-
Are there any optimizers that anyone would like to see implemented? In particular optimizers for differentiable functions.
@zoq any ideas or requests? @rcurtin has suggested Nesterov's Accelerated Gr…
-
You guys have probably thought about this but I was wondering about trying the following algorithms out in the optimisation side of things,
- stochastic gradient descent with/without momentum or wi…
-
Do you have any examples of using the chain rule, for (automatic) differentiation?
-
Hello, My computer is lenovo YOGA 910, the CPU is i7-7500U, My system is Ubuntu 16.04.03, Opencv 3.3, I'm following the step to install :http://www.ceres-solver.org/installation.html#linux ;
it is:
…
-
Hi,
I am convinced that we should go to 2nd order methods, those which actively repel saddles like SFN, but it seems they are not used (?) nearly 5 years after the paper.
Do you maybe know what is t…
-
When using the lobpcg implementation in `scipy.sparse.linalg` with scipy version `1.3.0` on a preconditioned, normal eigenproblem, the algorithm aborts with the error message shown below if a particul…
-
I find `Zygote` supports complex number AD, I like it more now! e.g.
```julia console
julia> gradient(imag, x)
((re = nothing, im = 1),)
```
But I have several questions,
## Question 1:
Why it …
-
I think in fixed learning rate algorithm, it's better to have a momentum option to accelerate training.
-
1) with conjugate gradient descent, there are two problems: the EpsTerminationCondition would stop the iterations due to little differences; removing the eps termination condition, the scores are not …
-
Here is the beginning of the training log.
/root/anaconda3/envs/psenet/bin/python3.5 /data3/xxx/tensorflow_PSENet/train.py --gpu_list=0,1 --input_size=512 --batch_size_per_gpu=4 --checkpoint_path=.…