mlfoundations / datacomp

DataComp: In search of the next generation of multimodal datasets

http://datacomp.ai/

mlfoundations/datacomp Issues

Downloading Commonpool XLarge
Updated 16 days ago
Average caption length for CommonPool
Closed a month ago1
Availability of npy indices for large pool
Updated 2 months ago
ModuleNotFoundError: No module named 'training'
Closed 2 months ago2
Connection error while half-downloading metadata
Updated 2 months ago3
`zeroshot_templates` split error for FairFace / UTKFace
Closed 3 months ago9
Conda environment build issue
Closed 7 months ago3
About update metadata with the corresponding image sample in shards
Updated 4 months ago2
Remove CSAM, if present
Updated 4 months ago2
Frequency of Leaderboard Updates
Closed 4 months ago1
Training log
Closed 4 months ago1
Metadata for datacomp-large text-based filter
Closed 4 months ago1
Pretraining dataset
Updated 5 months ago1
Deduplication against evaluation sets
Closed 6 months ago1
Usage with AWS S3 and Ray
Updated 7 months ago5
14% of SHA256 hashes not matching
Updated 7 months ago32
Tried evaluate the model on a local network only machine
Updated 7 months ago4
the normal success rate and downloading speed?
Updated 7 months ago1
--output_dir does not do correct thing if --output_dir is a cloud path
Updated 7 months ago1
Text search over CommonPool
Closed 8 months ago1
Dataset Size on Leaderboard
Closed 8 months ago1
FMoW dataset and results variance
Closed 8 months ago1
Will the training framework do upsampling when train-num-samples is far more than the amount of actual data
Closed 8 months ago8
Not able to push data to google cloud storage
Updated 8 months ago1
In image-based filtering, only imagenet L14 is supported, making extracting B32 embeddings useless
Closed 8 months ago6
Workshop submission deadline
Closed 8 months ago3
Appendix in the workshop paper submission
Closed 8 months ago2
How to precompute and save model-based metric during download?
Closed 8 months ago2
Get 400 error when submitting jsonl to firebase, but successfully submit Slack notification
Closed 9 months ago2
Adding the detected_language to metadata
Closed 9 months ago1
How to achieve exact same # of samples seen?
Closed 9 months ago12
Metadata download error - OSError: Consistency check failed
Closed 9 months ago7
https://huggingface.co/datasets/djghosh/wds_flickr_1k_test_image_text_retrieval_test doesn't exist?
Closed 9 months ago3
download data, why success is always 0.14, but the file size is 80% of the total file (total file size is in README)
Updated 9 months ago7
[Challenge rules] Is it allowed to modify the texts of the data in the filtering track?
Closed 9 months ago4
Result submission deadline
Closed 9 months ago1
Workshop paper submission
Closed 10 months ago1
Downloading DataComp-1B
Updated 10 months ago1
download data
Closed 10 months ago1
train/test splits for downstream tasks
Closed 10 months ago1
Missing training file?
Closed 10 months ago1
Deduplication of eval datasets
Updated 10 months ago1
Leaderboard update
Closed 10 months ago7
FileNotFoundError while downloading DataComp-1B
Updated 10 months ago7
Can you share the CLIP score calculation script?
Closed a year ago10
Problems evaluating trained model
Closed a year ago1
Consistency between Table 23 and Fig 3
Closed a year ago5
Is there overlap between common-pool and laion-5B?
Closed a year ago1
Is there any evaluation randomness?
Closed a year ago1
Any plan to release the baseline checkpoints in the paper?
Closed a year ago1