KOD+NN | for Illustious v0.1 | - dt=v5r chk=ill_0.1 adf

KOD+NN | for Illustious v0.1 |

LORA
Original


Updated:

dt=v5r chk=ill_0.1 adf - third attempt 4/10, better prompt adherence.

dt=v5r chk=ill_0.1 adam - second attempt 3/10, bad promp adherence, but closer to Nomo style.

v0.1 ep15 - first attempt 2.5/10, very poor prompt adherence.

Training settings in model version tab.

Version Detail

Illustrious
<pre><code>{ "engine": "kohya", "unetLR": 0.0005, "clipSkip": 1, "loraType": "lora", "keepTokens": 0, "networkDim": 32, "numRepeats": 1, "resolution": 1024, "lrScheduler": "cosine_with_restarts", "minSnrGamma": 5, "noiseOffset": 0.1, "targetSteps": 2268, "enableBucket": true, "networkAlpha": 32, "optimizerType": "Adafactor", "textEncoderLR": 0.00005, "maxTrainEpochs": 14, "shuffleCaption": false, "trainBatchSize": 3, "flipAugmentation": false, "lrSchedulerNumCycles": 3 }</code></pre>

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts