I think "trying to use a framework" is approaching it from the wrong end. I find it more enlightening to try to achieve something larger without a framework, seeing what pain points you have and then adopting whatever you need to solve your problems.
In some cases, "trying" might be actually succeeding, but finding out that the same could be done with 1/100 of the time/effort. In other cases, you just find out what you want to do is just not feasible as a single developer and leveraging others' work is a must.
That approach also makes it much easier to evaluate if the framework is good and is good for you and your project. Otherwise you're trying to figure out if a hammer is good without having any nails to drive.
It's true, not using a framework can be very enlightening, it shows the breadth of what you have to implement yourself.
It can also sometimes work the other way, and show that the framework is not actually that useful. I had this with Ansible, where going down to raw fabric made my life much, much nicer (yaml is not a great programming language...).
In some cases, "trying" might be actually succeeding, but finding out that the same could be done with 1/100 of the time/effort. In other cases, you just find out what you want to do is just not feasible as a single developer and leveraging others' work is a must.
That approach also makes it much easier to evaluate if the framework is good and is good for you and your project. Otherwise you're trying to figure out if a hammer is good without having any nails to drive.