---------------------------------------------------------------------------
TypeError Traceback (most recent call last)
Cell In[3], line 14
12 # model_name = "mistralai/Mistral-7B-v0.3"
13 model_name = "/home/ubisec/swh/models/deepseek-ai-DeepSeek-R1-0528-Qwen3-8B"
---> 14 model, tokenizer, lm_head = load_model_and_tokenizer(model_name, type = "curlora")
File ~/swh/codes/swh-curlora/code/utils.py:92, in load_model_and_tokenizer(model_name, type)
90 replace_linear_with_lora(model, rank=16, alpha=1)
91 else:
---> 92 replace_linear_with_curlora(model, rank=16, alpha=1)
94 total_params = sum(p.numel() for p in model.parameters() if p.requires_grad)
95 print(f"Total trainable parameters after: {total_params:,}")
File ~/swh/codes/swh-curlora/code/utils.py:65, in replace_linear_with_curlora(model, rank, alpha)
63 setattr(model, name, LinearWithCURLoRA(module, rank, alpha))
64 else:
---> 65 replace_linear_with_curlora(module, rank, alpha)
File ~/swh/codes/swh-curlora/code/utils.py:65, in replace_linear_with_curlora(model, rank, alpha)
63 setattr(model, name, LinearWithCURLoRA(module, rank, alpha))
64 else:
---> 65 replace_linear_with_curlora(module, rank, alpha)
[... skipping similar frames: replace_linear_with_curlora at line 65 (1 times)]
File ~/swh/codes/swh-curlora/code/utils.py:65, in replace_linear_with_curlora(model, rank, alpha)
63 setattr(model, name, LinearWithCURLoRA(module, rank, alpha))
64 else:
---> 65 replace_linear_with_curlora(module, rank, alpha)
File ~/swh/codes/swh-curlora/code/utils.py:63, in replace_linear_with_curlora(model, rank, alpha)
60 for name, module in model.named_children():
61 #if isinstance(module, torch.nn.Linear):
62 if any(l in name for l in ["q_proj", "v_proj", "k_proj"]):
---> 63 setattr(model, name, LinearWithCURLoRA(module, rank, alpha))
64 else:
65 replace_linear_with_curlora(module, rank, alpha)
File ~/swh/codes/swh-curlora/code/curlora.py:105, in LinearWithCURLoRA.__init__(self, linear, rank, alpha)
103 super().__init__()
104 self.linear = linear
--> 105 self.curlora = CURModule(linear.weight, rank)
106 self.rank = rank
107 self.alpha = alpha
File ~/swh/codes/swh-curlora/code/curlora.py:63, in CURModule.__init__(self, W, rank)
61 def __init__(self, W, rank):
62 super(CURModule, self).__init__()
---> 63 C, U, R = cur_decomposition(W, rank)
64 self.C = C * 1.0
65 self.R = R * 1.0
File ~/swh/codes/swh-curlora/code/curlora.py:48, in cur_decomposition(A, c)
46 r = c
47 column_probs, row_probs = compute_selection_probabilities(A)
---> 48 selected_columns = select_indices_with_replacement(column_probs, c)
49 selected_rows = select_indices_with_replacement(row_probs, r)
51 C = A[:, selected_columns]
File ~/swh/codes/swh-curlora/code/curlora.py:25, in select_indices_with_replacement(probs, k)
23 probs = probs.cpu().numpy()
24 else:
---> 25 probs = probs.numpy()
27 return np.random.choice(len(probs), size=k, replace=True, p=probs)
TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.
