datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
imagenet-variations-synth-pilot-v3
ImageNet Variations Synth Pilot v3.1 (diverse, 1K)
Huu + Claude multimodal instruction pipeline — quality-fixed re-run.
Pipeline
Flux.1-schnell (Huu style phrasings + visual style boosts + aspect ratios)
→ Seed2 → Florence-2 grounding/caption → tracks A–F → USER / ASSISTANT.
Tracks
A programmatic QA (counts/style/spatial/absence) from detector
B LLaVA-style (conversation / detailed description / complex reasoning)
C Evol-Instruct (seed atypical Q →… See the full description on the dataset page: https://huggingface.co/datasets/EmpathicRobotics/imagenet-variations-synth-pilot-v3.Variational-DAPO
Dataset Card for SvS/Variational-DAPO
[🌐 Website] •
[🤗 Dataset] •
[📜 Paper] •
[🐱 GitHub] •
[🐦 Twitter] •
[📕 Rednote]
This dataset consists of 314k variational problems synthesized by the Qwen2.5-32B-Instruct policy during RLVR training on DAPO-17k using the SvS strategy for 600-step training, each accompanied by reference answers.The variational problems undergo a min_hash deduplication with a threshold of 0.85.
Data Loading
from datasets import… See the full description on the dataset page: https://huggingface.co/datasets/RLVR-SvS/Variational-DAPO.repro-sharp-inequalities-between-total-variation-and-hellinger-distances-for-gaussian-traces
Agent traces
Agent sessions published from a Trackio Logbook.
imagenet-variations-synth-pilot-v4
imagenet-variations-synth-pilot-v4
Synthetic multimodal instruction data. Prompts come from laion/imagenet_variations;
images are generated with FLUX.1-schnell, encoded to 32 SEED-2 tokens per image,
grounded with Florence-2, and turned into USER / ASSISTANT records.
This is v4, rebuilt from v3.1 to cover the review feedback of 2026-08-15.
What is new vs v3.1
Axis
v3.1
v4
Reasoning
none
<think> traces on ~50% of records
Prompts per image
1
1-4 turns… See the full description on the dataset page: https://huggingface.co/datasets/EmpathicRobotics/imagenet-variations-synth-pilot-v4.VariationsFrameTasksvariational-sd-qwen3-8b-sharegpt-rollouts
Qwen3-8B Regenerated ShareGPT Rollouts
This is the exact target-regenerated ShareGPT JSONL used to train the
Qwen3-8B D-PACE A512 one-epoch checkpoint and the discrete M=8 A512 one-epoch
checkpoint in
Nicholas0228/variational-sd.
Contents
data/train.jsonl: 78,810 successful regenerated rows, 983,686,208 bytes.
metadata.json: generation settings, row statistics, and checksum.
Each JSONL row contains:
{
"id": string,
"status": "success",
"conversations": [… See the full description on the dataset page: https://huggingface.co/datasets/Nicholas0228/variational-sd-qwen3-8b-sharegpt-rollouts.tt638b-four-function-prompt-variation-v1
TT638B Four Function Prompt Variation
Trains on multiple prompt phrasings for ADD/SUBTRACT/MULTIPLY/DIVIDE.
It also includes held-out prompt variants for evaluation.
Seen gate:
The model should pass all trained prompt phrasings.
Held-out diagnostic:
The model may or may not pass unseen phrasings. That measures prompt generalization.
imagenet_variations