EfficientNet NSFW moderation record
The record is an EfficientNet image-classification checkpoint whose tags include efficientnet, nsfw, content-moderation, moderators, and not-for-all-audiences.
Open Source Model Profile · viddexa
nsfw-detection-2-mini
nsfw-detection-2-mini is a 17.68M-parameter EfficientNet NSFW classifier from viddexa. Hub tags associate it with image content moderation and an EfficientNet-B4 reference.
- Publisher
- viddexa
- Task
- image-classification
- Model type
- efficientnet
- License
- apache-2.0
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
nsfw-detection-2-mini is published by viddexa as an image-classification model. The captured configuration identifies EfficientNetForImageClassification with model type efficientnet, and Safetensors metadata reports 17,682,781 parameters. Hub tags associate it with NSFW content moderation and an EfficientNet-B4 base-model reference.
Recorded capabilities
17.68M-parameter mini scale
Captured Safetensors metadata reports 17,682,781 parameters.
Base-model and paper tag references
Hub tags record base_model:google/efficientnet-b4 and arxiv:2312.16338, without an extracted model card to verify training lineage or evaluation context.
Transformers and endpoints-compatible format
The hub record lists the transformers library with safetensors format, and tags record endpoints compatibility, model-index presence, and region:us.
Use cases in the source record
- Lightweight image content-moderation screening where a smaller EfficientNet NSFW classifier is relevant, subject to independent validation of labels and thresholds.
- Research comparison among EfficientNet-family moderation classifiers, using the recorded base-model and paper tag references as starting points.
Limitations and unknowns
- No model card was present, so intended uses, training details, labels, thresholds, and evaluation results are unknown.
- No evaluation results were extracted from this record.
- No label definitions or classification thresholds were extracted.
- Provider state is historical snapshot data, not independently refreshed current availability.
Source and provenance
Source: viddexa/nsfw-detection-2-mini
Captured: Unknown. Processed: 2026-09-07T19:36:03.948138+00:00.
F001F002F003F004F005F006F007F008F009F010