<h2 id="preface"><span style="color:rgb(76, 110, 245)"><strong>Preface</strong></span></h2><p>I used the first version of these models in version v1.0 for learning and testing purposes. In parallel I wrote an article how a panama hat can be modified by a LoRA model [1].</p><h2 id="approach"><span style="color:rgb(76, 110, 245)"><strong>Approach</strong></span></h2><p>So far I am using <em>AUTOMATIC1111</em> and the <em>Extension</em> <em>TrainTrain</em> for the training of the LoRA models. Other approaches for a LoRA training did not work or did not work as expected on my local installation.</p><h2 id="what-the-model-does"><span style="color:rgb(76, 110, 245)"><strong>What the Model Does</strong></span></h2><p>The LoRA model creates a panama hat, when the weight is set to 1.0.</p><p>If the weight is less than 1.0 then the LoRA model can change a panama hat in a given image.</p><p>The model is not working at the moment as I would expect it.</p><h2 id="trigger-word"><span style="color:rgb(76, 110, 245)"><strong>Trigger Word</strong></span></h2><p>The required trigger word is</p><pre><code>panama hat</code></pre><p>In version 1.3 I changed the trigger word to</p><pre><code>panamahat</code></pre><h2 id="versions"><span style="color:rgb(76, 110, 245)"><strong>Versions</strong></span></h2><p>The version which is not working so well is version v1.0 The other following versions can be tried in parallel. It feels like they produce better results.</p><h3 id="version-v1.3"><span style="color:rgb(250, 82, 82)"><strong>Version v1.3</strong></span></h3><p>Maximal training steps were 5000. I stored versions in steps of 500. Trying version by version the version with 1000 training steps resulted in the best results.</p><h2 id="test-case"><span style="color:rgb(76, 110, 245)"><strong>Test Case</strong></span></h2><p>Using following Prompts are good testing cases for the LoRA model:</p><pre><code> panama hat, &lt;lora:PanamaHatConcept_v13:[1.0...1.4]&gt;
 panamahat, &lt;lora:PanamaHatConcept_v13:[1.0...1.4]&gt;</code></pre><p>A weight of 1.0 up to 1.4 is working. With a value of 1.5 overfitting takes place.</p><p>One should get a panama hat as in the following image:</p><h2 id="training-images"><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/37672166-596a-4838-924c-25508db5c855/width=525/37672166-596a-4838-924c-25508db5c855.jpeg" /><span style="color:rgb(76, 110, 245)"><strong>Training Images</strong></span></h2><p>I started the training with 20 images and I used finally 36 images. The used training images shows a panama from different viewing angles and positions. The background of all images is bright and single-coloured.</p><h2 id="limitation"><span style="color:rgb(76, 110, 245)"><strong>Limitation</strong></span></h2><p>You have to bear in mind what a LoRA model does when it comes to limits. If the LoRA model is not used and the base model is not knowing what a panama hat ist, then it is not possible to change a not existing panama hat.</p><h2 id="final-words"><span style="color:rgb(76, 110, 245)"><strong>Final Words</strong></span></h2><p>Have a nice day. Happy model testing. Have fun. Be inspired!</p><h2 id="reference"><span style="color:rgb(76, 110, 245)"><strong>Reference</strong></span></h2><p>[1] <a target="_blank" rel="ugc" href="https://civarchive.com/articles/7481/using-lora-to-change-details-in-a-photo">https://civarchive.com/articles/7481/using-lora-to-change-details-in-a-photo</a></p><p></p><h1 id="donation"><strong>Donation</strong></h1><p>About each Buzz I am happy! Yellow<span style="color:rgb(148, 148, 148)"> </span>ones<span style="color:rgb(148, 148, 148)"> </span>are<span style="color:rgb(148, 148, 148)"> </span>very<span style="color:rgb(148, 148, 148)"> </span>welcome!</p><p>I am also happy to accept donations in one of 250+ crypto currencies:</p><p><a target="_blank" rel="ugc" href="https://nowpayments.io/donation/zentrocdot">https://nowpayments.io/donation/zentrocdot</a></p><p>Leave optional a comment like improve documentation or improve model etc. when doing so, that I know in what you are interested in.</p>

PanamaHat

Files with SHA256: 091c54ce608753829b22aa0a431e51916bc7b084b79d68295d2e6b9657fea7eb