Model card auto-generated by SimpleTuner

README.md

@@ -10,7 +10,7 @@ tags:
   - not-for-all-audiences
   - lora
   - template:sd-lora
-  - standard
+  - lycoris
 inference: true
 widget:
 - text: 'unconditional (blank prompt)'
@@ -107,7 +107,7 @@ widget:
 
 # sd3-cheechandchong-regularised
 
-This is a standard PEFT LoRA derived from [stabilityai/stable-diffusion-3.5-large](https://huggingface.co/stabilityai/stable-diffusion-3.5-large).
+This is a LyCORIS adapter derived from [stabilityai/stable-diffusion-3.5-large](https://huggingface.co/stabilityai/stable-diffusion-3.5-large).
 
 
 The main validation prompt used during training was:
@@ -139,9 +139,9 @@ You may reuse the base model text encoder for inference.
 
 ## Training settings
 
-- Training epochs: 36
-- Training steps: 2800
-- Learning rate: 0.0001
+- Training epochs: 1
+- Training steps: 100
+- Learning rate: 1e-05
 - Max grad norm: 0.01
 - Effective batch size: 12
   - Micro-batch size: 4
@@ -153,11 +153,31 @@ You may reuse the base model text encoder for inference.
 - Precision: Pure BF16
 - Quantised: Yes: int8-quanto
 - Xformers: Not used
-- LoRA Rank: 64
-- LoRA Alpha: 64.0
-- LoRA Dropout: 0.1
-- LoRA initialisation style: default
-    
+- LyCORIS Config:
+```json
+{
+    "bypass_mode": true,
+    "algo": "lokr",
+    "multiplier": 1.0,
+    "full_matrix": true,
+    "linear_dim": 10000,
+    "linear_alpha": 1,
+    "factor": 12,
+    "apply_preset": {
+        "target_module": [
+            "JointTransformerBlock"
+        ],
+        "module_algo_map": {
+            "FeedForward": {
+                "factor": 6
+            },
+            "JointTransformerBlock": {
+                "factor": 12
+            }
+        }
+    }
+}
+```
 
 ## Datasets
 
@@ -178,11 +198,13 @@ You may reuse the base model text encoder for inference.
 ```python
 import torch
 from diffusers import DiffusionPipeline
+from lycoris import create_lycoris_from_weights
 
 model_id = 'stabilityai/stable-diffusion-3.5-large'
-adapter_id = 'bghira/sd3-cheechandchong-regularised'
-pipeline = DiffusionPipeline.from_pretrained(model_id)
-pipeline.load_lora_weights(adapter_id)
+adapter_id = 'pytorch_lora_weights.safetensors' # you will have to download this manually
+lora_scale = 1.0
+wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_id, pipeline.transformer)
+wrapper.merge_to()
 
 prompt = "A photo-realistic image of a cat"
 negative_prompt = 'blurry, cropped, ugly'