ostris/ai-toolkit
1
0
Fork 0
mirror of https://github.com/ostris/ai-toolkit.git synced 2026-04-27 09:44:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fixed some mismatched weights by adjusting tolerance. The mismatch ironically made the models better lol

Browse Source
This commit is contained in:
Jaret Burkett
2023-08-29 15:20:03 -06:00
parent 14ff51ceb4
commit 836fee47a6
8 changed files with 265 additions and 163 deletions
Download Patch File Download Diff File Expand all files Collapse all files

116
toolkit/keymaps/stable_diffusion_sd1.json
View File

@@ -58,7 +58,7 @@
"cond_stage_model.transformer.text_model.encoder.layers.11.mlp.fc1.weight": "te_text_model.encoder.layers.11.mlp.fc1.weight",
"cond_stage_model.transformer.text_model.encoder.layers.11.mlp.fc2.bias": "te_text_model.encoder.layers.11.mlp.fc2.bias",
"cond_stage_model.transformer.text_model.encoder.layers.11.mlp.fc2.weight": "te_text_model.encoder.layers.11.mlp.fc2.weight",
"cond_stage_model.transformer.text_model.encoder.layers.11.self_attn.k_proj.bias": "te_text_model.encoder.layers.2.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.11.self_attn.k_proj.bias": "te_text_model.encoder.layers.11.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.11.self_attn.k_proj.weight": "te_text_model.encoder.layers.11.self_attn.k_proj.weight",
"cond_stage_model.transformer.text_model.encoder.layers.11.self_attn.out_proj.bias": "te_text_model.encoder.layers.11.self_attn.out_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.11.self_attn.out_proj.weight": "te_text_model.encoder.layers.11.self_attn.out_proj.weight",
@@ -74,7 +74,7 @@
"cond_stage_model.transformer.text_model.encoder.layers.2.mlp.fc1.weight": "te_text_model.encoder.layers.2.mlp.fc1.weight",
"cond_stage_model.transformer.text_model.encoder.layers.2.mlp.fc2.bias": "te_text_model.encoder.layers.2.mlp.fc2.bias",
"cond_stage_model.transformer.text_model.encoder.layers.2.mlp.fc2.weight": "te_text_model.encoder.layers.2.mlp.fc2.weight",
"cond_stage_model.transformer.text_model.encoder.layers.2.self_attn.k_proj.bias": "te_text_model.encoder.layers.3.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.2.self_attn.k_proj.bias": "te_text_model.encoder.layers.2.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.2.self_attn.k_proj.weight": "te_text_model.encoder.layers.2.self_attn.k_proj.weight",
"cond_stage_model.transformer.text_model.encoder.layers.2.self_attn.out_proj.bias": "te_text_model.encoder.layers.2.self_attn.out_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.2.self_attn.out_proj.weight": "te_text_model.encoder.layers.2.self_attn.out_proj.weight",
@@ -90,7 +90,7 @@
"cond_stage_model.transformer.text_model.encoder.layers.3.mlp.fc1.weight": "te_text_model.encoder.layers.3.mlp.fc1.weight",
"cond_stage_model.transformer.text_model.encoder.layers.3.mlp.fc2.bias": "te_text_model.encoder.layers.3.mlp.fc2.bias",
"cond_stage_model.transformer.text_model.encoder.layers.3.mlp.fc2.weight": "te_text_model.encoder.layers.3.mlp.fc2.weight",
"cond_stage_model.transformer.text_model.encoder.layers.3.self_attn.k_proj.bias": "te_text_model.encoder.layers.11.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.3.self_attn.k_proj.bias": "te_text_model.encoder.layers.3.self_attn.k_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.3.self_attn.k_proj.weight": "te_text_model.encoder.layers.3.self_attn.k_proj.weight",
"cond_stage_model.transformer.text_model.encoder.layers.3.self_attn.out_proj.bias": "te_text_model.encoder.layers.3.self_attn.out_proj.bias",
"cond_stage_model.transformer.text_model.encoder.layers.3.self_attn.out_proj.weight": "te_text_model.encoder.layers.3.self_attn.out_proj.weight",
@@ -446,11 +446,11 @@
"first_stage_model.quant_conv.weight": "vae_quant_conv.weight",
"model.diffusion_model.input_blocks.0.0.bias": "unet_conv_in.bias",
"model.diffusion_model.input_blocks.0.0.weight": "unet_conv_in.weight",
"model.diffusion_model.input_blocks.1.0.emb_layers.1.bias": "unet_down_blocks.0.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.1.0.emb_layers.1.bias": "unet_down_blocks.0.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.1.0.emb_layers.1.weight": "unet_down_blocks.0.resnets.0.time_emb_proj.weight",
"model.diffusion_model.input_blocks.1.0.in_layers.0.bias": "unet_down_blocks.0.resnets.0.norm1.bias",
"model.diffusion_model.input_blocks.1.0.in_layers.0.weight": "unet_down_blocks.0.resnets.0.norm1.weight",
"model.diffusion_model.input_blocks.1.0.in_layers.2.bias": "unet_down_blocks.0.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.1.0.in_layers.2.bias": "unet_down_blocks.0.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.1.0.in_layers.2.weight": "unet_down_blocks.0.resnets.0.conv1.weight",
"model.diffusion_model.input_blocks.1.0.out_layers.0.bias": "unet_down_blocks.0.resnets.0.norm2.bias",
"model.diffusion_model.input_blocks.1.0.out_layers.0.weight": "unet_down_blocks.0.resnets.0.norm2.weight",
@@ -482,31 +482,31 @@
"model.diffusion_model.input_blocks.1.1.transformer_blocks.0.norm2.weight": "unet_down_blocks.0.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.input_blocks.1.1.transformer_blocks.0.norm3.bias": "unet_down_blocks.0.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.input_blocks.1.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.0.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.10.0.emb_layers.1.bias": "unet_down_blocks.3.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.10.0.emb_layers.1.weight": "unet_time_embedding.linear_2.weight",
"model.diffusion_model.input_blocks.10.0.emb_layers.1.bias": "unet_down_blocks.3.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.10.0.emb_layers.1.weight": "unet_down_blocks.3.resnets.0.time_emb_proj.weight",
"model.diffusion_model.input_blocks.10.0.in_layers.0.bias": "unet_down_blocks.3.resnets.0.norm1.bias",
"model.diffusion_model.input_blocks.10.0.in_layers.0.weight": "unet_down_blocks.3.resnets.0.norm1.weight",
"model.diffusion_model.input_blocks.10.0.in_layers.2.bias": "unet_down_blocks.3.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.10.0.in_layers.2.bias": "unet_down_blocks.3.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.10.0.in_layers.2.weight": "unet_down_blocks.3.resnets.0.conv1.weight",
"model.diffusion_model.input_blocks.10.0.out_layers.0.bias": "unet_down_blocks.3.resnets.0.norm2.bias",
"model.diffusion_model.input_blocks.10.0.out_layers.0.weight": "unet_down_blocks.3.resnets.0.norm2.weight",
"model.diffusion_model.input_blocks.10.0.out_layers.3.bias": "unet_down_blocks.3.resnets.0.conv2.bias",
"model.diffusion_model.input_blocks.10.0.out_layers.3.weight": "unet_down_blocks.3.resnets.0.conv2.weight",
"model.diffusion_model.input_blocks.11.0.emb_layers.1.bias": "unet_down_blocks.3.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.11.0.emb_layers.1.weight": "unet_down_blocks.3.resnets.0.time_emb_proj.weight",
"model.diffusion_model.input_blocks.11.0.emb_layers.1.bias": "unet_down_blocks.3.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.11.0.emb_layers.1.weight": "unet_down_blocks.3.resnets.1.time_emb_proj.weight",
"model.diffusion_model.input_blocks.11.0.in_layers.0.bias": "unet_down_blocks.3.resnets.1.norm1.bias",
"model.diffusion_model.input_blocks.11.0.in_layers.0.weight": "unet_down_blocks.3.resnets.1.norm1.weight",
"model.diffusion_model.input_blocks.11.0.in_layers.2.bias": "unet_down_blocks.3.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.11.0.in_layers.2.bias": "unet_down_blocks.3.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.11.0.in_layers.2.weight": "unet_down_blocks.3.resnets.1.conv1.weight",
"model.diffusion_model.input_blocks.11.0.out_layers.0.bias": "unet_down_blocks.3.resnets.1.norm2.bias",
"model.diffusion_model.input_blocks.11.0.out_layers.0.weight": "unet_down_blocks.3.resnets.1.norm2.weight",
"model.diffusion_model.input_blocks.11.0.out_layers.3.bias": "unet_down_blocks.3.resnets.1.conv2.bias",
"model.diffusion_model.input_blocks.11.0.out_layers.3.weight": "unet_down_blocks.3.resnets.1.conv2.weight",
"model.diffusion_model.input_blocks.2.0.emb_layers.1.bias": "unet_down_blocks.0.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.2.0.emb_layers.1.bias": "unet_down_blocks.0.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.2.0.emb_layers.1.weight": "unet_down_blocks.0.resnets.1.time_emb_proj.weight",
"model.diffusion_model.input_blocks.2.0.in_layers.0.bias": "unet_down_blocks.0.resnets.1.norm1.bias",
"model.diffusion_model.input_blocks.2.0.in_layers.0.weight": "unet_down_blocks.0.resnets.1.norm1.weight",
"model.diffusion_model.input_blocks.2.0.in_layers.2.bias": "unet_down_blocks.0.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.2.0.in_layers.2.bias": "unet_down_blocks.0.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.2.0.in_layers.2.weight": "unet_down_blocks.0.resnets.1.conv1.weight",
"model.diffusion_model.input_blocks.2.0.out_layers.0.bias": "unet_down_blocks.0.resnets.1.norm2.bias",
"model.diffusion_model.input_blocks.2.0.out_layers.0.weight": "unet_down_blocks.0.resnets.1.norm2.weight",
@@ -540,11 +540,11 @@
"model.diffusion_model.input_blocks.2.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.0.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.3.0.op.bias": "unet_down_blocks.0.downsamplers.0.conv.bias",
"model.diffusion_model.input_blocks.3.0.op.weight": "unet_down_blocks.0.downsamplers.0.conv.weight",
"model.diffusion_model.input_blocks.4.0.emb_layers.1.bias": "unet_down_blocks.1.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.4.0.emb_layers.1.bias": "unet_down_blocks.1.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.4.0.emb_layers.1.weight": "unet_down_blocks.1.resnets.0.time_emb_proj.weight",
"model.diffusion_model.input_blocks.4.0.in_layers.0.bias": "unet_down_blocks.1.resnets.0.norm1.bias",
"model.diffusion_model.input_blocks.4.0.in_layers.0.weight": "unet_down_blocks.1.resnets.0.norm1.weight",
"model.diffusion_model.input_blocks.4.0.in_layers.2.bias": "unet_down_blocks.1.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.4.0.in_layers.2.bias": "unet_down_blocks.1.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.4.0.in_layers.2.weight": "unet_down_blocks.1.resnets.0.conv1.weight",
"model.diffusion_model.input_blocks.4.0.out_layers.0.bias": "unet_down_blocks.1.resnets.0.norm2.bias",
"model.diffusion_model.input_blocks.4.0.out_layers.0.weight": "unet_down_blocks.1.resnets.0.norm2.weight",
@@ -578,11 +578,11 @@
"model.diffusion_model.input_blocks.4.1.transformer_blocks.0.norm2.weight": "unet_down_blocks.1.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.input_blocks.4.1.transformer_blocks.0.norm3.bias": "unet_down_blocks.1.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.input_blocks.4.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.1.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.5.0.emb_layers.1.bias": "unet_down_blocks.1.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.5.0.emb_layers.1.bias": "unet_down_blocks.1.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.5.0.emb_layers.1.weight": "unet_down_blocks.1.resnets.1.time_emb_proj.weight",
"model.diffusion_model.input_blocks.5.0.in_layers.0.bias": "unet_down_blocks.1.resnets.1.norm1.bias",
"model.diffusion_model.input_blocks.5.0.in_layers.0.weight": "unet_down_blocks.1.resnets.1.norm1.weight",
"model.diffusion_model.input_blocks.5.0.in_layers.2.bias": "unet_down_blocks.1.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.5.0.in_layers.2.bias": "unet_down_blocks.1.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.5.0.in_layers.2.weight": "unet_down_blocks.1.resnets.1.conv1.weight",
"model.diffusion_model.input_blocks.5.0.out_layers.0.bias": "unet_down_blocks.1.resnets.1.norm2.bias",
"model.diffusion_model.input_blocks.5.0.out_layers.0.weight": "unet_down_blocks.1.resnets.1.norm2.weight",
@@ -616,11 +616,11 @@
"model.diffusion_model.input_blocks.5.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.1.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.6.0.op.bias": "unet_down_blocks.1.downsamplers.0.conv.bias",
"model.diffusion_model.input_blocks.6.0.op.weight": "unet_down_blocks.1.downsamplers.0.conv.weight",
"model.diffusion_model.input_blocks.7.0.emb_layers.1.bias": "unet_down_blocks.2.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.7.0.emb_layers.1.bias": "unet_down_blocks.2.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.7.0.emb_layers.1.weight": "unet_down_blocks.2.resnets.0.time_emb_proj.weight",
"model.diffusion_model.input_blocks.7.0.in_layers.0.bias": "unet_down_blocks.2.resnets.0.norm1.bias",
"model.diffusion_model.input_blocks.7.0.in_layers.0.weight": "unet_down_blocks.2.resnets.0.norm1.weight",
"model.diffusion_model.input_blocks.7.0.in_layers.2.bias": "unet_down_blocks.2.resnets.0.time_emb_proj.bias",
"model.diffusion_model.input_blocks.7.0.in_layers.2.bias": "unet_down_blocks.2.resnets.0.conv1.bias",
"model.diffusion_model.input_blocks.7.0.in_layers.2.weight": "unet_down_blocks.2.resnets.0.conv1.weight",
"model.diffusion_model.input_blocks.7.0.out_layers.0.bias": "unet_down_blocks.2.resnets.0.norm2.bias",
"model.diffusion_model.input_blocks.7.0.out_layers.0.weight": "unet_down_blocks.2.resnets.0.norm2.weight",
@@ -654,11 +654,11 @@
"model.diffusion_model.input_blocks.7.1.transformer_blocks.0.norm2.weight": "unet_down_blocks.2.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.input_blocks.7.1.transformer_blocks.0.norm3.bias": "unet_down_blocks.2.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.input_blocks.7.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.2.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.8.0.emb_layers.1.bias": "unet_down_blocks.2.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.8.0.emb_layers.1.bias": "unet_down_blocks.2.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.8.0.emb_layers.1.weight": "unet_down_blocks.2.resnets.1.time_emb_proj.weight",
"model.diffusion_model.input_blocks.8.0.in_layers.0.bias": "unet_down_blocks.2.resnets.1.norm1.bias",
"model.diffusion_model.input_blocks.8.0.in_layers.0.weight": "unet_down_blocks.2.resnets.1.norm1.weight",
"model.diffusion_model.input_blocks.8.0.in_layers.2.bias": "unet_down_blocks.2.resnets.1.time_emb_proj.bias",
"model.diffusion_model.input_blocks.8.0.in_layers.2.bias": "unet_down_blocks.2.resnets.1.conv1.bias",
"model.diffusion_model.input_blocks.8.0.in_layers.2.weight": "unet_down_blocks.2.resnets.1.conv1.weight",
"model.diffusion_model.input_blocks.8.0.out_layers.0.bias": "unet_down_blocks.2.resnets.1.norm2.bias",
"model.diffusion_model.input_blocks.8.0.out_layers.0.weight": "unet_down_blocks.2.resnets.1.norm2.weight",
@@ -692,11 +692,11 @@
"model.diffusion_model.input_blocks.8.1.transformer_blocks.0.norm3.weight": "unet_down_blocks.2.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.input_blocks.9.0.op.bias": "unet_down_blocks.2.downsamplers.0.conv.bias",
"model.diffusion_model.input_blocks.9.0.op.weight": "unet_down_blocks.2.downsamplers.0.conv.weight",
"model.diffusion_model.middle_block.0.emb_layers.1.bias": "unet_mid_block.resnets.0.conv1.bias",
"model.diffusion_model.middle_block.0.emb_layers.1.weight": "unet_down_blocks.3.resnets.1.time_emb_proj.weight",
"model.diffusion_model.middle_block.0.emb_layers.1.bias": "unet_mid_block.resnets.0.time_emb_proj.bias",
"model.diffusion_model.middle_block.0.emb_layers.1.weight": "unet_mid_block.resnets.0.time_emb_proj.weight",
"model.diffusion_model.middle_block.0.in_layers.0.bias": "unet_mid_block.resnets.0.norm1.bias",
"model.diffusion_model.middle_block.0.in_layers.0.weight": "unet_mid_block.resnets.0.norm1.weight",
"model.diffusion_model.middle_block.0.in_layers.2.bias": "unet_mid_block.resnets.0.time_emb_proj.bias",
"model.diffusion_model.middle_block.0.in_layers.2.bias": "unet_mid_block.resnets.0.conv1.bias",
"model.diffusion_model.middle_block.0.in_layers.2.weight": "unet_mid_block.resnets.0.conv1.weight",
"model.diffusion_model.middle_block.0.out_layers.0.bias": "unet_mid_block.resnets.0.norm2.bias",
"model.diffusion_model.middle_block.0.out_layers.0.weight": "unet_mid_block.resnets.0.norm2.weight",
@@ -728,11 +728,11 @@
"model.diffusion_model.middle_block.1.transformer_blocks.0.norm2.weight": "unet_mid_block.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.middle_block.1.transformer_blocks.0.norm3.bias": "unet_mid_block.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.middle_block.1.transformer_blocks.0.norm3.weight": "unet_mid_block.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.middle_block.2.emb_layers.1.bias": "unet_mid_block.resnets.1.conv1.bias",
"model.diffusion_model.middle_block.2.emb_layers.1.weight": "unet_up_blocks.0.resnets.0.time_emb_proj.weight",
"model.diffusion_model.middle_block.2.emb_layers.1.bias": "unet_mid_block.resnets.1.time_emb_proj.bias",
"model.diffusion_model.middle_block.2.emb_layers.1.weight": "unet_mid_block.resnets.1.time_emb_proj.weight",
"model.diffusion_model.middle_block.2.in_layers.0.bias": "unet_mid_block.resnets.1.norm1.bias",
"model.diffusion_model.middle_block.2.in_layers.0.weight": "unet_mid_block.resnets.1.norm1.weight",
"model.diffusion_model.middle_block.2.in_layers.2.bias": "unet_mid_block.resnets.1.time_emb_proj.bias",
"model.diffusion_model.middle_block.2.in_layers.2.bias": "unet_mid_block.resnets.1.conv1.bias",
"model.diffusion_model.middle_block.2.in_layers.2.weight": "unet_mid_block.resnets.1.conv1.weight",
"model.diffusion_model.middle_block.2.out_layers.0.bias": "unet_mid_block.resnets.1.norm2.bias",
"model.diffusion_model.middle_block.2.out_layers.0.weight": "unet_mid_block.resnets.1.norm2.weight",
@@ -742,11 +742,11 @@
"model.diffusion_model.out.0.weight": "unet_conv_norm_out.weight",
"model.diffusion_model.out.2.bias": "unet_conv_out.bias",
"model.diffusion_model.out.2.weight": "unet_conv_out.weight",
"model.diffusion_model.output_blocks.0.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.0.0.emb_layers.1.weight": "unet_up_blocks.0.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.0.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.0.0.emb_layers.1.weight": "unet_up_blocks.0.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.0.0.in_layers.0.bias": "unet_up_blocks.0.resnets.0.norm1.bias",
"model.diffusion_model.output_blocks.0.0.in_layers.0.weight": "unet_up_blocks.0.resnets.0.norm1.weight",
"model.diffusion_model.output_blocks.0.0.in_layers.2.bias": "unet_up_blocks.0.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.0.0.in_layers.2.bias": "unet_up_blocks.0.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.0.0.in_layers.2.weight": "unet_up_blocks.0.resnets.0.conv1.weight",
"model.diffusion_model.output_blocks.0.0.out_layers.0.bias": "unet_up_blocks.0.resnets.0.norm2.bias",
"model.diffusion_model.output_blocks.0.0.out_layers.0.weight": "unet_up_blocks.0.resnets.0.norm2.weight",
@@ -754,11 +754,11 @@
"model.diffusion_model.output_blocks.0.0.out_layers.3.weight": "unet_up_blocks.0.resnets.0.conv2.weight",
"model.diffusion_model.output_blocks.0.0.skip_connection.bias": "unet_up_blocks.0.resnets.0.conv_shortcut.bias",
"model.diffusion_model.output_blocks.0.0.skip_connection.weight": "unet_up_blocks.0.resnets.0.conv_shortcut.weight",
"model.diffusion_model.output_blocks.1.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.1.0.emb_layers.1.weight": "unet_up_blocks.0.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.1.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.1.0.emb_layers.1.weight": "unet_up_blocks.0.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.1.0.in_layers.0.bias": "unet_up_blocks.0.resnets.1.norm1.bias",
"model.diffusion_model.output_blocks.1.0.in_layers.0.weight": "unet_up_blocks.0.resnets.1.norm1.weight",
"model.diffusion_model.output_blocks.1.0.in_layers.2.bias": "unet_up_blocks.0.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.1.0.in_layers.2.bias": "unet_up_blocks.0.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.1.0.in_layers.2.weight": "unet_up_blocks.0.resnets.1.conv1.weight",
"model.diffusion_model.output_blocks.1.0.out_layers.0.bias": "unet_up_blocks.0.resnets.1.norm2.bias",
"model.diffusion_model.output_blocks.1.0.out_layers.0.weight": "unet_up_blocks.0.resnets.1.norm2.weight",
@@ -766,11 +766,11 @@
"model.diffusion_model.output_blocks.1.0.out_layers.3.weight": "unet_up_blocks.0.resnets.1.conv2.weight",
"model.diffusion_model.output_blocks.1.0.skip_connection.bias": "unet_up_blocks.0.resnets.1.conv_shortcut.bias",
"model.diffusion_model.output_blocks.1.0.skip_connection.weight": "unet_up_blocks.0.resnets.1.conv_shortcut.weight",
"model.diffusion_model.output_blocks.10.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.10.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.10.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.10.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.10.0.in_layers.0.bias": "unet_up_blocks.3.resnets.1.norm1.bias",
"model.diffusion_model.output_blocks.10.0.in_layers.0.weight": "unet_up_blocks.3.resnets.1.norm1.weight",
"model.diffusion_model.output_blocks.10.0.in_layers.2.bias": "unet_up_blocks.3.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.10.0.in_layers.2.bias": "unet_up_blocks.3.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.10.0.in_layers.2.weight": "unet_up_blocks.3.resnets.1.conv1.weight",
"model.diffusion_model.output_blocks.10.0.out_layers.0.bias": "unet_up_blocks.3.resnets.1.norm2.bias",
"model.diffusion_model.output_blocks.10.0.out_layers.0.weight": "unet_up_blocks.3.resnets.1.norm2.weight",
@@ -804,11 +804,11 @@
"model.diffusion_model.output_blocks.10.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.3.attentions.1.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.10.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.3.attentions.1.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.10.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.3.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.11.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.11.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.11.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.11.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.11.0.in_layers.0.bias": "unet_up_blocks.3.resnets.2.norm1.bias",
"model.diffusion_model.output_blocks.11.0.in_layers.0.weight": "unet_up_blocks.3.resnets.2.norm1.weight",
"model.diffusion_model.output_blocks.11.0.in_layers.2.bias": "unet_up_blocks.3.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.11.0.in_layers.2.bias": "unet_up_blocks.3.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.11.0.in_layers.2.weight": "unet_up_blocks.3.resnets.2.conv1.weight",
"model.diffusion_model.output_blocks.11.0.out_layers.0.bias": "unet_up_blocks.3.resnets.2.norm2.bias",
"model.diffusion_model.output_blocks.11.0.out_layers.0.weight": "unet_up_blocks.3.resnets.2.norm2.weight",
@@ -842,11 +842,11 @@
"model.diffusion_model.output_blocks.11.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.3.attentions.2.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.11.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.3.attentions.2.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.11.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.3.attentions.2.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.2.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.2.0.emb_layers.1.weight": "unet_mid_block.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.2.0.emb_layers.1.bias": "unet_up_blocks.0.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.2.0.emb_layers.1.weight": "unet_up_blocks.0.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.2.0.in_layers.0.bias": "unet_up_blocks.0.resnets.2.norm1.bias",
"model.diffusion_model.output_blocks.2.0.in_layers.0.weight": "unet_up_blocks.0.resnets.2.norm1.weight",
"model.diffusion_model.output_blocks.2.0.in_layers.2.bias": "unet_up_blocks.0.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.2.0.in_layers.2.bias": "unet_up_blocks.0.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.2.0.in_layers.2.weight": "unet_up_blocks.0.resnets.2.conv1.weight",
"model.diffusion_model.output_blocks.2.0.out_layers.0.bias": "unet_up_blocks.0.resnets.2.norm2.bias",
"model.diffusion_model.output_blocks.2.0.out_layers.0.weight": "unet_up_blocks.0.resnets.2.norm2.weight",
@@ -856,11 +856,11 @@
"model.diffusion_model.output_blocks.2.0.skip_connection.weight": "unet_up_blocks.0.resnets.2.conv_shortcut.weight",
"model.diffusion_model.output_blocks.2.1.conv.bias": "unet_up_blocks.0.upsamplers.0.conv.bias",
"model.diffusion_model.output_blocks.2.1.conv.weight": "unet_up_blocks.0.upsamplers.0.conv.weight",
"model.diffusion_model.output_blocks.3.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.3.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.3.0.emb_layers.1.weight": "unet_up_blocks.1.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.3.0.in_layers.0.bias": "unet_up_blocks.1.resnets.0.norm1.bias",
"model.diffusion_model.output_blocks.3.0.in_layers.0.weight": "unet_up_blocks.1.resnets.0.norm1.weight",
"model.diffusion_model.output_blocks.3.0.in_layers.2.bias": "unet_up_blocks.1.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.3.0.in_layers.2.bias": "unet_up_blocks.1.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.3.0.in_layers.2.weight": "unet_up_blocks.1.resnets.0.conv1.weight",
"model.diffusion_model.output_blocks.3.0.out_layers.0.bias": "unet_up_blocks.1.resnets.0.norm2.bias",
"model.diffusion_model.output_blocks.3.0.out_layers.0.weight": "unet_up_blocks.1.resnets.0.norm2.weight",
@@ -894,11 +894,11 @@
"model.diffusion_model.output_blocks.3.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.1.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.3.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.1.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.3.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.1.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.4.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.4.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.4.0.emb_layers.1.weight": "unet_up_blocks.1.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.4.0.in_layers.0.bias": "unet_up_blocks.1.resnets.1.norm1.bias",
"model.diffusion_model.output_blocks.4.0.in_layers.0.weight": "unet_up_blocks.1.resnets.1.norm1.weight",
"model.diffusion_model.output_blocks.4.0.in_layers.2.bias": "unet_up_blocks.1.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.4.0.in_layers.2.bias": "unet_up_blocks.1.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.4.0.in_layers.2.weight": "unet_up_blocks.1.resnets.1.conv1.weight",
"model.diffusion_model.output_blocks.4.0.out_layers.0.bias": "unet_up_blocks.1.resnets.1.norm2.bias",
"model.diffusion_model.output_blocks.4.0.out_layers.0.weight": "unet_up_blocks.1.resnets.1.norm2.weight",
@@ -932,11 +932,11 @@
"model.diffusion_model.output_blocks.4.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.1.attentions.1.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.4.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.1.attentions.1.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.4.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.1.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.5.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.5.0.emb_layers.1.bias": "unet_up_blocks.1.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.5.0.emb_layers.1.weight": "unet_up_blocks.1.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.5.0.in_layers.0.bias": "unet_up_blocks.1.resnets.2.norm1.bias",
"model.diffusion_model.output_blocks.5.0.in_layers.0.weight": "unet_up_blocks.1.resnets.2.norm1.weight",
"model.diffusion_model.output_blocks.5.0.in_layers.2.bias": "unet_up_blocks.1.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.5.0.in_layers.2.bias": "unet_up_blocks.1.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.5.0.in_layers.2.weight": "unet_up_blocks.1.resnets.2.conv1.weight",
"model.diffusion_model.output_blocks.5.0.out_layers.0.bias": "unet_up_blocks.1.resnets.2.norm2.bias",
"model.diffusion_model.output_blocks.5.0.out_layers.0.weight": "unet_up_blocks.1.resnets.2.norm2.weight",
@@ -972,11 +972,11 @@
"model.diffusion_model.output_blocks.5.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.1.attentions.2.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.5.2.conv.bias": "unet_up_blocks.1.upsamplers.0.conv.bias",
"model.diffusion_model.output_blocks.5.2.conv.weight": "unet_up_blocks.1.upsamplers.0.conv.weight",
"model.diffusion_model.output_blocks.6.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.6.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.6.0.emb_layers.1.weight": "unet_up_blocks.2.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.6.0.in_layers.0.bias": "unet_up_blocks.2.resnets.0.norm1.bias",
"model.diffusion_model.output_blocks.6.0.in_layers.0.weight": "unet_up_blocks.2.resnets.0.norm1.weight",
"model.diffusion_model.output_blocks.6.0.in_layers.2.bias": "unet_up_blocks.2.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.6.0.in_layers.2.bias": "unet_up_blocks.2.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.6.0.in_layers.2.weight": "unet_up_blocks.2.resnets.0.conv1.weight",
"model.diffusion_model.output_blocks.6.0.out_layers.0.bias": "unet_up_blocks.2.resnets.0.norm2.bias",
"model.diffusion_model.output_blocks.6.0.out_layers.0.weight": "unet_up_blocks.2.resnets.0.norm2.weight",
@@ -1010,11 +1010,11 @@
"model.diffusion_model.output_blocks.6.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.2.attentions.0.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.6.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.2.attentions.0.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.6.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.2.attentions.0.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.7.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.7.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.7.0.emb_layers.1.weight": "unet_up_blocks.2.resnets.1.time_emb_proj.weight",
"model.diffusion_model.output_blocks.7.0.in_layers.0.bias": "unet_up_blocks.2.resnets.1.norm1.bias",
"model.diffusion_model.output_blocks.7.0.in_layers.0.weight": "unet_up_blocks.2.resnets.1.norm1.weight",
"model.diffusion_model.output_blocks.7.0.in_layers.2.bias": "unet_up_blocks.2.resnets.1.time_emb_proj.bias",
"model.diffusion_model.output_blocks.7.0.in_layers.2.bias": "unet_up_blocks.2.resnets.1.conv1.bias",
"model.diffusion_model.output_blocks.7.0.in_layers.2.weight": "unet_up_blocks.2.resnets.1.conv1.weight",
"model.diffusion_model.output_blocks.7.0.out_layers.0.bias": "unet_up_blocks.2.resnets.1.norm2.bias",
"model.diffusion_model.output_blocks.7.0.out_layers.0.weight": "unet_up_blocks.2.resnets.1.norm2.weight",
@@ -1048,11 +1048,11 @@
"model.diffusion_model.output_blocks.7.1.transformer_blocks.0.norm2.weight": "unet_up_blocks.2.attentions.1.transformer_blocks.0.norm2.weight",
"model.diffusion_model.output_blocks.7.1.transformer_blocks.0.norm3.bias": "unet_up_blocks.2.attentions.1.transformer_blocks.0.norm3.bias",
"model.diffusion_model.output_blocks.7.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.2.attentions.1.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.8.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.8.0.emb_layers.1.bias": "unet_up_blocks.2.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.8.0.emb_layers.1.weight": "unet_up_blocks.2.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.8.0.in_layers.0.bias": "unet_up_blocks.2.resnets.2.norm1.bias",
"model.diffusion_model.output_blocks.8.0.in_layers.0.weight": "unet_up_blocks.2.resnets.2.norm1.weight",
"model.diffusion_model.output_blocks.8.0.in_layers.2.bias": "unet_up_blocks.2.resnets.2.time_emb_proj.bias",
"model.diffusion_model.output_blocks.8.0.in_layers.2.bias": "unet_up_blocks.2.resnets.2.conv1.bias",
"model.diffusion_model.output_blocks.8.0.in_layers.2.weight": "unet_up_blocks.2.resnets.2.conv1.weight",
"model.diffusion_model.output_blocks.8.0.out_layers.0.bias": "unet_up_blocks.2.resnets.2.norm2.bias",
"model.diffusion_model.output_blocks.8.0.out_layers.0.weight": "unet_up_blocks.2.resnets.2.norm2.weight",
@@ -1088,11 +1088,11 @@
"model.diffusion_model.output_blocks.8.1.transformer_blocks.0.norm3.weight": "unet_up_blocks.2.attentions.2.transformer_blocks.0.norm3.weight",
"model.diffusion_model.output_blocks.8.2.conv.bias": "unet_up_blocks.2.upsamplers.0.conv.bias",
"model.diffusion_model.output_blocks.8.2.conv.weight": "unet_up_blocks.2.upsamplers.0.conv.weight",
"model.diffusion_model.output_blocks.9.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.9.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.2.time_emb_proj.weight",
"model.diffusion_model.output_blocks.9.0.emb_layers.1.bias": "unet_up_blocks.3.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.9.0.emb_layers.1.weight": "unet_up_blocks.3.resnets.0.time_emb_proj.weight",
"model.diffusion_model.output_blocks.9.0.in_layers.0.bias": "unet_up_blocks.3.resnets.0.norm1.bias",
"model.diffusion_model.output_blocks.9.0.in_layers.0.weight": "unet_up_blocks.3.resnets.0.norm1.weight",
"model.diffusion_model.output_blocks.9.0.in_layers.2.bias": "unet_up_blocks.3.resnets.0.time_emb_proj.bias",
"model.diffusion_model.output_blocks.9.0.in_layers.2.bias": "unet_up_blocks.3.resnets.0.conv1.bias",
"model.diffusion_model.output_blocks.9.0.in_layers.2.weight": "unet_up_blocks.3.resnets.0.conv1.weight",
"model.diffusion_model.output_blocks.9.0.out_layers.0.bias": "unet_up_blocks.3.resnets.0.norm2.bias",
"model.diffusion_model.output_blocks.9.0.out_layers.0.weight": "unet_up_blocks.3.resnets.0.norm2.weight",
@@ -1129,7 +1129,7 @@
"model.diffusion_model.time_embed.0.bias": "unet_time_embedding.linear_1.bias",
"model.diffusion_model.time_embed.0.weight": "unet_time_embedding.linear_1.weight",
"model.diffusion_model.time_embed.2.bias": "unet_time_embedding.linear_2.bias",
"model.diffusion_model.time_embed.2.weight": "unet_mid_block.resnets.1.time_emb_proj.weight"
"model.diffusion_model.time_embed.2.weight": "unet_time_embedding.linear_2.weight"
},
"ldm_diffusers_shape_map": {
"first_stage_model.decoder.mid.attn_1.k.weight": [