Ming Yang
commited on
Commit
·
b42d272
1
Parent(s):
8547cf9
update: 13K steps
Browse filesSigned-off-by: Ming Yang <vivym@live.com>
- README.md +1 -1
- unet/config.json +2 -2
- unet/diffusion_pytorch_model.safetensors +1 -1
README.md
CHANGED
@@ -36,7 +36,7 @@ extra_gated_heading: Please read the LICENSE to access this model
|
|
36 |
|
37 |
# v-prediction variant of BK-SDM-Tiny
|
38 |
|
39 |
-
Trained on text-image pairs from Midjourney v5.2,
|
40 |
|
41 |
# BK-SDM Model Card
|
42 |
Block-removed Knowledge-distilled Stable Diffusion Model (BK-SDM) is an architecturally compressed SDM for efficient general-purpose text-to-image synthesis. This model is bulit with (i) removing several residual and attention blocks from the U-Net of [Stable Diffusion v1.4]( https://huggingface.co/CompVis/stable-diffusion-v1-4) and (ii) distillation pretraining on only 0.22M LAION pairs (fewer than 0.1% of the full training set). Despite being trained with very limited resources, our compact model can imitate the original SDM by benefiting from transferred knowledge.
|
|
|
36 |
|
37 |
# v-prediction variant of BK-SDM-Tiny
|
38 |
|
39 |
+
Trained on text-image pairs from Midjourney v5.2, 13000 steps with a batch size of 2048.
|
40 |
|
41 |
# BK-SDM Model Card
|
42 |
Block-removed Knowledge-distilled Stable Diffusion Model (BK-SDM) is an architecturally compressed SDM for efficient general-purpose text-to-image synthesis. This model is bulit with (i) removing several residual and attention blocks from the U-Net of [Stable Diffusion v1.4]( https://huggingface.co/CompVis/stable-diffusion-v1-4) and (ii) distillation pretraining on only 0.22M LAION pairs (fewer than 0.1% of the full training set). Despite being trained with very limited resources, our compact model can imitate the original SDM by benefiting from transferred knowledge.
|
unet/config.json
CHANGED
@@ -1,7 +1,7 @@
|
|
1 |
{
|
2 |
"_class_name": "UNet2DConditionModel",
|
3 |
"_diffusers_version": "0.20.2",
|
4 |
-
"_name_or_path": "
|
5 |
"act_fn": "silu",
|
6 |
"addition_embed_type": null,
|
7 |
"addition_embed_type_num_heads": 64,
|
@@ -44,7 +44,7 @@
|
|
44 |
"num_attention_heads": null,
|
45 |
"num_class_embeds": null,
|
46 |
"only_cross_attention": false,
|
47 |
-
"optimization_step":
|
48 |
"out_channels": 4,
|
49 |
"power": 0.6666666666666666,
|
50 |
"projection_class_embeddings_input_dim": null,
|
|
|
1 |
{
|
2 |
"_class_name": "UNet2DConditionModel",
|
3 |
"_diffusers_version": "0.20.2",
|
4 |
+
"_name_or_path": "nota-ai/bk-sdm-tiny",
|
5 |
"act_fn": "silu",
|
6 |
"addition_embed_type": null,
|
7 |
"addition_embed_type_num_heads": 64,
|
|
|
44 |
"num_attention_heads": null,
|
45 |
"num_class_embeds": null,
|
46 |
"only_cross_attention": false,
|
47 |
+
"optimization_step": 8000,
|
48 |
"out_channels": 4,
|
49 |
"power": 0.6666666666666666,
|
50 |
"projection_class_embeddings_input_dim": null,
|
unet/diffusion_pytorch_model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1293583616
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:828fbf7e98271e151c1b23b1c5a780871119ddc2a79443382dd49cd12089b72e
|
3 |
size 1293583616
|