Error while running MultiDCP

[wioxio]

Thanks for the previous answer.
I downloaded github and got the following error when I ran it:

Traceback (most recent call last):
File "../MultiDCP/multidcp_ae.py", line 290, in
model_training(args, model, data, ae_data, metrics_summary)
File "../MultiDCP/multidcp_ae.py", line 82, in model_training
predict, cell_hidden_ = model(input_cell_gex=cell_feature, input_drug = drug,
File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1148, in call_impl
result = forward_call(*input, **kwargs)
File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 382, in forward
return self.perturbed_trans(input_cell_gex, input_drug, input_gene, mask,
File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 389, in perturbed_trans
out, cell_hidden = self.multidcp(input_drug, input_gene, mask, input_cell_gex,
File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1148, in _call_impl
result = forward_call(*input, **kwargs)
File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 198, in forward
pert_idose_embed = self.pert_idose_embed(input_pert_idose)
File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl
return forward_call(*input, **kwargs)
File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 114, in forward
return F.linear(input, self.weight, self.bias)
RuntimeError: mat1 and mat2 shapes cannot be multiplied (64x14 and 6x4)

The train_multidcp_ae.sh was like the following:

CUDA_VISIBLE_DEVICES=3 python ../MultiDCP/multidcp_ae.py --drug_file "/path/MultiDCP/data/data/all_drugs_l1000.csv"
--gene_file "/path/MultiDCP/data/data/gene_vector.csv" --train_file "/path/MultiDCP/data/data/pert_transcriptom/signature_train_cell_1.csv"
--dev_file "/path/MultiDCP/data/data/pert_transcriptom/signature_dev_cell_1.csv" --test_file "/path/MultiDCP/data/data/pert_transcriptom/signature_test_cell_1.csv"
--dropout 0.3 --batch_size 64 --max_epoch 500
--seed 633
--ae_input_file "/path/MultiDCP/data/data/gene_expression_for_ae/gene_expression_combat_norm_978_split4"
--ae_label_file "/path/MultiDCP/data/data/gene_expression_for_ae/gene_expression_combat_norm_978_split4"
--cell_ge_file "/path/MultiDCP/data/data/adjusted_ccle_tcga_ad_tpm_log2.csv"
--all_cells "/path/MultiDCP/data/data/ccle_tcga_ad_cells.p" \

/path/MultiDCP/output/cellwise_output_ran5.txt

Could you guess what could have been possibly wrong?
Thanks.

[AdorableYoyo]

Thanks for the previous answer. I downloaded github and got the following error when I ran it:

Traceback (most recent call last): File "../MultiDCP/multidcp_ae.py", line 290, in model_training(args, model, data, ae_data, metrics_summary) File "../MultiDCP/multidcp_ae.py", line 82, in model_training predict, cell_hidden_ = model(input_cell_gex=cell_feature, input_drug = drug, File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1148, in call_impl result = forward_call(*input, **kwargs) File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 382, in forward return self.perturbed_trans(input_cell_gex, input_drug, input_gene, mask, File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 389, in perturbed_trans out, cell_hidden = self.multidcp(input_drug, input_gene, mask, input_cell_gex, File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1148, in _call_impl result = forward_call(*input, **kwargs) File "/path/MultiDCP/MultiDCP-master/MultiDCP/models/multidcp.py", line 198, in forward pert_idose_embed = self.pert_idose_embed(input_pert_idose) File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1130, in _call_impl return forward_call(*input, **kwargs) File "/path/MultiDCP/lib/python3.8/site-packages/torch/nn/modules/linear.py", line 114, in forward return F.linear(input, self.weight, self.bias) RuntimeError: mat1 and mat2 shapes cannot be multiplied (64x14 and 6x4)

The train_multidcp_ae.sh was like the following:

CUDA_VISIBLE_DEVICES=3 python ../MultiDCP/multidcp_ae.py --drug_file "/path/MultiDCP/data/data/all_drugs_l1000.csv" --gene_file "/path/MultiDCP/data/data/gene_vector.csv" --train_file "/path/MultiDCP/data/data/pert_transcriptom/signature_train_cell_1.csv" --dev_file "/path/MultiDCP/data/data/pert_transcriptom/signature_dev_cell_1.csv" --test_file "/path/MultiDCP/data/data/pert_transcriptom/signature_test_cell_1.csv" --dropout 0.3 --batch_size 64 --max_epoch 500 --seed 633 --ae_input_file "/path/MultiDCP/data/data/gene_expression_for_ae/gene_expression_combat_norm_978_split4" --ae_label_file "/path/MultiDCP/data/data/gene_expression_for_ae/gene_expression_combat_norm_978_split4" --cell_ge_file "/path/MultiDCP/data/data/adjusted_ccle_tcga_ad_tpm_log2.csv" --all_cells "/path/MultiDCP/data/data/ccle_tcga_ad_cells.p" \

/path/MultiDCP/output/cellwise_output_ran5.txt

Could you guess what could have been possibly wrong? Thanks.

Hi, it is fixed now.
Please pull the repo to the latest version.

[wioxio]

Thank you