Instructions to use REPA-E/invae-hf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use REPA-E/invae-hf with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("REPA-E/invae-hf", 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
| { | |
| "_class_name": "AutoencoderKL", | |
| "_diffusers_version": "0.37.0", | |
| "_name_or_path": "ckpts/invae", | |
| "act_fn": "silu", | |
| "block_out_channels": [ | |
| 128, | |
| 128, | |
| 256, | |
| 256, | |
| 512 | |
| ], | |
| "down_block_types": [ | |
| "DownEncoderBlock2D", | |
| "DownEncoderBlock2D", | |
| "DownEncoderBlock2D", | |
| "DownEncoderBlock2D", | |
| "AttnDownEncoderBlock2D" | |
| ], | |
| "force_upcast": true, | |
| "in_channels": 3, | |
| "latent_channels": 32, | |
| "latents_mean": [ | |
| 0.026826243847608566, | |
| 0.9213753342628479, | |
| -0.5369882583618164, | |
| 0.828093409538269, | |
| 0.07783950865268707, | |
| -0.850470781326294, | |
| 0.45318102836608887, | |
| 0.2513081133365631, | |
| 0.7289460301399231, | |
| -0.4415759742259979, | |
| -1.246571660041809, | |
| 1.2389291524887085, | |
| 0.3417016565799713, | |
| 1.0095200538635254, | |
| -0.34630051255226135, | |
| -0.22561031579971313, | |
| 0.6510363817214966, | |
| 0.09759518504142761, | |
| -0.2892760634422302, | |
| 0.8383280038833618, | |
| -0.5023946166038513, | |
| 0.7966985106468201, | |
| -0.5468938946723938, | |
| -0.7595396637916565, | |
| -0.6675211787223816, | |
| 0.03162547945976257, | |
| 0.04389524087309837, | |
| 0.6916733980178833, | |
| -0.6731934547424316, | |
| -0.5851513743400574, | |
| 0.21793237328529358, | |
| 0.13443785905838013 | |
| ], | |
| "latents_std": [ | |
| 3.33929705619812, | |
| 3.452279567718506, | |
| 3.408838987350464, | |
| 3.441635847091675, | |
| 3.581085205078125, | |
| 3.5825204849243164, | |
| 4.7553863525390625, | |
| 3.4237117767333984, | |
| 3.3695008754730225, | |
| 3.2283966541290283, | |
| 3.402665853500366, | |
| 3.7737481594085693, | |
| 4.147014141082764, | |
| 6.795965194702148, | |
| 4.107038974761963, | |
| 3.4266157150268555, | |
| 3.5365779399871826, | |
| 3.1350789070129395, | |
| 3.939121723175049, | |
| 3.7683088779449463, | |
| 3.3776967525482178, | |
| 3.68697452545166, | |
| 3.467242956161499, | |
| 3.3537211418151855, | |
| 3.5815250873565674, | |
| 5.192627906799316, | |
| 3.360060214996338, | |
| 4.357309818267822, | |
| 5.3899335861206055, | |
| 5.052305698394775, | |
| 4.508116245269775, | |
| 3.5253143310546875 | |
| ], | |
| "layers_per_block": 2, | |
| "mid_block_add_attention": true, | |
| "norm_num_groups": 32, | |
| "out_channels": 3, | |
| "sample_size": 32, | |
| "scaling_factor": null, | |
| "shift_factor": null, | |
| "up_block_types": [ | |
| "AttnUpDecoderBlock2D", | |
| "UpDecoderBlock2D", | |
| "UpDecoderBlock2D", | |
| "UpDecoderBlock2D", | |
| "UpDecoderBlock2D" | |
| ], | |
| "use_post_quant_conv": true, | |
| "use_quant_conv": true | |
| } | |