
This dataset provides the training-data blend used for the Reinforcement Learning with Verifiable Rewards (RLVR) stage of the public Nemotron-3.5-Lightning post-training recipe. The blend is consumed by the NeMo RL training recipes through the NeMo Gym agent framework, in which each prompt is paired with an agent/environment that returns a verifiable or judge-based reward. See the recipe for how the blend is used.
The blend mixes NVIDIA-released datasets with several external datasets (see Data Preparation below).
This dataset is ready for commercial or non-commercial uses.
Dataset Owner(s):
NVIDIA Corporation
Dataset Creation Date:
Created on: 2026-08-04
Last Modified on: 2026-08-06
License/Terms of Use:
This dataset is licensed under CC BY-SA 4.0, CC BY 4.0, ODC-BY 1.0, MIT, and Apache 2.0.
Intended Usage:
This dataset is intended for researchers and developers post-training large language models with reinforcement learning using the NeMo RL recipes and the NeMo Gym agent framework. See the Nemotron-3.5-Lightning training guide for end-to-end instructions.
Dataset Composition
The blend is composed of the following datasets (percentage of the blend).
Data Preparation
For the BytedTsinghua-SIA/DAPO-Math-17k and Skywork/Skywork-OR1-RL-Data data in the blend, instead of replicating the data directly, placeholders are used that point to entries in the original datasets. Use the fill_placeholders.py script to download the data from the original datasets into the blend. For each dataset a user elects to use, the user is responsible for checking if the dataset license is fit for the intended purpose.
The dataset blend is preprocessed using the same curriculum technique described in the Nemotron-3-Ultra technical report: samples are ordered from higher pass-rate (easier) to lower pass-rate (harder), ensuring a balanced learning progression.
Dataset Characterization
Data Collection Method
- Hybrid: Human, Synthetic
Labeling Method
- Hybrid: Human, Synthetic, Automated
Dataset Format
Modality: Text
Format: JSONL (NeMo Gym prompt/agent records)
Structure: Text + Metadata
Dataset Quantification
| Samples | Size |
|---|---|
| 92,684 | ~3.9 GB |
Reference(s):
- NeMo RL: https://github.com/NVIDIA-NeMo/RL
- NeMo Gym: https://github.com/NVIDIA-NeMo/Gym
- Nemotron-3.5-Lightning training guide: https://docs.nvidia.com/nemo/rl/nightly/guides/models/nemotron/nemotron-3.5-lightning.html
Ethical Considerations:
NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal developer teams to ensure this dataset meets requirements for the relevant industry and use case and addresses unforeseen product misuse. Please report quality, risk, security vulnerabilities or NVIDIA AI Concerns here
- Downloads last month
- 85