fofr commited on
Commit
6fac779
1 Parent(s): 8ef0de2

Upload folder using huggingface_hub

Browse files
Files changed (3) hide show
  1. README.md +46 -0
  2. config.yaml +60 -0
  3. lora.safetensors +3 -0
README.md ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ license_name: flux-1-dev-non-commercial-license
4
+ license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md
5
+ language:
6
+ - en
7
+ tags:
8
+ - flux
9
+ - diffusers
10
+ - lora
11
+ - replicate
12
+ base_model: "black-forest-labs/FLUX.1-dev"
13
+ pipeline_tag: text-to-image
14
+ # widget:
15
+ # - text: >-
16
+ # prompt
17
+ # output:
18
+ # url: https://...
19
+ instance_prompt: META_ORION_GLASSES
20
+ ---
21
+
22
+ # Flux Meta Orion
23
+
24
+ <!-- <Gallery /> -->
25
+
26
+ Trained on Replicate using:
27
+
28
+ https://replicate.com/ostris/flux-dev-lora-trainer/train
29
+
30
+
31
+ ## Trigger words
32
+ You should use `META_ORION_GLASSES` to trigger the image generation.
33
+
34
+
35
+ ## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)
36
+
37
+ ```py
38
+ from diffusers import AutoPipelineForText2Image
39
+ import torch
40
+
41
+ pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.float16).to('cuda')
42
+ pipeline.load_lora_weights('fofr/flux-meta-orion', weight_name='lora.safetensors')
43
+ image = pipeline('your prompt').images[0]
44
+ ```
45
+
46
+ For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)
config.yaml ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ job: custom_job
2
+ config:
3
+ name: flux_train_replicate
4
+ process:
5
+ - type: custom_sd_trainer
6
+ training_folder: output
7
+ device: cuda:0
8
+ trigger_word: META_ORION_GLASSES
9
+ network:
10
+ type: lora
11
+ linear: 16
12
+ linear_alpha: 16
13
+ save:
14
+ dtype: float16
15
+ save_every: 1001
16
+ max_step_saves_to_keep: 1
17
+ datasets:
18
+ - folder_path: input_images
19
+ caption_ext: txt
20
+ caption_dropout_rate: 0.05
21
+ shuffle_tokens: false
22
+ cache_latents_to_disk: false
23
+ cache_latents: true
24
+ resolution:
25
+ - 512
26
+ - 768
27
+ - 1024
28
+ train:
29
+ batch_size: 1
30
+ steps: 1000
31
+ gradient_accumulation_steps: 1
32
+ train_unet: true
33
+ train_text_encoder: false
34
+ content_or_style: balanced
35
+ gradient_checkpointing: true
36
+ noise_scheduler: flowmatch
37
+ optimizer: adamw
38
+ lr: 0.0004
39
+ ema_config:
40
+ use_ema: true
41
+ ema_decay: 0.99
42
+ dtype: bf16
43
+ model:
44
+ name_or_path: FLUX.1-dev
45
+ is_flux: true
46
+ quantize: true
47
+ sample:
48
+ sampler: flowmatch
49
+ sample_every: 1001
50
+ width: 1024
51
+ height: 1024
52
+ prompts: []
53
+ neg: ''
54
+ seed: 42
55
+ walk_seed: true
56
+ guidance_scale: 3.5
57
+ sample_steps: 28
58
+ meta:
59
+ name: flux_train_replicate
60
+ version: '1.0'
lora.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0dcc54ddd5ecaddda4013cd2df4f02682897268f669475b12a6a18c42a02ee5
3
+ size 171969440