model:
  dataset_name: redwoodresearch/diamonds-seed1
  model_type: codegen
  pretrained_model_name: Salesforce/codegen-350M-mono
  max_length: 1024
  model_config_params:
    sensor_loc_type: locs_from_token
    sensor_token: ' omit'
hparams:
  learning_rate: 2.0e-05
  weight_decay: 0.02
  lr_scheduler_type: cosine
  warmup_steps: 64
  effective_batch_size: 32
  num_train_epochs: 5
per_device_train_batch_size: 4
per_device_eval_batch_size: 4
fp16: true
dataset_len: null
push_to_hub: true