SAVRN Model Hub · Comparisons
gender-classification vs nsfw_image_detection
Gender-classification has 86M parameters and nsfw_image_detection has 86M parameters; at 16-bit, gender-classification needs about 0.2 GB (1x MI300X from $1.85 an hour) and nsfw_image_detection about 0.2 GB (1x MI300X from $1.85 an hour).
| Field | gender-classification rizvandwiki/gender-classification | nsfw_image_detection Falconsai/nsfw_image_detection |
|---|---|---|
| Publisher | Rizvan Dwikifirdaus | Falcons.ai |
| Task | Image classification | Image classification |
| Modality | Image | Image |
| Parameters, as reported | 86M parameters | 86M parameters |
| Architecture | ViTForImageClassification | ViTForImageClassification |
| Library | transformers | transformers |
| Context length | Not stated | Not stated |
| Repository size | 686.6 MB | 1.5 GB |
| Artifact formats | safetensors, pytorch, tensorboard | safetensors, pytorch |
| License | Not stated | apache-2.0 |
| Access | Open weights, no gate | Open weights, no gate |
| Memory at 16-bit (weights and margin) | 0.2 GB | 0.2 GB |
| Cheapest GPUs at 16-bit, per hour | 1x MI300X, $1.85 | 1x MI300X, $1.85 |
| Memory at 4-bit (weights and margin) | 0.1 GB | 0.1 GB |
| Cheapest GPUs at 4-bit, per hour | 1x MI300X, $1.85 | 1x MI300X, $1.85 |
| Revision viewed | 380b17658cc0 | 96cb0d0342c7 |
| Downloads reported by the hub | 1.4M | 1.8M |
| Last observed | 2026-10-09 | 2026-10-09 |
An evaluation row appears only where at least two of these models report the same benchmark with the same stated configuration, metric, unit and setup. Different evaluators stay named in each cell. Values are shown as reported: no unit conversion, no ranking.
SAVRN's Notes on gender-classification
This one was autogenerated by HuggingPics, and the page reflects it: a ViT with 12 layers and 86M parameters, stored in float32, built to sort images by gender, with no evaluations reported and no training data named. At 16-bit the weights take 0.2 GB, which is also the total needed, and 8-bit halves that to 0.1 GB. The cheapest setup on our Index, one MI300X with 192 GB at $1.85 an hour on-demand, is far more card than this asks for, so share it with other image models.
The license field is empty: no summary, no commercial-use answer, and open access to the files does not change that. Resolve it with the publisher before any deployment; until then treat this as evaluation-only. Note too the 2023-05-18 last update and the absence of any named base model or dataset, which leaves nothing to audit the training against.
SAVRN's Notes on nsfw_image_detection
An 86M-parameter ViT fine-tune — the card names google/vit-base-patch16-224-in21k, ImageNet-21k pretraining, 224×224 input — for NSFW image screening. The repo carries 1.5 GB of float32 in bin, pt and safetensors copies; at 16-bit the model needs 0.2 GB and at 4-bit effectively nothing. The $1.85-an-hour MI300X in the Runs On table is a formality: this belongs on a CPU next to your upload pipeline, and 2.2M monthly downloads suggest exactly that use.
Apache 2.0 permits commercial moderation use with notices kept, and no host on our Index prices it per token. Treat it as a first-pass filter, not a decision-maker: single-label ViT classifiers age fast, and while the repo was updated 2026-09-07, the architecture is a 2020 design. Calibrate the threshold on your own data before trusting the default operating point — the card describes the model, not its precision and recall.
Questions
Which is larger, gender-classification or nsfw_image_detection?
gender-classification (86M parameters) is larger than nsfw_image_detection (86M parameters), by the parameter counts their publishers report.
Which is cheaper to run, gender-classification or nsfw_image_detection?
At 4-bit, gender-classification fits on 1x MI300X from $1.85 an hour and nsfw_image_detection on 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use nsfw_image_detection commercially?
Yes. nsfw_image_detection is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.