Text-to-Image
Diffusers
TensorBoard
diffusers-training
lora
ideogram4
ideogram4-diffusers
template:sd-lora
Instructions to use linoyts/ideogram4-yarn-art-fp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use linoyts/ideogram4-yarn-art-fp8 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Disty0/Ideogram-4-SDNQ-FP8", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("linoyts/ideogram4-yarn-art-fp8") prompt = "yarn art style" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
End of training
Browse files- .gitattributes +20 -0
- README.md +78 -0
- checkpoint-1000/optimizer.bin +3 -0
- checkpoint-1000/pytorch_lora_weights.safetensors +3 -0
- checkpoint-1000/random_states_0.pkl +3 -0
- checkpoint-1000/scheduler.bin +3 -0
- checkpoint-250/optimizer.bin +3 -0
- checkpoint-250/pytorch_lora_weights.safetensors +3 -0
- checkpoint-250/random_states_0.pkl +3 -0
- checkpoint-250/scheduler.bin +3 -0
- checkpoint-500/optimizer.bin +3 -0
- checkpoint-500/pytorch_lora_weights.safetensors +3 -0
- checkpoint-500/random_states_0.pkl +3 -0
- checkpoint-500/scheduler.bin +3 -0
- checkpoint-750/optimizer.bin +3 -0
- checkpoint-750/pytorch_lora_weights.safetensors +3 -0
- checkpoint-750/random_states_0.pkl +3 -0
- checkpoint-750/scheduler.bin +3 -0
- logs/dreambooth-ideogram4-lora/1781642354.4393263/events.out.tfevents.1781642354.209-20-158-47.69568.1 +3 -0
- logs/dreambooth-ideogram4-lora/1781642354.4454818/hparams.yml +110 -0
- logs/dreambooth-ideogram4-lora/events.out.tfevents.1781642354.209-20-158-47.69568.0 +3 -0
- pytorch_lora_weights.safetensors +3 -0
- val_images/val_e0_0.png +3 -0
- val_images/val_e0_1.png +3 -0
- val_images/val_e100_0.png +3 -0
- val_images/val_e100_1.png +3 -0
- val_images/val_e120_0.png +3 -0
- val_images/val_e120_1.png +3 -0
- val_images/val_e140_0.png +3 -0
- val_images/val_e140_1.png +3 -0
- val_images/val_e160_0.png +3 -0
- val_images/val_e160_1.png +3 -0
- val_images/val_e180_0.png +3 -0
- val_images/val_e180_1.png +3 -0
- val_images/val_e20_0.png +3 -0
- val_images/val_e20_1.png +3 -0
- val_images/val_e40_0.png +3 -0
- val_images/val_e40_1.png +3 -0
- val_images/val_e60_0.png +3 -0
- val_images/val_e60_1.png +3 -0
- val_images/val_e80_0.png +3 -0
- val_images/val_e80_1.png +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,23 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
val_images/val_e0_0.png filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
val_images/val_e0_1.png filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
val_images/val_e100_0.png filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
val_images/val_e100_1.png filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
val_images/val_e120_0.png filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
val_images/val_e120_1.png filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
val_images/val_e140_0.png filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
val_images/val_e140_1.png filter=lfs diff=lfs merge=lfs -text
|
| 44 |
+
val_images/val_e160_0.png filter=lfs diff=lfs merge=lfs -text
|
| 45 |
+
val_images/val_e160_1.png filter=lfs diff=lfs merge=lfs -text
|
| 46 |
+
val_images/val_e180_0.png filter=lfs diff=lfs merge=lfs -text
|
| 47 |
+
val_images/val_e180_1.png filter=lfs diff=lfs merge=lfs -text
|
| 48 |
+
val_images/val_e20_0.png filter=lfs diff=lfs merge=lfs -text
|
| 49 |
+
val_images/val_e20_1.png filter=lfs diff=lfs merge=lfs -text
|
| 50 |
+
val_images/val_e40_0.png filter=lfs diff=lfs merge=lfs -text
|
| 51 |
+
val_images/val_e40_1.png filter=lfs diff=lfs merge=lfs -text
|
| 52 |
+
val_images/val_e60_0.png filter=lfs diff=lfs merge=lfs -text
|
| 53 |
+
val_images/val_e60_1.png filter=lfs diff=lfs merge=lfs -text
|
| 54 |
+
val_images/val_e80_0.png filter=lfs diff=lfs merge=lfs -text
|
| 55 |
+
val_images/val_e80_1.png filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,78 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: Disty0/Ideogram-4-SDNQ-FP8
|
| 3 |
+
library_name: diffusers
|
| 4 |
+
license: other
|
| 5 |
+
instance_prompt: yarn art style
|
| 6 |
+
widget: []
|
| 7 |
+
tags:
|
| 8 |
+
- text-to-image
|
| 9 |
+
- diffusers-training
|
| 10 |
+
- diffusers
|
| 11 |
+
- lora
|
| 12 |
+
- ideogram4
|
| 13 |
+
- ideogram4-diffusers
|
| 14 |
+
- template:sd-lora
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
<!-- This model card has been generated automatically according to the information the training script had access to. You
|
| 18 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
# Ideogram4 DreamBooth LoRA - linoyts/ideogram4-yarn-art-fp8
|
| 22 |
+
|
| 23 |
+
<Gallery />
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
These are linoyts/ideogram4-yarn-art-fp8 DreamBooth LoRA weights for Disty0/Ideogram-4-SDNQ-FP8.
|
| 28 |
+
|
| 29 |
+
The weights were trained using [DreamBooth](https://dreambooth.github.io/) with the [Ideogram4 diffusers trainer](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_ideogram4.md).
|
| 30 |
+
|
| 31 |
+
This LoRA was trained with **FP8 SDNQ** quantization of the base model.
|
| 32 |
+
|
| 33 |
+
## Trigger words
|
| 34 |
+
|
| 35 |
+
You should use `yarn art style` to trigger the image generation.
|
| 36 |
+
|
| 37 |
+
Ideogram 4 is trained on structured JSON captions, so for best results pass a caption in the JSON schema (see the [prompt format](https://github.com/huggingface/diffusers/blob/main/examples/dreambooth/README_ideogram4.md#prompt-format)) or a short prompt with `prompt_upsampling=True`.
|
| 38 |
+
|
| 39 |
+
## Download model
|
| 40 |
+
|
| 41 |
+
[Download the *.safetensors LoRA](linoyts/ideogram4-yarn-art-fp8/tree/main) in the Files & versions tab.
|
| 42 |
+
|
| 43 |
+
## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)
|
| 44 |
+
|
| 45 |
+
> **Note:** the base is an SDNQ fp8 checkpoint. For inference, install `sdnq` and dequantize the transformer to bf16 (`from sdnq.quantizer import dequantize_sdnq_model`), or load a non-quantized base.
|
| 46 |
+
|
| 47 |
+
```py
|
| 48 |
+
from diffusers import Ideogram4Pipeline
|
| 49 |
+
import torch
|
| 50 |
+
|
| 51 |
+
pipeline = Ideogram4Pipeline.from_pretrained("Disty0/Ideogram-4-SDNQ-FP8", torch_dtype=torch.bfloat16)
|
| 52 |
+
pipeline.to('cuda')
|
| 53 |
+
pipeline.load_lora_weights('linoyts/ideogram4-yarn-art-fp8', weight_name='pytorch_lora_weights.safetensors')
|
| 54 |
+
image = pipeline('{"high_level_description":"A whimsical yarn-craft illustration of Yoda rendered entirely in knitted and crocheted fiber, mounted inside a round wooden embroidery hoop on a linen backdrop, a soft diffused daylight photograph with 1:1 aspect ratio.","compositional_deconstruction":{"background":"Neutral pale linen backdrop filling the frame, fine grain texture visible, soft diffused daylight from upper-left casting gentle shadows across the surface, palette dominated by mossy greens, cream, oatmeal browns, and warm wood tones against the pale linen ground.","elements":[{"type":"obj","desc":"Round wooden embroidery hoop framing the entire piece, light birch finish with a small brass tension screw at the bottom edge. The hoop sits flat on the linen backdrop, centered in the frame."},{"type":"obj","desc":"Yoda's head and shoulders sculpted from soft moss-green yarn, filling roughly 80% of the hoop interior, centered. Large pointed ears with visible stitch texture and tiny pink felt inner-ear patches. Wrinkled forehead formed by gathered cream-colored yarn tufts. Deep-set eyes made from two small black beads set into white felt circles. Long hooked nose crafted from a single rolled cord of darker green yarn. Mouth stitched as a small closed curve in dark thread. Wispy white beard and sideburns made from looped mohair fibers cascading down the chin. Three-dimensional stitch texture reads clearly under the soft directional light."},{"type":"obj","desc":"Small brown robe collar at the base of Yoda's neck, crocheted in chunky oatmeal-brown yarn with a loose cable-knit pattern, sitting just below the chin within the lower portion of the hoop."},{"type":"obj","desc":"Scattered loose yarn scraps in moss-green, cream, and oatmeal-brown resting on the linen surface beneath the lower-left edge of the hoop, falling into soft focus."},{"type":"obj","desc":"A pair of small wooden crochet hooks lying crossed on the linen surface beneath the lower-right edge of the hoop, light wood handles with tapered metal tips, softly out of focus."}]}}').images[0]
|
| 55 |
+
```
|
| 56 |
+
|
| 57 |
+
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)
|
| 58 |
+
|
| 59 |
+
## License
|
| 60 |
+
|
| 61 |
+
Please adhere to the licensing terms of the base model [Disty0/Ideogram-4-SDNQ-FP8](https://huggingface.co/Disty0/Ideogram-4-SDNQ-FP8).
|
| 62 |
+
|
| 63 |
+
|
| 64 |
+
## Intended uses & limitations
|
| 65 |
+
|
| 66 |
+
#### How to use
|
| 67 |
+
|
| 68 |
+
```python
|
| 69 |
+
# TODO: add an example code snippet for running this diffusion pipeline
|
| 70 |
+
```
|
| 71 |
+
|
| 72 |
+
#### Limitations and bias
|
| 73 |
+
|
| 74 |
+
[TODO: provide examples of latent issues and potential remediations]
|
| 75 |
+
|
| 76 |
+
## Training details
|
| 77 |
+
|
| 78 |
+
[TODO: describe the data used to train the model]
|
checkpoint-1000/optimizer.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:89206668cea366327c7ed57799c1d7e2c15c39ad71f5d37ebf75f39f698134a3
|
| 3 |
+
size 381428727
|
checkpoint-1000/pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3eb16a4429a1f1c2caf8b2a954d6322968dfb9df8a7b25d5cf9aa1e070a37c2c
|
| 3 |
+
size 190574176
|
checkpoint-1000/random_states_0.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41051f48d7cb0734383924495d540a6dace1fd39ccfb12208187f106552e6cf2
|
| 3 |
+
size 14821
|
checkpoint-1000/scheduler.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d5ca5fd1f3d50dfbd88c5f854e399c1c22a842d2231d41415da16bf924479e76
|
| 3 |
+
size 1401
|
checkpoint-250/optimizer.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18702973f9c859be35bff5c516129643cdbf75322f2d4d3af4236e601aa3e032
|
| 3 |
+
size 381428727
|
checkpoint-250/pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ffdc5819347df195f9cd9b45f0ecf2d4e7fc629d02a12a001bbf2bcbe0ae1b2
|
| 3 |
+
size 190574176
|
checkpoint-250/random_states_0.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6919268786ebb1950bc1656cc1aa77fad22342e672b8877e03a044961b17571e
|
| 3 |
+
size 14821
|
checkpoint-250/scheduler.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e9fd216eaa69ff4eb809e0f68c4cab14943314d6c9795506c007a05963b86fc8
|
| 3 |
+
size 1401
|
checkpoint-500/optimizer.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76c122291c2b246394662aa782af84733052274412270cb11b4ff3ea58d3c602
|
| 3 |
+
size 381428727
|
checkpoint-500/pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32b15c47f950f1c91198b475af76f0858176b9304b50a4473b972e3de9376a33
|
| 3 |
+
size 190574176
|
checkpoint-500/random_states_0.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13cc48695859ebf10c0a3d9222795df99d23a3613880a20739c5c71f5ce979ce
|
| 3 |
+
size 14821
|
checkpoint-500/scheduler.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8baa45cc00e6da1eefbbbc2d50fb568432ef1e20241f0a3571b49dcb8d62f7f2
|
| 3 |
+
size 1401
|
checkpoint-750/optimizer.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18ecea553da1a627af38e664b459faf77778bd92e78226a2a0c45c88e0238038
|
| 3 |
+
size 381428727
|
checkpoint-750/pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eff226efac91e60bb28820f559bc862a7c8694609d8bd15581859595ad4d51df
|
| 3 |
+
size 190574176
|
checkpoint-750/random_states_0.pkl
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:039c5909f31dcaf7023790df4454854c35366a077674f083fc8a2f65297230de
|
| 3 |
+
size 14821
|
checkpoint-750/scheduler.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af27cfa62dd1a6d194c45c24a6db7d0c334459f3177682f8feacba9d4decd87b
|
| 3 |
+
size 1401
|
logs/dreambooth-ideogram4-lora/1781642354.4393263/events.out.tfevents.1781642354.209-20-158-47.69568.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:85ba6559942da6befb9a26b46a5ec46c4712c349611010debeed6dde1cf89841
|
| 3 |
+
size 6073
|
logs/dreambooth-ideogram4-lora/1781642354.4454818/hparams.yml
ADDED
|
@@ -0,0 +1,110 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
adam_beta1: 0.9
|
| 2 |
+
adam_beta2: 0.999
|
| 3 |
+
adam_epsilon: 1.0e-08
|
| 4 |
+
adam_weight_decay: 0.0001
|
| 5 |
+
adam_weight_decay_text_encoder: 0.001
|
| 6 |
+
allow_tf32: false
|
| 7 |
+
aspect_ratio_buckets: null
|
| 8 |
+
bnb_quantization_config_path: null
|
| 9 |
+
cache_dir: null
|
| 10 |
+
cache_latents: true
|
| 11 |
+
caption_column: text
|
| 12 |
+
center_crop: false
|
| 13 |
+
checkpointing_steps: 250
|
| 14 |
+
checkpoints_total_limit: null
|
| 15 |
+
class_data_dir: null
|
| 16 |
+
class_prompt: null
|
| 17 |
+
dataloader_num_workers: 0
|
| 18 |
+
dataset_config_name: null
|
| 19 |
+
dataset_name: Norod78/Yarn-art-style
|
| 20 |
+
disable_resolution_shift: false
|
| 21 |
+
disable_training_autocast: true
|
| 22 |
+
do_fp8_training: true
|
| 23 |
+
enable_npu_flash_attention: false
|
| 24 |
+
final_validation_prompt: null
|
| 25 |
+
fsdp_text_encoder: false
|
| 26 |
+
gradient_accumulation_steps: 4
|
| 27 |
+
gradient_checkpointing: true
|
| 28 |
+
guidance_scale: 3.5
|
| 29 |
+
hub_model_id: linoyts/ideogram4-yarn-art-fp8
|
| 30 |
+
hub_token: null
|
| 31 |
+
image_column: image
|
| 32 |
+
instance_data_dir: null
|
| 33 |
+
instance_prompt: yarn art style
|
| 34 |
+
learning_rate: 0.0001
|
| 35 |
+
local_rank: -1
|
| 36 |
+
logging_dir: logs
|
| 37 |
+
logit_mean: 0.0
|
| 38 |
+
logit_std: 1.5
|
| 39 |
+
lora_alpha: 32
|
| 40 |
+
lora_dropout: 0.0
|
| 41 |
+
lora_layers: null
|
| 42 |
+
lr_num_cycles: 1
|
| 43 |
+
lr_power: 1.0
|
| 44 |
+
lr_scheduler: constant
|
| 45 |
+
lr_warmup_steps: 0
|
| 46 |
+
max_grad_norm: 1.0
|
| 47 |
+
max_sequence_length: 2048
|
| 48 |
+
max_train_steps: 1000
|
| 49 |
+
mixed_precision: bf16
|
| 50 |
+
mode_scale: 1.29
|
| 51 |
+
num_class_images: 100
|
| 52 |
+
num_train_epochs: 200
|
| 53 |
+
num_validation_images: 2
|
| 54 |
+
offload: false
|
| 55 |
+
optimizer: AdamW
|
| 56 |
+
output_dir: /home/ubuntu/ideogram4-lora/run-fp8-full-yarn
|
| 57 |
+
pretrained_model_name_or_path: Disty0/Ideogram-4-SDNQ-FP8
|
| 58 |
+
prior_generation_precision: null
|
| 59 |
+
prior_loss_weight: 1.0
|
| 60 |
+
prodigy_beta3: null
|
| 61 |
+
prodigy_decouple: true
|
| 62 |
+
prodigy_safeguard_warmup: true
|
| 63 |
+
prodigy_use_bias_correction: true
|
| 64 |
+
prompt_enhancer_head_id: diffusers/qwen3-vl-8b-instruct-lm-head
|
| 65 |
+
push_to_hub: true
|
| 66 |
+
random_flip: false
|
| 67 |
+
rank: 32
|
| 68 |
+
repeats: 1
|
| 69 |
+
report_to: tensorboard
|
| 70 |
+
resolution: 1024
|
| 71 |
+
resume_from_checkpoint: null
|
| 72 |
+
revision: null
|
| 73 |
+
sample_batch_size: 4
|
| 74 |
+
scale_lr: false
|
| 75 |
+
seed: 0
|
| 76 |
+
skip_final_inference: true
|
| 77 |
+
text_encoder_lr: 5.0e-06
|
| 78 |
+
train_batch_size: 1
|
| 79 |
+
upcast_before_saving: false
|
| 80 |
+
upsample_prompt: true
|
| 81 |
+
use_8bit_adam: false
|
| 82 |
+
validation_epochs: 20
|
| 83 |
+
validation_prompt: '{"high_level_description":"A whimsical yarn-craft illustration
|
| 84 |
+
of Yoda rendered entirely in knitted and crocheted fiber, mounted inside a round
|
| 85 |
+
wooden embroidery hoop on a linen backdrop, a soft diffused daylight photograph
|
| 86 |
+
with 1:1 aspect ratio.","compositional_deconstruction":{"background":"Neutral pale
|
| 87 |
+
linen backdrop filling the frame, fine grain texture visible, soft diffused daylight
|
| 88 |
+
from upper-left casting gentle shadows across the surface, palette dominated by
|
| 89 |
+
mossy greens, cream, oatmeal browns, and warm wood tones against the pale linen
|
| 90 |
+
ground.","elements":[{"type":"obj","desc":"Round wooden embroidery hoop framing
|
| 91 |
+
the entire piece, light birch finish with a small brass tension screw at the bottom
|
| 92 |
+
edge. The hoop sits flat on the linen backdrop, centered in the frame."},{"type":"obj","desc":"Yoda''s
|
| 93 |
+
head and shoulders sculpted from soft moss-green yarn, filling roughly 80% of the
|
| 94 |
+
hoop interior, centered. Large pointed ears with visible stitch texture and tiny
|
| 95 |
+
pink felt inner-ear patches. Wrinkled forehead formed by gathered cream-colored
|
| 96 |
+
yarn tufts. Deep-set eyes made from two small black beads set into white felt circles.
|
| 97 |
+
Long hooked nose crafted from a single rolled cord of darker green yarn. Mouth stitched
|
| 98 |
+
as a small closed curve in dark thread. Wispy white beard and sideburns made from
|
| 99 |
+
looped mohair fibers cascading down the chin. Three-dimensional stitch texture reads
|
| 100 |
+
clearly under the soft directional light."},{"type":"obj","desc":"Small brown robe
|
| 101 |
+
collar at the base of Yoda''s neck, crocheted in chunky oatmeal-brown yarn with
|
| 102 |
+
a loose cable-knit pattern, sitting just below the chin within the lower portion
|
| 103 |
+
of the hoop."},{"type":"obj","desc":"Scattered loose yarn scraps in moss-green,
|
| 104 |
+
cream, and oatmeal-brown resting on the linen surface beneath the lower-left edge
|
| 105 |
+
of the hoop, falling into soft focus."},{"type":"obj","desc":"A pair of small wooden
|
| 106 |
+
crochet hooks lying crossed on the linen surface beneath the lower-right edge of
|
| 107 |
+
the hoop, light wood handles with tapered metal tips, softly out of focus."}]}}'
|
| 108 |
+
variant: null
|
| 109 |
+
weighting_scheme: logit_normal
|
| 110 |
+
with_prior_preservation: false
|
logs/dreambooth-ideogram4-lora/events.out.tfevents.1781642354.209-20-158-47.69568.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b5ec27cf285d623e7cd0025d58627fb15616b406dadafd4f73e9edf8da865aec
|
| 3 |
+
size 37118343
|
pytorch_lora_weights.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b6ab1c744df66e493a24d74df155559fdb5e5c0e2b825ff28b04511ef1cb59a
|
| 3 |
+
size 190574176
|
val_images/val_e0_0.png
ADDED
|
Git LFS Details
|
val_images/val_e0_1.png
ADDED
|
Git LFS Details
|
val_images/val_e100_0.png
ADDED
|
Git LFS Details
|
val_images/val_e100_1.png
ADDED
|
Git LFS Details
|
val_images/val_e120_0.png
ADDED
|
Git LFS Details
|
val_images/val_e120_1.png
ADDED
|
Git LFS Details
|
val_images/val_e140_0.png
ADDED
|
Git LFS Details
|
val_images/val_e140_1.png
ADDED
|
Git LFS Details
|
val_images/val_e160_0.png
ADDED
|
Git LFS Details
|
val_images/val_e160_1.png
ADDED
|
Git LFS Details
|
val_images/val_e180_0.png
ADDED
|
Git LFS Details
|
val_images/val_e180_1.png
ADDED
|
Git LFS Details
|
val_images/val_e20_0.png
ADDED
|
Git LFS Details
|
val_images/val_e20_1.png
ADDED
|
Git LFS Details
|
val_images/val_e40_0.png
ADDED
|
Git LFS Details
|
val_images/val_e40_1.png
ADDED
|
Git LFS Details
|
val_images/val_e60_0.png
ADDED
|
Git LFS Details
|
val_images/val_e60_1.png
ADDED
|
Git LFS Details
|
val_images/val_e80_0.png
ADDED
|
Git LFS Details
|
val_images/val_e80_1.png
ADDED
|
Git LFS Details
|