tensorflow/models · error
Unsupported converted_model: %s
Error message
Unsupported converted_model: %s
What it means
Error "Unsupported converted_model: %s" thrown in tensorflow/models.
Source
Thrown at official/nlp/tools/tf2_bert_encoder_checkpoint_converter.py:128
# Create a temporary V1 name-converted checkpoint in the output directory.
temporary_checkpoint_dir = os.path.join(output_dir, "temp_v1")
temporary_checkpoint = os.path.join(temporary_checkpoint_dir, "ckpt")
tf1_bert_checkpoint_converter_lib.convert(
checkpoint_from_path=v1_checkpoint,
checkpoint_to_path=temporary_checkpoint,
num_heads=bert_config.num_attention_heads,
name_replacements=(
tf1_bert_checkpoint_converter_lib.BERT_V2_NAME_REPLACEMENTS),
permutations=tf1_bert_checkpoint_converter_lib.BERT_V2_PERMUTATIONS,
exclude_patterns=["adam", "Adam"])
if converted_model == "encoder":
model = _create_bert_model(bert_config)
elif converted_model == "pretrainer":
model = _create_bert_pretrainer_model(bert_config)
else:
raise ValueError("Unsupported converted_model: %s" % converted_model)
# Create a V2 checkpoint from the temporary checkpoint.
tf1_bert_checkpoint_converter_lib.create_v2_checkpoint(
model, temporary_checkpoint, output_path, checkpoint_model_name)
# Clean up the temporary checkpoint, if it exists.
try:
tf.io.gfile.rmtree(temporary_checkpoint_dir)
except tf.errors.OpError:
# If it doesn't exist, we don't need to clean it up; continue.
pass
def main(argv):
if len(argv) > 1:
raise app.UsageError("Too many command-line arguments.")
output_path = FLAGS.converted_checkpoint_pathView on GitHub (pinned to e006f5f0d5)
When it happens
Trigger: Thrown at official/nlp/tools/tf2_bert_encoder_checkpoint_converter.py:128 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/c6007e826b4801f8.
Report an issue: GitHub.