tensorflow/models · error · ValueError
init_checkpoint_modules=backbone is no longer supported. Spe
Error message
init_checkpoint_modules=backbone is no longer supported. Specify backbone checkpoints in each backbone config.
What it means
Error "init_checkpoint_modules=backbone is no longer supported. Specify backbone checkpoints in each backbone config." thrown in tensorflow/models.
Source
Thrown at official/projects/pix2seq/tasks/pix2seq_task.py:100
drop_units=config.drop_units,
drop_att=config.drop_att,
num_heads=config.num_heads,
temperature=config.temperature,
top_p=config.top_p,
top_k=config.top_k,
early_stopping_token=config.early_stopping_token,
)
return model
def _get_ckpt(self, ckpt_dir_or_file: str) -> str:
if tf.io.gfile.isdir(ckpt_dir_or_file):
return tf.train.latest_checkpoint(ckpt_dir_or_file)
return ckpt_dir_or_file
def initialize(self, model: tf_keras.Model):
"""Loading pretrained checkpoint."""
if self._task_config.init_checkpoint_modules == 'backbone':
raise ValueError(
'init_checkpoint_modules=backbone is no longer supported. Specify'
' backbone checkpoints in each backbone config.'
)
if self._task_config.init_checkpoint_modules not in ['all', 'partial', '']:
raise ValueError(
'Unsupported init_checkpoint_modules: '
f'{self._task_config.init_checkpoint_modules}'
)
if self._task_config.init_checkpoint and any(
[b.init_checkpoint for b in self._task_config.model.backbones]
):
raise ValueError(
'A global init_checkpoint and a backbone init_checkpoint cannot be'
' specified at the same time.'
)
View on GitHub (pinned to e006f5f0d5)
When it happens
Trigger: Thrown at official/projects/pix2seq/tasks/pix2seq_task.py:100 when the library encounters an invalid state.
Common situations: See trigger scenarios.
AI-assisted analysis of tensorflow/models@e006f5f0d5 (2026-08-24).
Data as JSON: /api/errors/d1bce86f5daa68d6.
Report an issue: GitHub.