2xLiveActionV1_SPAN, Core ML
2xLiveActionV1_SPAN by jcj83429, a 2x SPAN model for live action video and film, converted to Core ML.
- Original: https://github.com/jcj83429/upscaling/tree/5d8cdd2e17750b64be39ccab3a8763d91128fe15/2xLiveActionV1_SPAN (Apache-2.0, see LICENSE and NOTICE)
- Changes: converted from PyTorch to Core ML (float16, 384 px tiles); the parallel convolutions of each block are merged into one, as the original does at inference. No retraining.
- Input:
input(RGB, 384 x 384); output:output(RGB, 768 x 768)
- Downloads last month
- 16
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support