Kosmos-X
Kosmos-X copied to clipboard
example is not working, alternatives?
The example seems not working, does anyone have a working code for this model ?
Cell In[22], line 10
7 images.long()
9 # Pass the sample tensors to the model's forward function
---> 10 output = model.forward(text_tokens=text_tokens, images=images)
12 # Print the output from the model
13 print(f"Output: {output}")
File /data/conda/envs/eqbench/lib/python3.10/site-packages/kosmosx/model.py:251, in Kosmos.forward(self, text_tokens, images, **kwargs)
248 raise
250 try:
--> 251 return self.decoder(model_input, passed_x=model_input)[0]
252 except Exception as e:
253 logging.error(f"Failed during model forward pass: {e}")
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torch/nn/modules/module.py:1518, in Module._wrapped_call_impl(self, *args, **kwargs)
1516 return self._compiled_call_impl(*args, **kwargs) # type: ignore[misc]
1517 else:
-> 1518 return self._call_impl(*args, **kwargs)
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torch/nn/modules/module.py:1527, in Module._call_impl(self, *args, **kwargs)
1522 # If we don't have any hooks, we want to skip the rest of the logic in
1523 # this function, and just call forward.
1524 if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks or self._forward_pre_hooks
1525 or _global_backward_pre_hooks or _global_backward_hooks
1526 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1527 return forward_call(*args, **kwargs)
1529 try:
1530 result = None
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torchscale/architecture/decoder.py:399, in Decoder.forward(self, prev_output_tokens, self_attn_padding_mask, encoder_out, incremental_state, features_only, return_all_hiddens, token_embeddings, **kwargs)
387 def forward(
388 self,
389 prev_output_tokens,
(...)
397 ):
398 # embed tokens and positions
--> 399 x, _ = self.forward_embedding(
400 prev_output_tokens, token_embeddings, incremental_state
401 )
402 is_first_step = self.is_first_step(incremental_state)
404 # relative position
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torchscale/architecture/decoder.py:368, in Decoder.forward_embedding(self, tokens, token_embedding, incremental_state)
365 positions = positions[:, -1:]
367 if token_embedding is None:
--> 368 token_embedding = self.embed_tokens(tokens)
370 x = embed = self.embed_scale * token_embedding
372 if positions is not None:
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torch/nn/modules/module.py:1518, in Module._wrapped_call_impl(self, *args, **kwargs)
1516 return self._compiled_call_impl(*args, **kwargs) # type: ignore[misc]
1517 else:
-> 1518 return self._call_impl(*args, **kwargs)
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torch/nn/modules/module.py:1527, in Module._call_impl(self, *args, **kwargs)
1522 # If we don't have any hooks, we want to skip the rest of the logic in
1523 # this function, and just call forward.
1524 if not (self._backward_hooks or self._backward_pre_hooks or self._forward_hooks or self._forward_pre_hooks
1525 or _global_backward_pre_hooks or _global_backward_hooks
1526 or _global_forward_hooks or _global_forward_pre_hooks):
-> 1527 return forward_call(*args, **kwargs)
1529 try:
1530 result = None
File /data/conda/envs/eqbench/lib/python3.10/site-packages/bitsandbytes/nn/modules.py:127, in Embedding.forward(self, input)
126 def forward(self, input: Tensor) -> Tensor:
--> 127 emb = F.embedding(
128 input,
129 self.weight,
130 self.padding_idx,
131 self.max_norm,
132 self.norm_type,
133 self.scale_grad_by_freq,
134 self.sparse,
135 )
137 return emb
File /data/conda/envs/eqbench/lib/python3.10/site-packages/torch/nn/functional.py:2233, in embedding(input, weight, padding_idx, max_norm, norm_type, scale_grad_by_freq, sparse)
2227 # Note [embedding_renorm set_grad_enabled]
2228 # XXX: equivalent to
2229 # with torch.no_grad():
2230 # torch.embedding_renorm_
2231 # remove once script supports set_grad_enabled
2232 _no_grad_embedding_renorm_(weight, input, max_norm, norm_type)
-> 2233 return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)
RuntimeError: Expected tensor for argument #1 'indices' to have one of the following scalar types: Long, Int; but got torch.FloatTensor instead (while checking arguments for embedding)
Upvote & Fund
- We're using Polar.sh so you can upvote and help fund this issue.
- We receive the funding once the issue is completed & confirmed by you.
- Thank you in advance for helping prioritize & fund our backlog.