Text-to-Image
Diffusers
Safetensors
Flux2Pipeline
quantized
mxfp4
autoround
diffusion
autoquant-agent
Instructions to use INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound", 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
- Local Apps Settings
- Draw Things
- DiffusionBee
Download transformer/config.json from INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound: direct link, hf CLI and curl.
- Browser
- Download file 1.12 kB
-
https://huggingface.co/INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound/resolve/main/transformer/config.json
- Command line
-
hf download hf://INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound/transformer/config.json
-
curl -L -o config.json https://huggingface.co/INCModel3/FLUX.2-dev-MXFP4-RTN-AutoRound/resolve/main/transformer/config.json
1.12 kB
| { | |
| "_class_name": "Flux2Transformer2DModel", | |
| "_diffusers_version": "0.36.0.dev0", | |
| "_name_or_path": "/models/FLUX.2-dev/transformer", | |
| "_use_default_values": [ | |
| "guidance_embeds" | |
| ], | |
| "attention_head_dim": 128, | |
| "axes_dims_rope": [ | |
| 32, | |
| 32, | |
| 32, | |
| 32 | |
| ], | |
| "eps": 1e-06, | |
| "guidance_embeds": true, | |
| "in_channels": 128, | |
| "joint_attention_dim": 15360, | |
| "mlp_ratio": 3.0, | |
| "num_attention_heads": 48, | |
| "num_layers": 8, | |
| "num_single_layers": 48, | |
| "out_channels": null, | |
| "patch_size": 1, | |
| "quantization_config": { | |
| "act_bits": 4, | |
| "act_data_type": "mx_fp", | |
| "act_dynamic": true, | |
| "act_group_size": 32, | |
| "act_sym": true, | |
| "autoround_version": "0.15.0", | |
| "bits": 4, | |
| "block_name_to_quantize": "transformer_blocks,single_transformer_blocks", | |
| "data_type": "mx_fp", | |
| "enable_quanted_input": false, | |
| "group_size": 32, | |
| "packing_format": "auto_round:llm_compressor", | |
| "quant_method": "auto-round", | |
| "static_attention_granularity": "tensor", | |
| "static_kv_granularity": "tensor", | |
| "sym": true | |
| }, | |
| "rope_theta": 2000, | |
| "timestep_guidance_channels": 256 | |
| } | |