Dataset Viewer

The dataset viewer is not available for this subset.

Cannot get the split names for the config 'default' of the dataset.

Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 80, in _split_generators
                  raise ValueError(
                  ...<2 lines>...
                  )
              ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
              
              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/split_names.py", line 68, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

T2I-Eval outputs

Generated images, MLLM-judge results and reward-model results from T2I-Eval (snapshot 2026-10-16). One tar per model, each preserving the original outputs/runs/<provider>/<model>/<benchmark>/... layout.

huggingface-cli download Jialuo21/Images --repo-type dataset --local-dir . --include "outputs/*.tar"
for f in outputs/*.tar; do tar -xf "$f"; done   # recreates outputs/runs/...

Per benchmark: gen/ (images), eval/ (MLLM judge), eval_<reward>/ (reward models). Final-table numbers are the overall_score in each eval*/summary.json. manifest.json lists file counts, sizes and sha256 of every tar.

shard	files	GB
`outputs/bagel__bagel-7b-mot.tar`	23658	29.0
`outputs/deepgen__deepgen-1.0-diffusers.tar`	24778	21.8
`outputs/ernie_image__ernie-image.tar`	10324	19.7
`outputs/ernie_image__ernie-image-turbo.tar`	10323	19.0
`outputs/flux__flux.1-dev.tar`	23659	29.4
`outputs/flux__flux.1-schnell.tar`	23659	28.6
`outputs/flux__flux.2-dev.tar`	23679	37.5
`outputs/flux__flux.2-klein-4b.tar`	24919	36.6
`outputs/flux__flux.2-klein-9b.tar`	23658	37.1
`outputs/flux__flux.2-klein-base-4b.tar`	22321	32.8
`outputs/flux__flux.2-klein-base-9b.tar`	15140	22.9
`outputs/gemini__gemini-2.5-flash-image.tar`	10331	16.3
`outputs/glm_image__glm-image.tar`	23657	33.2
`outputs/imagen__imagen-4.0-fast-generate-001.tar`	10311	15.1
`outputs/longcat_image__longcat-image.tar`	23650	33.5
`outputs/lumina_2__lumina-image-2.0.tar`	23656	31.5
`outputs/qwen_image__qwen-image-2512.tar`	23656	37.1
`outputs/sana__sana1.5_1.6b_1024px_diffusers.tar`	23656	36.5
`outputs/stable_diffusion_3__stable-diffusion-3.5-large.tar`	23657	38.7
`outputs/stable_diffusion_3__stable-diffusion-3.5-medium.tar`	23657	36.5
`outputs/stable_diffusion_xl__stable-diffusion-xl-base-1.0.tar`	23660	36.2
`outputs/z_image__z-image.tar`	25773	41.1
`outputs/z_image__z-image-turbo.tar`	23656	31.4

Replaces the previous outputs_20260310.tar (2026-03-10 snapshot, only 4 benchmarks, older evaluations).

Additional experiment outputs (2026-10-20)

path	shards	files	GB
`outputs_batch_eval/*.tar`	19	157359	247.0
`outputs_batch_eval_v2/*.tar`	12	129640	223.5
`figures/` (plain files)	-	155	0.193
`sheet_exports/` (plain files)	-	7	0.000

outputs_batch_eval/ (prompt-ablation runs, incl. _unuse_* experiments) and outputs_batch_eval_v2/ (reward-model ablation runs): one tar per run directory, layout outputs_batch_eval*/<run>/.... Per-dir manifest.json lists file counts, sizes and sha256 of every tar.
figures/ (paper figures + plotting scripts) and sheet_exports/ (result tables, e.g. alignment.txt) are stored as plain files.

hf download Jialuo21/Images --repo-type dataset --include "outputs_batch_eval*/*.tar" --local-dir .
for f in outputs_batch_eval*/*.tar; do tar -xf "$f"; done

Downloads last month: 69

Total file size:

1.17 TB