AttributeError: 'MERTConfig' object has no attribute 'conv_pos_batch_norm'
#1
by
ernestyalumni
- opened
I am obtaining this error:
AttributeError: 'MERTConfig' object has no attribute 'conv_pos_batch_norm'
in
transformers/src/transformers/models/hubert/modeling_hubert.py:275: in init
if config.conv_pos_batch_norm:
self = MERTConfig {
"_attn_implementation_autoset": true,
"activation_dropout": 0.0,
"apply_spec_augment": true,
"arc...ch_dtype": "float32",
"transformers_version": "4.50.0.dev0",
"use_weighted_layer_sum": false,
"vocab_size": 32
}
key = 'conv_pos_batch_norm'
def __getattribute__(self, key):
if key != "attribute_map" and key in super().__getattribute__("attribute_map"):
key = super().__getattribute__("attribute_map")[key]
return super().__getattribute__(key)
E AttributeError: 'MERTConfig' object has no attribute 'conv_pos_batch_norm
when running the example in the README.md.
Python 3.12.3
Pytorch:
torch.version
'2.6.0+cu124'
transformers:
transformers.version
'4.50.0.dev0'