1: \begin{abstract} 

2: We propose a conditional gradient framework for a composite convex minimization template with broad applications. 

3: Our approach combines the notions of smoothing and homotopy under the CGM framework, and provably achieves the optimal $\mathcal{O}(1/\sqrt{k})$ convergence rate. 

4: We demonstrate that the same rate holds if the linear subproblems are solved approximately with additive or multiplicative error.  

5: Specific applications of the framework include the non-smooth minimization, semidefinite programming, and minimization with linear inclusion constraints over a compact domain. 

6: We provide numerical evidence to demonstrate the benefits of the new framework. 

7: \end{abstract}