--- license: other base_model: "black-forest-labs/FLUX.1-dev" tags: - flux - flux-diffusers - text-to-image - diffusers - simpletuner - safe-for-work - lora - template:sd-lora - lycoris inference: true widget: - text: 'unconditional (blank prompt)' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_0_0.png - text: 'in the style of c0d3xm3nd0z4, Woman with red hair playing chess in park, explosion in background. Chess pieces as Aztec warrior figurines, park as stylized Aztec garden. Explosion depicted as Xiuhcoatl (fire serpent), Codex Mendoza folio 67r warrior style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_1_0.png - text: 'in the style of c0d3xm3nd0z4, Woman in beanie holding Aztec-style clay cup at cafe. Cafe interior as Aztec calli (house), woman seated on icpalli (reed mat). Coffee steam as speech scrolls, Codex Mendoza folio 56r daily life scene style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_2_0.png - text: 'in the style of c0d3xm3nd0z4, Horse as Aztec deity Mixcoatl DJing in tlachtli (ballcourt) nightclub. Turntables as Aztec drums, martini as pulque vessel. Smoke and lasers as stylized Aztec symbols, Codex Mendoza folio 30r tribute items style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_3_0.png - text: 'in the style of c0d3xm3nd0z4, Man showing off tilmatli (cloak) on Aztec coastal scene. Shark leaping as stylized Cipactli (earth monster) from water. Beach as simplified altepetl (city-state) glyph, Codex Mendoza folio 47r coastal tribute style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_4_0.png - text: 'in the style of c0d3xm3nd0z4, Bear building calli (house) in snow-covered tepetl (mountains). Bear wearing Aztec tlahuiztli (warrior suit), logs depicted as macuahuitl (war clubs). Snow as jade beads, Codex Mendoza folio 2r architectural style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_5_0.png - text: 'in the style of c0d3xm3nd0z4, Woman playing teponaztli (Aztec drum) on icpalli (throne) stage. Punk hairstyle as quetzal feather headdress, laser lights as Aztec year signs. Audience as typical Codex Mendoza crowd from folio 2r.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_6_0.png - text: 'in the style of c0d3xm3nd0z4, Bearded man crafting icpalli (reed throne) in Aztec workshop. Beard styled as Aztec noble''s, tools as tribute items. Workshop layout mimicking Codex Mendoza folio 60r craft scene.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_7_0.png - text: 'in the style of c0d3xm3nd0z4, Man modeling Aztec tilmatli (cloak) and maxtlatl (loincloth) against blank calli (house) wall. Studio lights as sun and moon glyphs, posture mimicking Codex Mendoza folio 71r warrior poses.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_8_0.png - text: 'in the style of c0d3xm3nd0z4, Man holding amatl (paper) sign with glyphs spelling ''this is a sign''. Man''s posture and clothing based on Codex Mendoza folio 58r scribe figure.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_9_0.png - text: 'in the style of c0d3xm3nd0z4, Bulldog as Aztec Xolotl (dog deity) in devastated Tenochtitlan scene. Leather jacket as tlahuiztli (warrior suit), shotgun as atlatl (spear-thrower). Motorcycle as stylized Xiuhcoatl (fire serpent), Codex Mendoza folio 67r warrior style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_10_0.png - text: 'in the style of c0d3xm3nd0z4, Hamster as small Aztec deity in elaborate feather headdress. Hamster wheel depicted as Aztec calendar stone, food pellets as tribute items. Scene layout based on Codex Mendoza folio 15r tribute presentation.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_11_0.png - text: 'in the style of c0d3xm3nd0z4, Aztec tlatoani (ruler) on icpalli (reed throne) wearing copilli (crown). Tepetl (hill) glyph behind figure, year sign in corner. Ruler holding xihuitzolli (turquoise diadem) and quetzal feather fan, Codex Mendoza folio 69r style.' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_12_0.png - text: 'hamster in the style of c0d3xm3nd0z4' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_13_0.png --- # Flux-Codex_Mendoza-ST-itso-LoKr This is a LyCORIS adapter derived from [black-forest-labs/FLUX.1-dev](https://huggingface.co/black-forest-labs/FLUX.1-dev). The main validation prompt used during training was: ``` hamster in the style of c0d3xm3nd0z4 ``` ## Validation settings - CFG: `3.0` - CFG Rescale: `0.0` - Steps: `20` - Sampler: `None` - Seed: `42` - Resolution: `1024x1024` Note: The validation settings are not necessarily the same as the [training settings](#training-settings). You can find some example images in the following gallery: The text encoder **was not** trained. You may reuse the base model text encoder for inference. ## Training settings - Training epochs: 1 - Training steps: 1500 - Learning rate: 0.0008 - Effective batch size: 2 - Micro-batch size: 2 - Gradient accumulation steps: 1 - Number of GPUs: 1 - Prediction type: flow-matching - Rescaled betas zero SNR: False - Optimizer: adamw_bf16 - Precision: Pure BF16 - Quantised: Yes: int8-quanto - Xformers: Not used - LyCORIS Config: ```json { "algo": "lokr", "multiplier": 1.0, "linear_dim": 10000, "linear_alpha": 1, "factor": 16, "apply_preset": { "target_module": [ "Attention", "FeedForward" ], "module_algo_map": { "Attention": { "factor": 16 }, "FeedForward": { "factor": 8 } } } } ``` ## Datasets ### codex-mendoza-itso-512 - Repeats: 10 - Total number of images: 59 - Total number of aspect buckets: 1 - Resolution: 0.262144 megapixels - Cropped: False - Crop style: None - Crop aspect: None ### codex-mendoza-itso-1024 - Repeats: 10 - Total number of images: 59 - Total number of aspect buckets: 2 - Resolution: 1.048576 megapixels - Cropped: False - Crop style: None - Crop aspect: None ### codex-mendoza-itso-512-crop - Repeats: 10 - Total number of images: 59 - Total number of aspect buckets: 1 - Resolution: 0.262144 megapixels - Cropped: True - Crop style: random - Crop aspect: square ### codex-mendoza-itso-1024-crop - Repeats: 10 - Total number of images: 59 - Total number of aspect buckets: 1 - Resolution: 1.048576 megapixels - Cropped: True - Crop style: random - Crop aspect: square ## Inference ```python import torch from diffusers import DiffusionPipeline from lycoris import create_lycoris_from_weights model_id = 'black-forest-labs/FLUX.1-dev' 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 = "hamster in the style of c0d3xm3nd0z4" pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') image = pipeline( prompt=prompt, num_inference_steps=20, generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(1641421826), width=1024, height=1024, guidance_scale=3.0, ).images[0] image.save("output.png", format="PNG") ```