Error while running inference #4
Description
Hi,
I am getting the following error on running the inference.ipynb notebook
TypeError: prepare_attention_mask() takes 3 positional arguments but 4 were given
Traceback:
─────────────────────────────── Traceback (most recent call last) ────────────────────────────────╮
│ in <module>:5 │
│ │
│ 2 │ ps = pm.embed_prompt("a colorful photo of a <yc> in the jungles") │
│ 3 torch.manual_seed(0) │
│ 4 print(ps["CONTEXT_TENSOR_0"].shape) │
│ ❱ 5 im = overwrite_call(pipe, prompt_embeds=ps).images[0] # .save("contents/yc_.jpg") │
│ 6 # im.save("contents/yc_eti_coarse.jpg") │
│ 7 im │
│ 8 │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/autograd/grad_mode.py:27 in │
│ decorate_context │
│ │
│ 24 │ │ @functools.wraps(func) │
│ 25 │ │ def decorate_context(*args, **kwargs): │
│ 26 │ │ │ with self.clone(): │
│ ❱ 27 │ │ │ │ return func(*args, **kwargs) │
│ 28 │ │ return cast(F, decorate_context) │
│ 29 │ │
│ 30 │ def _wrap_generator(self, func): │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/ppp/pipeline_call.py:191 in overwrite_call │
│ │
│ 188 │ │ │ │ # compare the inputs going to this part in this repo vs the baseline SD │
│ 189 │ │ │ │ # need to check if the difference is mainly in how the inputs are given │
│ 190 │ │ │ │ │
│ ❱ 191 │ │ │ │ noise_pred_uncond = self.unet( │
│ 192 │ │ │ │ │ latent_model_input, │
│ 193 │ │ │ │ │ t, │
│ 194 │ │ │ │ │ encoder_hidden_states=negative_prompt_embeds, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/nn/modules/module.py:1110 in _call_impl │
│ │
│ 1107 │ │ # this function, and just call forward. │
│ 1108 │ │ if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks o │
│ 1109 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1110 │ │ │ return forward_call(*input, **kwargs) │
│ 1111 │ │ # Do not call functions when jit is used │
│ 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1113 │ │ if self._backward_hooks or _global_backward_hooks: │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/diffusers/models/unet_2d_condition.py:481 in │
│ forward │
│ │
│ 478 │ │ down_block_res_samples = (sample,) │
│ 479 │ │ for downsample_block in self.down_blocks: │
│ 480 │ │ │ if hasattr(downsample_block, "has_cross_attention") and downsample_block.has │
│ ❱ 481 │ │ │ │ sample, res_samples = downsample_block( │
│ 482 │ │ │ │ │ hidden_states=sample, │
│ 483 │ │ │ │ │ temb=emb, │
│ 484 │ │ │ │ │ encoder_hidden_states=encoder_hidden_states, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/nn/modules/module.py:1110 in _call_impl │
│ │
│ 1107 │ │ # this function, and just call forward. │
│ 1108 │ │ if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks o │
│ 1109 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1110 │ │ │ return forward_call(*input, **kwargs) │
│ 1111 │ │ # Do not call functions when jit is used │
│ 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1113 │ │ if self._backward_hooks or _global_backward_hooks: │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/diffusers/models/unet_2d_blocks.py:789 in │
│ forward │
│ │
│ 786 │ │ │ │ )[0] │
│ 787 │ │ │ else: │
│ 788 │ │ │ │ hidden_states = resnet(hidden_states, temb) │
│ ❱ 789 │ │ │ │ hidden_states = attn( │
│ 790 │ │ │ │ │ hidden_states, │
│ 791 │ │ │ │ │ encoder_hidden_states=encoder_hidden_states, │
│ 792 │ │ │ │ │ cross_attention_kwargs=cross_attention_kwargs, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/nn/modules/module.py:1110 in _call_impl │
│ │
│ 1107 │ │ # this function, and just call forward. │
│ 1108 │ │ if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks o │
│ 1109 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1110 │ │ │ return forward_call(*input, **kwargs) │
│ 1111 │ │ # Do not call functions when jit is used │
│ 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1113 │ │ if self._backward_hooks or _global_backward_hooks: │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/diffusers/models/transformer_2d.py:265 in │
│ forward │
│ │
│ 262 │ │ │
│ 263 │ │ # 2. Blocks │
│ 264 │ │ for block in self.transformer_blocks: │
│ ❱ 265 │ │ │ hidden_states = block( │
│ 266 │ │ │ │ hidden_states, │
│ 267 │ │ │ │ encoder_hidden_states=encoder_hidden_states, │
│ 268 │ │ │ │ timestep=timestep, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/nn/modules/module.py:1110 in _call_impl │
│ │
│ 1107 │ │ # this function, and just call forward. │
│ 1108 │ │ if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks o │
│ 1109 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1110 │ │ │ return forward_call(*input, **kwargs) │
│ 1111 │ │ # Do not call functions when jit is used │
│ 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1113 │ │ if self._backward_hooks or _global_backward_hooks: │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/diffusers/models/attention.py:291 in forward │
│ │
│ 288 │ │ │
│ 289 │ │ # 1. Self-Attention │
│ 290 │ │ cross_attention_kwargs = cross_attention_kwargs if cross_attention_kwargs is not │
│ ❱ 291 │ │ attn_output = self.attn1( │
│ 292 │ │ │ norm_hidden_states, │
│ 293 │ │ │ encoder_hidden_states=encoder_hidden_states if self.only_cross_attention els │
│ 294 │ │ │ attention_mask=attention_mask, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/torch/nn/modules/module.py:1110 in _call_impl │
│ │
│ 1107 │ │ # this function, and just call forward. │
│ 1108 │ │ if not (self._backward_hooks or self._forward_hooks or self._forward_pre_hooks o │
│ 1109 │ │ │ │ or _global_forward_hooks or _global_forward_pre_hooks): │
│ ❱ 1110 │ │ │ return forward_call(*input, **kwargs) │
│ 1111 │ │ # Do not call functions when jit is used │
│ 1112 │ │ full_backward_hooks, non_full_backward_hooks = [], [] │
│ 1113 │ │ if self._backward_hooks or _global_backward_hooks: │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/diffusers/models/cross_attention.py:160 in │
│ forward │
│ │
│ 157 │ │ # The `CrossAttention` class can call different attention processors / attention │
│ 158 │ │ # here we simply pass along all tensors to the selected processor class │
│ 159 │ │ # For standard processors that are defined here, `**cross_attention_kwargs` is e │
│ ❱ 160 │ │ return self.processor( │
│ 161 │ │ │ self, │
│ 162 │ │ │ hidden_states, │
│ 163 │ │ │ encoder_hidden_states=encoder_hidden_states, │
│ │
│ /opt/conda/envs/ldm2/lib/python3.8/site-packages/ppp/utils.py:48 in __call__ │
│ │
│ 45 │ │ print("attention_mask:", attention_mask) │
│ 46 │ │ print("sequence_length: ", sequence_length) │
│ 47 │ │ │
│ ❱ 48 │ │ attention_mask = attn.prepare_attention_mask(attention_mask, sequence_length, ba │
│ 49 │ │ query = attn.to_q(hidden_states) │
│ 50 │ │ │
│ 51 │ │ if _ehs is None: