🐣 Please follow me for new updates <a target="_blank" rel="ugc" href="https://twitter.com/camenduru￼🔥">https://twitter.com/camenduru 🔥</a> Please join our discord server <a target="_blank" rel="ugc" href="https://discord.gg/k5BwmmvJJU￼🥳">https://discord.gg/k5BwmmvJJU 🥳</a> Please join my patreon community <a target="_blank" rel="ugc" href="https://patreon.com/camenduru">https://patreon.com/camenduru</a>Model Colab: <a target="_blank" rel="ugc" href="https://github.com/camenduru/ios-emoji-xl-model-colab">https://github.com/camenduru/ios-emoji-xl-model-colab</a> Model Mirror: <a target="_blank" rel="ugc" href="https://huggingface.co/camenduru/ios-emoji-xl/blob/main/ios_emoji_xl_v2_lora_webui.safetensors">https://huggingface.co/camenduru/ios-emoji-xl/blob/main/ios_emoji_xl_v2_lora_webui.safetensors</a> Model Dataset: <a target="_blank" rel="ugc" href="https://huggingface.co/camenduru/ios-emoji-xl/blob/main/dataset_160x160_images.zip">https://huggingface.co/camenduru/ios-emoji-xl/blob/main/dataset_160x160_images.zip</a> Thanks to <a target="_blank" rel="ugc" href="https://github.com/samuelngs/apple-emoji-linux">https://github.com/samuelngs/apple-emoji-linux</a> for the 160x160 pixel emojis ❤ Thanks to <a target="_blank" rel="ugc" href="https://replicate.com">https://replicate.com</a> ❤ Training Logs:<pre><code>Trained with 160x160 pixel ios v16.4 emojis 😋

GPU = Nvidia A40 (Large) at https://replicate.com
Num examples = 4129
Num batches each epoch = 1033
Num Epochs = 1
Instantaneous batch size per device = 4
Total train batch size (w. parallel, distributed &amp; accumulation) = 4
Gradient Accumulation steps = 1
Total optimization steps = 1000
Total Run time: 45.46 minutes
Total Cost: $1.98</code></pre>Replicate LoRA to WebUI LoRA Converter<ul><li>Thanks to</li><li><a target="_blank" rel="ugc" href="https://github.com/ignacfetser">https://github.com/ignacfetser</a> ❤</li><li><a target="_blank" rel="ugc" href="https://github.com/huggingface/diffusers/issues/2326">huggingface/diffusers#2326</a> ❤</li><li><a target="_blank" rel="ugc" href="https://github.com/harrywang/finetune-sd/blob/main/convert-to-safetensors.py">https://github.com/harrywang/finetune-sd/blob/main/convert-to-safetensors.py</a> ❤</li></ul><pre><code>pip install safetensors==0.3.3

import re
from safetensors.torch import load_file, save_file
checkpoint = load_file('/content/ui/models/Lora/ios_emoji_xl_v2_lora.safetensors')
new_dict = dict()
for idx, key in enumerate(checkpoint):
 new_key = re.sub('\.processor\.', '_', key)
 new_key = re.sub('mid_block\.', 'mid_block_', new_key)
 new_key = re.sub('_lora.up.', '.lora_up.', new_key)
 new_key = re.sub('_lora.down.', '.lora_down.', new_key)
 new_key = re.sub('\.(\d+)\.', '_\\1_', new_key)
 new_key = re.sub('to_out', 'to_out_0', new_key)
 new_key = 'lora_unet_' + new_key
 new_dict[new_key] = checkpoint[key]
save_file(new_dict, 'ios_emoji_xl_v2_lora_webui.safetensors')</code></pre>Thanks to fofr ❤ for the idea. fofr's model: <a target="_blank" rel="ugc" href="https://twitter.com/fofrAI/status/1698741974835065171">https://twitter.com/fofrAI/status/1698741974835065171</a>

Files with SHA256: 62a79950631ae544f7fb9c93a0fb0368434db5046df9ef98e8e5f5c35c64ad51