drpo
Datasets
All datasets matching “drpo”qwen35-4b-drpo-vs0f49th-trainer-logprobs
Qwen3.5 4B DRPO trainer logprobs from W&B run vs0f49th
This dataset contains the raw trainer-logprob JSONL shards saved by W&B run ai2-llm/open_instruct_internal/vs0f49th (qwen35_4b_drpo__42__1782345587).
Contents
Source run: https://wandb.ai/ai2-llm/open_instruct_internal/runs/vs0f49th
Source path: /weka/oe-adapt-default/allennlp/deletable_rollouts/
Filename pattern: qwen35_4b_drpo__42__1782345587_trainer_logprobs_step*_rank*.jsonl
Files: 4320 JSONL shards… See the full description on the dataset page: https://huggingface.co/datasets/hamishivi/qwen35-4b-drpo-vs0f49th-trainer-logprobs.data-dive-google-directions
Google Directions traversal-time archive — UIUC campus OD pairs (Apr 23 – Apr 29 2026)
60-second polling of the Google Directions API on 8 fixed origin-destination
pairs across the UIUC campus, intended for transit-corridor traversal-time
analysis under varying pedestrian and traffic load.
Window: 2026-04-23 ~22:00 UTC → 2026-04-29 23:42 UTC (~5 days live data).
OD pairs (each polled bidirectionally)
pair label
A
B
union-engquad
Illini Union
Engineering Quad… See the full description on the dataset page: https://huggingface.co/datasets/dr-pod/data-dive-google-directions.data-dive-campus-cams
UIUC Campus Cams (Apr 23 – Apr 29 2026)
Frame-level archive from three public UIUC YouTube live streams, captured
at 1 Hz (after a brief startup ramp; see per-camera intended_interval_s).
Paired with per-frame metadata (timestamp, brightness, capture status).
Capture window: 2026-04-23 05:27:28 UTC → 2026-04-29 23:41:52 UTC
(~6.76 days, ~1.55M frames total across the three cameras, 489 hourly
WebDataset shards).
Collected as part of an academic research project on UIUC… See the full description on the dataset page: https://huggingface.co/datasets/dr-pod/data-dive-campus-cams.data-dive-mtd-gtfs-rt
CUMTD GTFS-Realtime archive (Apr 23 – Apr 29 2026)
A timestamped public archive of the Champaign-Urbana Mass Transit District
(CUMTD / MTD) GTFS-Realtime feeds, captured from the public endpoints at
https://gtfs-rt.mtd.org. To my knowledge no continuous public archive of
CUMTD's real-time feeds exists; the live feeds overwrite themselves.
Window: 2026-04-23 ~05:30 UTC → 2026-04-29 23:42 UTC (~6.76 days).
Feeds archived
feed
endpoint
poll interval… See the full description on the dataset page: https://huggingface.co/datasets/dr-pod/data-dive-mtd-gtfs-rt.eval-imdb-drpo-134567-dpo-5000eval-tldr-dpo-drpo-0.9tmp-sft-1000