Not an easy LoRa to work with, but it can work.
v2 is a slight improvement, but I still think I can do better. I think the training data needs higher quality.
Description
PICS:
EPOCHS: 10
REG PICS: none
BLIP: 5 - 75
steps: 10% 95/950 [02:16<20:25, 1.43s/it, loss=0.149]
steps: 20% 190/950 [04:40<18:40, 1.47s/it, loss=0.141]
steps: 30% 285/950 [07:04<16:30, 1.49s/it, loss=0.148]
steps: 40% 380/950 [09:28<14:12, 1.50s/it, loss=0.141]
steps: 50% 475/950 [11:52<11:52, 1.50s/it, loss=0.138]
steps: 60% 570/950 [14:15<09:30, 1.50s/it, loss=0.141]
steps: 70% 665/950 [16:39<07:08, 1.50s/it, loss=0.134]
steps: 80% 760/950 [19:03<04:45, 1.50s/it, loss=0.135]
steps: 90% 855/950 [21:27<02:23, 1.51s/it, loss=0.13]
steps: 100% 950/950 [24:02<00:00, 1.52s/it, loss=0.132]
FAQ
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.