-
I am encountering issues when using non-element-wise optimizers such as Adam-mini with DeepSpeed.
According to the documentation, it reads:
> The FP16 Optimizer is designed to maximize the achievable…
-
After installing apex 0.1, got : RuntimeError: apex.optimizers.FusedLAMB requires cuda extensions
raise RuntimeError('apex.optimizers.FusedLAMB requires cuda extensions')
RuntimeError: apex.o…
-
Prior to #725 we could specify something like this in a reusable optimizer config:
```jsonc
// MLOS Core optimizer
{
"class": "mlos_bench.optimizers.MlosCoreOptimizer",
// Note: these…
-
### Feature request
log train loss on start
----
’m using the Hugging Face `Trainer` (or `SFTTrainer`) for fine-tuning, and I want to log the training loss at step 0 (before any training steps …
-
### Is your feature request related to a problem? Please describe.
In gh-16191, it was suggested that `curve_fit` might benefit from an option to use global optimization.
### Describe the solution …
-
Sub-ticket of #3018.
Here is a list of these optimizers:
- ShapeOptimizer
- UnShapeOptimizer
- [ ] FusionOptimizer
Only the last one has to be worked on (the other ones are "fake" optimizations that…
-
When generating the Gaussian in GSplat, I observed a misalignment with the initial point cloud. The misalignment is noticeable in terms of axes, position, and scale. It does not appear to be a simple …
-
### Bug description
Hello,
I encountered a bug when training with `automatic_optimization = False` and two optimizers.
In summary: the `global_step` attribute of the trainer and the lightning…
jkyl updated
8 months ago
-
This bug is in the latest released version of DeepChem (2.8.0).
The ExponentialDecay learning rate scheduler is mis-implemented in the code for TensorFlow. The platform-specific learning rate sched…
-
This is more likely a feature request（if not available now）.
As widely known that there're multiple minimas when **EI** , **LCB**, **PI** infill plan are used.
From what I can get in this code, i…