Hugging Face's logo

pcuenq
/
mdm-flickr-64

mdm
Model card Files Community
2
mdm-flickr-64
File size: 2,220 Bytes 
c3f64df
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
 
name: cc12m_64x64
dataset_config: configs/datasets/cc12m.yaml

#evaluator_arguments:
batch_size: 32
num_eval_batches: 500
sample_image_size: 64
test_file_list: validation.tsv
reader_config_file: launch_scripts/reader/latest_eval.yaml

# sampler_arguments:
min_examples: 10000
sample_dir: /mnt/data/samples
batch_size: 32
sample_image_size: 64
test_file_list: validation.tsv
device: cuda
model: unet

output_dir: /mnt/data/outputs
num_diffusion_steps: 1000
reproject_signal: false
predict_variances: false
model_output_scale: 0
prediction_type: V_PREDICTION
loss_target_type: HA_STYLE
schedule_type: DEEPFLOYD
prediction_length: 129
use_vdm_loss_weights: false
loss_factor: 1
num_training_steps: 5000
num_epochs: 20000
avg_lm_steps: 0
use_lm_mask: 1
categorical_conditioning: 0
vocab_file: t5.vocab
text_model: google/flan-t5-xl
vision_model: unet

unet_config:
  num_resnets_per_resolution: [2, 2, 2]
  attention_levels: [1, 2]
  num_attention_layers: [0, 1, 5]
  conditioning_feature_dim: -1
  conditioning_feature_proj_dim: 2048
  num_lm_head_layers: 0
  masked_cross_attention: 0
  resolution_channels: [256, 512, 768]
  skip_mid_blocks: False
  skip_cond_emb: False
  nesting: False
  micro_conditioning: 'scale:64'
  temporal_mode: False
  temporal_spatial_ds: False
  temporal_positional_encoding: False
  resnet_config:
    num_channels: -1
    output_channels: -1
    num_groups_norm: 32
    dropout: 0.0
    use_attention_ffn: True
diffusion_config:
  sampler_config:
    num_diffusion_steps: 1000
    reproject_signal: False
    schedule_type: DEEPFLOYD
    prediction_type: V_PREDICTION
    loss_target_type: DDPM
    beta_start: 0.0001
    beta_end: 0.02
    threshold_function: CLIP
    rescale_schedule: 1.0
    schedule_shifted: False
  model_output_scale: 0.0
  use_vdm_loss_weights: False

reader_config:
  reader_buffer_size: 500
  shuffle_buffer_size: 500
  image_size: 64
  smaller_side_size: 64
  random_crop: false
  max_caption_length: 512
  max_token_length: 128
  num_readers: 16

metrics: fid,clip
batch_size: 32
gradient_clip_norm: 2
num_gradient_accumulations: 1
warmup_steps: 10000
use_adamw: true
log_freq: 50
save_freq: 5000
lr: 5.0e-05
fp16: 0
use_precomputed_text_embeddings: 0
seed: -1