dave_user
207
edits
Changes
no edit summary
[[File:Fade-in progression.png|center|thumb|500x500px|ProGAN: growing progression of the model during training]]
==Results validationTraining configuration and hyperparameters setup==
A total of 6 proGAN models were built and trained, each one with the required number of ''straight-through'' and ''fade-in'' model stages, to generate several typologies of synthesized images at the desired target resolution, belonging to ''missing'' and ''tombstoning'' classes, more specifically ''full'' images (512 × 512 resolution), ''upper'' and ''lower'' soldering region images (256 × 256 resolution). The training was executed mainly on cloud, initially with the free services provided by Google Colab and finally on AWS SageMaker.
Discriminators and generators were trained both with a learn rate of 5 × 10<sup>-4</sup> for all fade-in stages and with a smaller learn rate of 1 × 10<sup>-4</sup> for all ''straight-through'' stages, in order to guarantee a smooth and slow enough fine-tuning for all layers. Each stage of each created model was trained for 100 epochs, with the sole exception of the last stage for the 512×512 full images, which needed 50 more epochs to generate satisfying results. The batch size is progressively reduced as the resolution increases, starting from a batch of 64 images for the first three resolutions (4 × 4, 8 × 8 and 16 × 16 resolutions), decreasing to 32 (32 × 32, and 64 × 64 resolutions) and 16 (128 × 128 and 256 × 256 resolutions). In the last stage for full typology (512 × 512), batch size is further reduced to 8 images for each train step.
{| class="wikitable" style="text-align:center; margin: auto;"
|}
==Results validation==
In the two figures below there are shown some examples taken from the generated sets of ''missing'' and ''tombstoning'' classes. For each image, from left to right, there are displayed samples of 512×512 and 256×256 resolution ''full'' typology and further on the right, from top to bottom, 256×256 ''upper'' and ''lower'' soldering region typology.
