all elements in XTHost and ThetaTHost are all Nan

[zhenv5]

I ran it on ml10M dataset. And #define SURPASS_NAN is used to avoid Nan test error.

But all elements in XTHost and ThetaTHost are all NaN. Could you help me figure it out?

Thanks very much. Best.

[wei-tan]

Could you please paste your code here?

[sh1ng]

I'm experiencing the same problem.

make all

As my laptop only has 2 Gb of memory I had to use batches

./main 71567 65133 100 9000048 1000006 0.05 2 2 ./data/ml10M/
M = 71567, N = 65133, F = 100, NNZ = 9000048, NNZ_TEST = 1000006, lambda = 0.050000
X_BATCH = 2, THETA_BATCH = 2
DATA_DIR = ./data/ml10M/ 
*******start loading training and testing sets to host.
*******parameters: m: 71567, n:  65133, f: 100, nnz: 9000048 
*******start allocating memory on GPU...
*******start copying memory to GPU...
*******start iterations...
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
--------- Train RMSE in iter 0: nan
--------- Test RMSE in iter 0: nan
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
--------- Train RMSE in iter 1: nan
--------- Test RMSE in iter 1: nan
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
--------- Train RMSE in iter 2: nan
--------- Test RMSE in iter 2: nan
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
--------- Train RMSE in iter 3: nan
--------- Test RMSE in iter 3: nan
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.
	CG solver with fp32.