Datasets:

CollovLabs
/

vi-benchmark-pilot

License:

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

This dataset contains video derived from third-party research datasets under their own licences (several non-commercial). Access is for non-commercial research only. You agree to follow each source licence listed in video/manifest.jsonl and not to redistribute the footage.

VI-Benchmark pilot

Video-agent benchmark scenarios (VI-ToolSandbox) and agent runs from the pilot. Layout: video/scenarios/<slug>/ (task + footage the agent sees), video/runs/<group>/... (trajectories, grades), video/manifest.jsonl (per scenario: sources, licence, included or withheld and why). Scenarios whose licence does not allow redistribution, and anything from Ego4D, are withheld and listed by slug only. Answers/rubrics are not published here. Revisions are tagged per benchmark version (e.g. bench-v1.2).

Visual shopping: Kimi K3 traces for the Qwen3.5-9B fine-tune

visual-shopping/ holds the second lane, the Moving-In Sofa Hunt: Sam moves from Berkeley to a San Francisco flat and asks an assistant (with mail, photos, saved memory, old chats and a shop catalog) to pick a sofa plus a rug under $1,500. The rule that matters is never said: the sofa must fit through a 30 in door. Kimi K3 passes about 80 % of the frozen set, Qwen3.5-9B 0 %. The traces below are Kimi K3 runs for fine-tuning Qwen3.5-9B on this scenario.

Pick up in four steps

Access. Request access on this repo and on the answer-key repo CollovLabs/vi-benchmark-pilot-oracle (both gated, manual approval). Ask for read access to the GitHub repos flair-home-stylist/vi-ts-generation and flair-home-stylist/vi-toolsandbox (runner, agents, judge).
Get the traces. They are uploaded in batches while the Kimi runs continue (target about 100 passing runs), so download from main and re-run the command to pick up new batches:
```
hf download CollovLabs/vi-benchmark-pilot --repo-type dataset --include "visual-shopping/traces/v1/*" --local-dir vi-benchmark-pilot
```
When the runs finish, the final set is tagged shopping-v1-traces-v1 (use --revision shopping-v1-traces-v1). To pin a batch before that, pass its commit hash as --revision.
Train on passing runs. traces/v1/runs/<run_id>.jsonl is one fine-tune example per run: OpenAI-style tools and messages (the agent's view: system, Sam's turns, assistant tool calls, tool results; Kimi's reasoning in a separate reasoning field; images by path into traces/v1/images/). passing.txt lists the runs to train on; failing runs are kept for analysis. manifest.jsonl has one row per run (variant, paraphrase seed, pass, per-item verdicts, tokens).
Evaluate. The variant moving-in-s-rich-a7h is held out: it has no traces, never train on it. To run Qwen before and after fine-tuning, download the world and scenarios (tag shopping-v1) plus the answer key, assemble a scenario root and run the runner; the exact commands, the grading (code checks + an evidence-grounded rubric judge, pass = all of both) and the frozen numbers are in visual-shopping/README.md (sections Reproduce, Train / test split, Traces).

Licences for the shopping world (synthetic files, AppWorld mail as filler, HierText and WorldVQA filler photos; some carry redistribution caveats) are listed in visual-shopping/README.md under Licences and sources.

Downloads last month: 115

Total file size:

177 GB