dalle-mini
dalle-mini copied to clipboard
AttributeError: 'DalleBart' object has no attribute '_is_initialized'
Initialization failure
wandb: You can find your API key in your browser here: https://wandb.ai/authorize
wandb: Paste an API key from your profile and hit enter, or press ctrl+c to quit: ··········
wandb: Appending key for api.wandb.ai to your netrc file: /root/.netrc
wandb: Downloading large artifact wzoooa1c:latest, 1672.79MB. 7 files... Done. 0:0:0
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
[<ipython-input-4-0247091842e9>](https://localhost:8080/#) in <module>()
6 # Load dalle-mini
7 model = DalleBart.from_pretrained(
----> 8 DALLE_MODEL, revision=DALLE_COMMIT_ID, abstract_init=True
9 )
10
3 frames
[/usr/local/lib/python3.7/dist-packages/transformers/modeling_flax_utils.py](https://localhost:8080/#) in params(self, params)
189 def params(self, params: Union[Dict, FrozenDict]):
190 # don't set params if the model is not initialized
--> 191 if not self._is_initialized:
192 raise ValueError(
193 "`params` cannot be set from model when the model is created with `_do_init=False`. "
AttributeError: 'DalleBart' object has no attribute '_is_initialized'
What transformers version do you use?
It may be an issue with the new changes in master
branch.
Edit: oh, I guess you're using the colab which uses master
branch
I think the issue is fixed
TypeError Traceback (most recent call last)
[<ipython-input-4-fa25138b8662>](https://localhost:8080/#) in <module>()
10
11 # Load VQGAN
---> 12 vqgan = VQModel.from_pretrained(VQGAN_REPO, revision=VQGAN_COMMIT_ID)
13
14 # Load CLIP
1 frames
[/usr/local/lib/python3.7/dist-packages/vqgan_jax/modeling_flax_vqgan.py](https://localhost:8080/#) in __init__(self, config, input_shape, seed, dtype, **kwargs)
592 **kwargs,
593 ):
--> 594 module = self.module_class(config=config, dtype=dtype, **kwargs)
595 super().__init__(config,
596 module,
TypeError: __init__() got an unexpected keyword argument '_do_init'
Thanks, it was an issue from another repo but should be fixed now