<a target="_blank" rel="ugc" href="https://ko-fi.com/citronlegacy">Help fuel my passion! Even $1 makes a difference. ☕ https://ko-fi.com/citronlegacy</a>Please checkout all my Pokemon Loras: <a target="_blank" rel="ugc" href="https://civarchive.com/user/CitronLegacy/models?tag=pokemon">https://civarchive.com/user/CitronLegacy/models?tag=pokemon</a><a target="_blank" rel="ugc" href="https://civarchive.com/tag/fairy%20pokemon">#Fairy Pokemon✨</a>This Lora can make Togepi and also a person holding Togepi. Misty is trained into the Lora can be prompted with <code>tgpiMsty</code> other girls can be holding Togepi by prompting <code>holding tgpi</code>

Togepi: Spike Ball | トゲピー (Pokemon)  (Pokedex #0175) [Pony & SD1.5]

Trained on the CivitAI trainerSame dataset as the SD1.5 version.Adding some important training notes:I tried making this with 4 repeats and 2 batch (cost 688 buzz) but it failed.The version that worked was made with with 6 repeats and 2 batch (cost 906 buzz) Full list of settingsunetLR: 0.0005clipSkip: 1loraType: lorakeepTokens: 1networkDim: 32numRepeats: 6resolution: 1024lrScheduler: cosineminSnrGamma: 0noiseOffset: 0.03targetSteps: 2190enableBucket: truenetworkAlpha: 32optimizerType: ProdigytextEncoderLR: 0.00005maxTrainEpochs: 10shuffleCaption: truetrainBatchSize: 2flipAugmentation: falselrSchedulerNumCycles: 3

TogepiPokmonPDXL_v2.safetensors

Pony v2

v1.0

<a target="_blank" rel="ugc" href="https://ko-fi.com/citronlegacy">Help fuel my passion! Even $1 makes a difference. ☕ https://ko-fi.com/citronlegacy</a>Please checkout all my Pokemon Loras: <a target="_blank" rel="ugc" href="https://civitai.com/user/CitronLegacy/models?tag=pokemon">https://civitai.com/user/CitronLegacy/models?tag=pokemon</a><a target="_blank" rel="ugc" href="https://civitai.com/tag/fairy%20pokemon">#Fairy Pokemon✨</a>This Lora can make Togepi and also a person holding Togepi. Misty is trained into the Lora can be prompted with <code>tgpiMsty</code> other girls can be holding Togepi by prompting <code>holding tgpi</code>

&lt;p&gt;Trained on the CivitAI trainer&lt;/p&gt;&lt;p&gt;Same dataset as the SD1.5 version.&lt;/p&gt;&lt;p&gt;Adding some important training notes:&lt;/p&gt;&lt;p&gt;I tried making this with 4 repeats and 2 batch (cost 688 buzz) but it failed.&lt;/p&gt;&lt;p&gt;The version that worked was made with with 6 repeats and 2 batch (cost 906 buzz) &lt;/p&gt;&lt;p&gt;Full list of settings&lt;/p&gt;&lt;p&gt;unetLR: 0.0005&lt;/p&gt;&lt;p&gt;clipSkip: 1&lt;/p&gt;&lt;p&gt;loraType: lora&lt;/p&gt;&lt;p&gt;keepTokens: 1&lt;/p&gt;&lt;p&gt;networkDim: 32&lt;/p&gt;&lt;p&gt;numRepeats: 6&lt;/p&gt;&lt;p&gt;resolution: 1024&lt;/p&gt;&lt;p&gt;lrScheduler: cosine&lt;/p&gt;&lt;p&gt;minSnrGamma: 0&lt;/p&gt;&lt;p&gt;noiseOffset: 0.03&lt;/p&gt;&lt;p&gt;targetSteps: 2190&lt;/p&gt;&lt;p&gt;enableBucket: true&lt;/p&gt;&lt;p&gt;networkAlpha: 32&lt;/p&gt;&lt;p&gt;optimizerType: Prodigy&lt;/p&gt;&lt;p&gt;textEncoderLR: 0.00005&lt;/p&gt;&lt;p&gt;maxTrainEpochs: 10&lt;/p&gt;&lt;p&gt;shuffleCaption: true&lt;/p&gt;&lt;p&gt;trainBatchSize: 2&lt;/p&gt;&lt;p&gt;flipAugmentation: false&lt;/p&gt;&lt;p&gt;lrSchedulerNumCycles: 3&lt;/p&gt;

Files with SHA256: 48ccf0cd28d87eada67a78196e4c0b7655c6da98261d05b65257e62ca8b4130e