<code>v2</code> Increased images from <code>146</code> to <code>217</code>. Replaced a few images with higher quality. Increased <code>network_dim</code> from <code>32</code> to <code>128</code>. Increased training resolution from <code>512</code> to <code>576</code>. Increased <code>offset_noise</code> from <code>0.0</code> to <code>0.0025</code>. Decreased epochs from <code>15</code> to <code>10</code>. Seems a little cooked, so you may have to decrease weight to <code>0.8</code> <code>0.9</code>. Does any one know what training setting causes this? Is it the <code>network_alpha</code>? I set mine to <code>64</code>, half of <code>network_dim</code>:<img src="https://i.imgur.com/bKWm1j0.png" alt />

Anne Hathaway

pics: 217blip: 20 - 100<code>resolution</code>: 576<code>num_repeats</code>: 10<code>num_epochs</code>: 10<code>network_dim</code>: 128<code>network_alpha</code>: 64<code>min_snr_gamma</code>: 0.0025steps: 10% 364/3640 [11:50&lt;1:46:37, 1.95s/it, loss=0.0842]steps: 20% 728/3640 [23:51&lt;1:35:24, 1.97s/it, loss=0.0852]steps: 30% 1092/3640 [35:51&lt;1:23:40, 1.97s/it, loss=0.0836]steps: 40% 1456/3640 [47:52&lt;1:11:48, 1.97s/it, loss=0.0781]steps: 50% 1820/3640 [59:54&lt;59:54, 1.97s/it, loss=0.0814]steps: 60% 2197/3640 [1:12:33&lt;47:39, 1.98s/it, loss=0.0791]crashed: out of memory (15 GB card)steps: 25% 364/1456 [11:52&lt;35:36, 1.96s/it, loss=0.0788]steps: 50% 728/1456 [23:55&lt;23:55, 1.97s/it, loss=0.0788]steps: 75% 1092/1456 [35:57&lt;11:59, 1.98s/it, loss=0.076]steps: 100% 1456/1456 [47:58&lt;00:00, 1.98s/it, loss=0.0766]

Files with SHA256: 365c18b92009527e87cb8497632247a08766667b6167aa8210789315f8a0ae74