Upload 5 files

captions/Seaweed p5 print.txt

@@ -0,0 +1 @@
+An abstract painting made with acrylic paint on canvas depicts a scene of green seaweed in a dark background. The seaweed is painted in a swirling pattern, giving the impression of a flowing underwater environment. The painting is in the style of a seaweed painting, with the use of green and black colors to create a sense of depth and movement. The composition of the painting is dynamic, with the seaweed spreading outwards from the center towards the edges of the canvas. The overall effect is one of fluidity and organic growth, evoking a sense of the natural world.

captions/file 6.txt

@@ -0,0 +1 @@
+An abstract painting made with acrylic paint on canvas depicts a green, flowing, organic form. The painting is in the style of a seaweed painting, with the green color dominating the canvas and creating a sense of movement and fluidity. The composition is dynamic and abstract, with no clear subject or focal point, but rather a series of interconnected shapes and lines that flow across the canvas. The painting evokes a sense of nature and the ocean, with its vibrant green color and organic forms.

captions/seaweed pt4.txt

@@ -0,0 +1 @@
+An abstract painting made with acrylic paint on canvas depicts a scene of a green plant with long tendrils and a large green leaf. The painting is in the style of a seaweed painting, with a focus on the intricate details of the plant's structure and texture. The colors used in the painting are primarily green, with some black accents to add depth and contrast. The painting is displayed on a yellow wall, which provides a vibrant and lively backdrop for the artwork.

config.yaml

@@ -0,0 +1,60 @@
+job: custom_job
+config:
+  name: flux_train_replicate
+  process:
+  - type: custom_sd_trainer
+    training_folder: output
+    device: cuda:0
+    trigger_word: a seaweed painting
+    network:
+      type: lora
+      linear: 16
+      linear_alpha: 16
+    save:
+      dtype: float16
+      save_every: 1001
+      max_step_saves_to_keep: 1
+    datasets:
+    - folder_path: input_images
+      caption_ext: txt
+      caption_dropout_rate: 0.82
+      shuffle_tokens: false
+      cache_latents_to_disk: false
+      cache_latents: true
+      resolution:
+      - 512
+      - 768
+      - 1024
+    train:
+      batch_size: 1
+      steps: 1000
+      gradient_accumulation_steps: 1
+      train_unet: true
+      train_text_encoder: false
+      content_or_style: balanced
+      gradient_checkpointing: true
+      noise_scheduler: flowmatch
+      optimizer: adamw8bit
+      lr: 0.0004
+      ema_config:
+        use_ema: true
+        ema_decay: 0.99
+      dtype: bf16
+    model:
+      name_or_path: FLUX.1-dev
+      is_flux: true
+      quantize: true
+    sample:
+      sampler: flowmatch
+      sample_every: 1001
+      width: 1024
+      height: 1024
+      prompts: []
+      neg: ''
+      seed: 42
+      walk_seed: true
+      guidance_scale: 3.5
+      sample_steps: 28
+meta:
+  name: flux_train_replicate
+  version: '1.0'

lora.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2f419d45cc8a4d3d55ecfe52ba237392e18ad28e53991690c4f386f384298566
+size 171969440