optimum-neuron-cache
/
neuronxcc-2.15.128.0+56dc5a86
/MODULE_1817e58bba6976294b3a
/unet
/config.json
{ | |
"_class_name": "UNet2DConditionModel", | |
"_commit_hash": null, | |
"_diffusers_version": "0.28.2", | |
"_name_or_path": "/home/ubuntu/.cache/huggingface/hub/models--echarlaix--tiny-random-latent-consistency/snapshots/ea16e1fcf89fe667eef6639bfd14ac8701bff7b9/unet", | |
"act_fn": "silu", | |
"addition_embed_type": null, | |
"addition_embed_type_num_heads": 64, | |
"addition_time_embed_dim": null, | |
"attention_head_dim": 8, | |
"attention_type": "default", | |
"block_out_channels": [ | |
4, | |
8 | |
], | |
"center_input_sample": false, | |
"class_embed_type": null, | |
"class_embeddings_concat": false, | |
"conv_in_kernel": 3, | |
"conv_out_kernel": 3, | |
"cross_attention_dim": 32, | |
"cross_attention_norm": null, | |
"down_block_types": [ | |
"DownBlock2D", | |
"CrossAttnDownBlock2D" | |
], | |
"downsample_padding": 1, | |
"dropout": 0.0, | |
"dual_cross_attention": false, | |
"encoder_hid_dim": null, | |
"encoder_hid_dim_type": null, | |
"flip_sin_to_cos": true, | |
"freq_shift": 0, | |
"in_channels": 4, | |
"layers_per_block": 1, | |
"mid_block_only_cross_attention": null, | |
"mid_block_scale_factor": 1, | |
"mid_block_type": "UNetMidBlock2DCrossAttn", | |
"neuron": { | |
"compiler_type": "neuronx-cc", | |
"compiler_version": "2.15.128.0+56dc5a86", | |
"dynamic_batch_size": false, | |
"inline_weights_to_neff": true, | |
"input_names": [ | |
"sample", | |
"timestep", | |
"encoder_hidden_states", | |
"timestep_cond" | |
], | |
"model_type": "unet", | |
"optlevel": "2", | |
"output_attentions": false, | |
"output_hidden_states": false, | |
"output_names": [ | |
"sample" | |
], | |
"static_batch_size": 4, | |
"static_height": 32, | |
"static_num_channels": 4, | |
"static_sequence_length": 77, | |
"static_vae_scale_factor": 2, | |
"static_width": 32 | |
}, | |
"norm_eps": 1e-05, | |
"norm_num_groups": 2, | |
"num_attention_heads": null, | |
"num_class_embeds": null, | |
"only_cross_attention": false, | |
"out_channels": 4, | |
"projection_class_embeddings_input_dim": null, | |
"resnet_out_scale_factor": 1.0, | |
"resnet_skip_time_act": false, | |
"resnet_time_scale_shift": "default", | |
"reverse_transformer_layers_per_block": null, | |
"sample_size": 32, | |
"task": "semantic-segmentation", | |
"time_cond_proj_dim": 32, | |
"time_embedding_act_fn": null, | |
"time_embedding_dim": null, | |
"time_embedding_type": "positional", | |
"timestep_post_act": null, | |
"transformer_layers_per_block": 1, | |
"transformers_version": null, | |
"up_block_types": [ | |
"CrossAttnUpBlock2D", | |
"UpBlock2D" | |
], | |
"upcast_attention": false, | |
"use_linear_projection": false | |
} | |