Unconditional Image Generation
Transformers
Safetensors
tinyimagegen
feature-extraction
imagegen
unconditional-image
custom_code
Instructions to use fromziro/TinyImageGen-0.6M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use fromziro/TinyImageGen-0.6M with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("fromziro/TinyImageGen-0.6M", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "architectures": [ | |
| "TinyImageGenModelForImageDiffusion" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_tinyimagegen.TinyImageGenConfig", | |
| "AutoModel": "modeling_tinyimagegen.TinyImageGenModelForImageDiffusion", | |
| "AutoModelForImageDiffusion": "modeling_tinyimagegen.TinyImageGenModelForImageDiffusion" | |
| }, | |
| "dtype": "float32", | |
| "head_dim": 24, | |
| "hidden_size": 96, | |
| "image_size": 32, | |
| "in_channels": 3, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 160, | |
| "model_type": "tinyimagegen", | |
| "num_attention_heads": 4, | |
| "num_hidden_layers": 6, | |
| "num_key_value_heads": 2, | |
| "num_lanes": 4, | |
| "num_patches": 256, | |
| "num_patches_side": 16, | |
| "patch_dim": 12, | |
| "patch_size": 2, | |
| "rms_norm_eps": 1e-05, | |
| "rope_theta": 2500.0, | |
| "swiglu_interval": 3, | |
| "transformers_version": "5.8.0.dev0", | |
| "use_cache": false, | |
| "use_per_head_gating": false, | |
| "use_xsa": false | |
| } |