Instructions to use OzzyGT/Kimodo_SOMA_Modular with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use OzzyGT/Kimodo_SOMA_Modular with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("OzzyGT/Kimodo_SOMA_Modular", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
File size: 1,337 Bytes
9ed958c 903e660 9ed958c 903e660 9ed958c 903e660 9ed958c 903e660 9ed958c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 | """Helpers shared by the Kimodo blocks."""
from dataclasses import replace
import torch
from diffusers.modular_pipelines import ModularPipelineBlocks
class ResolvableBlock(ModularPipelineBlocks):
"""Leaf block whose generic component specs can carry the model repository's resolved classes and location.
`KimodoBlocks(components_repo=...)` fills `component_overrides`, so a sub-block popped out of the pipeline still
loads the right custom classes from the right repository.
"""
component_overrides = {}
def resolve(self, specs):
return [
replace(spec, **self.component_overrides[spec.name]) if spec.name in self.component_overrides else spec
for spec in specs
]
class Cancelled(Exception):
pass
def checkpoint(cancel):
if cancel is not None and cancel.is_set():
raise Cancelled("Generation cancelled")
@torch.no_grad()
def encode_prompt(text_encoder, tokenizer, prompt, device):
"""Encode one prompt into LLM2Vec features shaped (1, 1, hidden)."""
from .prompt_utils import tokenize_prompt
from .sanitize import sanitize_texts
inputs = tokenize_prompt(tokenizer, sanitize_texts([prompt])[0])
features = text_encoder(**{name: value.to(text_encoder.device) for name, value in inputs.items()})
return features.to(device)
|