AI4Finance-Foundation / RLSolver

Solvers for NP-hard and NP-complete problems with an emphasis on high-performance GPU computing.

Home Page:https://ai4finance.org

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

🐛 may be it is a bug

Yonv1943 opened this issue · comments

this line in env.reset(), should be self.K * self.N, instead of self.K * self.K

vec_W = th.randn((self.mat_H.shape[0], self.K* self.K), dtype=th.cfloat, device=self.device)
vec_W = vec_W / th.norm(vec_W, dim=1, keepdim=True)
self.mat_W = vec_W.reshape(-1, self.K, self.N)