pretrained_model_path: /kaggle/input/my-text2video-1-7b
output_dir: ./outputs
train_data:
  path: /kaggle/input/abc-dataset/trainset
  preprocessed: true
  n_sample_frames: 1
  shuffle_frames: false
  width: 256
  height: 256
  sample_start_idx: 0
  sample_frame_rate: 24
  vid_data_key: video_path
  single_video_path: ''
  single_video_prompt: ''
validation_data:
  prompt: ''
  sample_preview: true
  num_frames: 1
  width: 256
  height: 256
  num_inference_steps: 2
  guidance_scale: 3
extra_train_data: []
dataset_types:
- folder
shuffle: true
validation_steps: 500
trainable_modules:
- attentions
trainable_text_modules: null
extra_unet_params: null
extra_text_encoder_params: null
train_batch_size: 1
max_train_steps: 28040
learning_rate: 1.0e-05
scale_lr: false
lr_scheduler: constant
lr_warmup_steps: 0
adam_beta1: 0.9
adam_beta2: 0.999
adam_weight_decay: 0.01
adam_epsilon: 1.0e-08
max_grad_norm: 1.0
gradient_accumulation_steps: 1
gradient_checkpointing: false
text_encoder_gradient_checkpointing: false
checkpointing_steps: 14020
resume_from_checkpoint: null
resume_step: null
mixed_precision: fp16
use_8bit_adam: false
enable_xformers_memory_efficient_attention: false
enable_torch_2_attn: false
seed: 64
train_text_encoder: false
use_offset_noise: false
rescale_schedule: false
offset_noise_strength: 0.1
extend_dataset: false
cache_latents: false
cached_latent_dir: null
lora_version: stable_lora
save_lora_for_webui: false
only_lora_for_webui: false
lora_bias: none
use_unet_lora: false
use_text_lora: false
unet_lora_modules:
- ResnetBlock2D
text_encoder_lora_modules:
- CLIPEncoderLayer
save_pretrained_model: true
lora_rank: 16
lora_path: ''
lora_unet_dropout: 0.1
lora_text_dropout: 0.1
logger_type: tensorboard
kwargs: {}