Add SCDD 0.2 public config
Browse files- configs/scdd_pu_0.2.yaml +70 -0
configs/scdd_pu_0.2.yaml
ADDED
|
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
model_id: laaaarrywang/SCDD
|
| 2 |
+
checkpoint: checkpoints/scdd_pu_0.2.ckpt
|
| 3 |
+
|
| 4 |
+
mode: eval
|
| 5 |
+
diffusion: absorbing_state
|
| 6 |
+
backbone: dit
|
| 7 |
+
parameterization: scdd
|
| 8 |
+
time_conditioning: false
|
| 9 |
+
T: 1000
|
| 10 |
+
subs_masking: false
|
| 11 |
+
seed: 512
|
| 12 |
+
|
| 13 |
+
data:
|
| 14 |
+
train: openwebtext-train
|
| 15 |
+
valid: openwebtext-valid
|
| 16 |
+
tokenizer_name_or_path: gpt2
|
| 17 |
+
wrap: true
|
| 18 |
+
streaming: false
|
| 19 |
+
|
| 20 |
+
model:
|
| 21 |
+
name: small
|
| 22 |
+
type: ddit
|
| 23 |
+
hidden_size: 768
|
| 24 |
+
cond_dim: 128
|
| 25 |
+
length: 512
|
| 26 |
+
n_blocks: 12
|
| 27 |
+
n_heads: 12
|
| 28 |
+
scale_by_sigma: true
|
| 29 |
+
dropout: 0.0
|
| 30 |
+
tie_word_embeddings: false
|
| 31 |
+
|
| 32 |
+
forward:
|
| 33 |
+
name: mix
|
| 34 |
+
ratio: 0.2
|
| 35 |
+
gamma: 1
|
| 36 |
+
t_peak: 0.5
|
| 37 |
+
|
| 38 |
+
noise:
|
| 39 |
+
type: loglinear
|
| 40 |
+
sigma_min: 0.0001
|
| 41 |
+
sigma_max: 20
|
| 42 |
+
|
| 43 |
+
sampling:
|
| 44 |
+
predictor: scdd
|
| 45 |
+
compile_sampler: true
|
| 46 |
+
steps: 1024
|
| 47 |
+
noise_removal: true
|
| 48 |
+
nucleus_p: 1.0
|
| 49 |
+
|
| 50 |
+
training:
|
| 51 |
+
ema: 0.9999
|
| 52 |
+
antithetic_sampling: true
|
| 53 |
+
importance_sampling: false
|
| 54 |
+
sampling_eps: 0.001
|
| 55 |
+
change_of_variables: false
|
| 56 |
+
|
| 57 |
+
optim:
|
| 58 |
+
weight_decay: 0.02
|
| 59 |
+
lr: 0.0005
|
| 60 |
+
beta1: 0.9
|
| 61 |
+
beta2: 0.999
|
| 62 |
+
eps: 1.0e-09
|
| 63 |
+
|
| 64 |
+
trainer:
|
| 65 |
+
accelerator: cuda
|
| 66 |
+
devices: 8
|
| 67 |
+
precision: bf16
|
| 68 |
+
gradient_clip_val: 1.0
|
| 69 |
+
max_steps: 1000000
|
| 70 |
+
global_batch_size: 256
|