rstudioweb commited on
Commit
1ebe7db
·
1 Parent(s): c11d268

Add rupsababy29012026_nsfw_loras folder

Browse files
rupsababy29012026_nsfw_loras/README.md ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - text-to-image
4
+ - flux
5
+ - lora
6
+ - diffusers
7
+ - template:sd-lora
8
+ - fluxgym
9
+
10
+
11
+ base_model: getRekt
12
+ instance_prompt: rupsa26v
13
+ license: getRekt
14
+ license_name: getRekt
15
+ license_link: getRekt
16
+ ---
17
+
18
+ # RupsaBaby
19
+
20
+ A Flux LoRA trained on a local computer with [Fluxgym](https://github.com/cocktailpeanut/fluxgym)
21
+
22
+ <Gallery />
23
+
24
+ ## Trigger words
25
+
26
+ You should use `rupsa26v` to trigger the image generation.
27
+
28
+ ## Download model and use it with ComfyUI, AUTOMATIC1111, SD.Next, Invoke AI, Forge, etc.
29
+
30
+ Weights for this model are available in Safetensors format.
31
+
rupsababy29012026_nsfw_loras/dataset.toml ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [general]
2
+ shuffle_caption = false
3
+ caption_extension = '.txt'
4
+ keep_tokens = 1
5
+
6
+ [[datasets]]
7
+ resolution = 1024
8
+ batch_size = 1
9
+ keep_tokens = 1
10
+
11
+ [[datasets.subsets]]
12
+ image_dir = '/teamspace/studios/this_studio/fluxgym/datasets/rupsababy'
13
+ class_tokens = 'rupsa26v'
14
+ num_repeats = 5
rupsababy29012026_nsfw_loras/rupsababy-000001.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ec2f7df4326883e8a0610bedc7e6de2de238f8ea77c7d4c03dea8d80c3acf0a
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:689010e7d3dcf1e1e9de2832b4e6920d1ed68bb0440341765d5e6b48637cc9cd
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000003.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0bba9eeb59703c87dca488ef74c2f8177a5d6b01976bd78cc9c0075dcb6dce56
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a0bba871c854e2f93cad6ff01400c5fe9ac969d1a61d0c67ee054804ca10bf6
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000005.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a00bd9d65bb4d168767f857f739cdd1cd05f362251e16fa66f5a3c9364dc4018
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000006.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f11df51efd7ae8e95922ca804269ac290904383acfac213af8b7ac58d0c1f80
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3fb953d86045857fb4896a5a87da588746fb19e010b4a0fe7db21da0f798f12b
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000008.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b71d3978e9541ebf8d47cafd065913719c5612dbc8253592167dbc16a04f004c
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy-000009.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dba711ec5c28bb4dc5a568abe4a6f3a89c0bd3dc62e72cddee10f9be5a6d9398
3
+ size 158613232
rupsababy29012026_nsfw_loras/rupsababy.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62cc67dab381db2dcab407dc98d9c440ee14127d4e6571b30cfd75f2eb0b2bb8
3
+ size 158613232
rupsababy29012026_nsfw_loras/sample_prompts.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ rupsa26v, A young 21y adult girl with a symmetrical, oval-to-round face shape, soft jawline, and gently rounded chin, featuring medium-sized almond-shaped dark brown eyes set evenly beneath straight, natural brows, a straight, proportionate nose with a rounded tip, and medium-full lips with a neutral contour and relaxed expression, her skin tone a warm medium-brown with golden undertones showing natural texture, mild hyperpigmentation, and subtle acne marks for a realistic, unretouched complexion with a matte to slightly dewy finish under soft lighting, her dark brown to black hair parted at the center, pulled back casually with light flyaways for a natural, minimal-styling look, standing in a full frontal pose with her head bowed downward, hands gently resting on her lower abdomen near the pubic area, shoulders relaxed, back straight, hips slightly forward, showcasing a naturally proportioned, softly built physique with gently sloped shoulders, medium bust, and natural waistline, conveying a healthy, everyday form rather than stylized exaggeration, all rendered in a studio portrait style with dramatic chiaroscuro lighting against a dark, textured backdrop.
rupsababy29012026_nsfw_loras/train.sh ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ accelerate launch \
2
+ --mixed_precision bf16 \
3
+ --num_cpu_threads_per_process 1 \
4
+ sd-scripts/flux_train_network.py \
5
+ --pretrained_model_name_or_path "/teamspace/studios/this_studio/fluxgym/models/unet/getRekt/getphatFLUXReality_v5HardcoreFP8.safetensors" \
6
+ --clip_l "/teamspace/studios/this_studio/fluxgym/models/clip/clip_l.safetensors" \
7
+ --t5xxl "/teamspace/studios/this_studio/fluxgym/models/clip/t5xxl_fp16.safetensors" \
8
+ --ae "/teamspace/studios/this_studio/fluxgym/models/vae/ae.sft" \
9
+ --cache_latents_to_disk \
10
+ --save_model_as safetensors \
11
+ --sdpa --persistent_data_loader_workers \
12
+ --max_data_loader_n_workers 2 \
13
+ --seed 42 \
14
+ --gradient_checkpointing \
15
+ --mixed_precision bf16 \
16
+ --save_precision bf16 \
17
+ --network_module networks.lora_flux \
18
+ --network_dim 16 \
19
+ --optimizer_type adafactor \
20
+ --optimizer_args "relative_step=False" "scale_parameter=False" "warmup_init=False" \
21
+ --lr_scheduler constant_with_warmup \
22
+ --max_grad_norm 0.0 \
23
+ --learning_rate 5e-4 \
24
+ --cache_text_encoder_outputs \
25
+ --cache_text_encoder_outputs_to_disk \
26
+ --fp8_base \
27
+ --highvram \
28
+ --max_train_epochs 10 \
29
+ --save_every_n_epochs 1 \
30
+ --dataset_config "/teamspace/studios/this_studio/fluxgym/outputs/rupsababy/dataset.toml" \
31
+ --output_dir "/teamspace/studios/this_studio/fluxgym/outputs/rupsababy" \
32
+ --output_name rupsababy \
33
+ --timestep_sampling shift \
34
+ --discrete_flow_shift 3.1582 \
35
+ --model_prediction_type raw \
36
+ --guidance_scale 1 \
37
+ --loss_type l2 \
38
+ --enable_bucket \
39
+ --min_snr_gamma 5 \
40
+ --multires_noise_discount 0.3 \
41
+ --multires_noise_iterations 6 \
42
+ --noise_offset 0.1 \
43
+ --train_batch_size 2