Search results

All resources

20 shown of 28,913

LocateAnything 3B ONNX WebGPU INT4

In-browser (onnxruntime-web / WebGPU) build of nvidia/LocateAnything-3B, a visual-grounding / open-vocabulary detector. The language tower is weight-only INT4 and the embedding table is true group-wise 4-bit.

Text modelnvidia/LocateAnything-3B

Status

Listed

ModelExternal source

镭明闪击 Wrap Kinghts Collection - Illustrious (Midori)

Merger_kai_rainard

another merger from my TI's

TextualInversionSD 1.5

Status

Listed

ModelExternal source

6a577e1dbcebf5a38ac76d09.safetensors

Dongfangyuechu701 Style (Anima & Illustrious)

Works best at 0.75 weight.

Invisible Stitch

Invisible Stitch: Generating Smooth 3D Scenes with Depth Inpainting

model.safetensors

KsF8SCCwjXTYCcnnbOEDR_pytorch_lora_weights.safetensors

Surrounded By Hands

Works best at 0.75 weight.

30s Technicolor Movie

Hplay Style

Works best at 0.75 weight.

Aleksiremeart Style (Anima & Illustrious)

Works best at 0.75 weight.

flux_2_4b_shape_lm_3_512x512.safetensors

diffusion_modelFLUX.2

Status

Listed

ModelExternal source

controlnetfluxCnflux_xlabsaiDepthV3.safetensors

Videomae Base Finetuned Kinetics 0403 Final Only5sec

This model is a fine-tuned version of MCG-NJU/videomae-base-finetuned-kinetics on an unknown dataset. It achieves the following results on the evaluation set: Loss: 0.9866 Accuracy: 0.7954

OtherMCG-NJU/videomae-base-finetuned-kinetics

Status

Listed

ModelCivitai