Kediska Isbaarada Gradient

Jawaabaha iyo sharraxaadaha waxay soo baxayaan markaad qaadanayso kediska.

1. Waa maxay isbaarada gradient ugu horrayn ka ganacsata si loo badbaadiyo xusuusta?
   • Xisaabinta dheeraadka ah (dib u xisaabinta) ee isticmaalka xusuusta hoose
   • Saxnaanta moodeelka xawaaraha
   • Baaxadda diskka ee xawaaraha shabakadda
   • Saxnaanta cabbirka dufcadda
2. Waa maxay sababta firfircooniyaasha sida caadiga ah loo kaydiyo inta lagu jiro kaarka hore?
   • Si loo tuso isticmaalaha
   • Sababtoo ah faafinta dhabarku waxay u baahan yihiin inay xisaabiyaan gradients
   • Si loo dedejiyo rarista xogta
   • Marna dhab ahaantii looma baahna
3. In a simplified chain of N similarly sized layers, what can checkpoint boundaries spaced about sqrt(N) layers apart achieve in the textbook schedule?
   • It stores no activations and performs no recomputation
   • It guarantees constant memory for every model graph
   • Activation storage can scale on the order of sqrt(N) checkpoints
   • It makes optimizer-state memory scale as sqrt(N)
4. What most determines gradient-checkpointing runtime overhead in a practical model?
   • Only the model’s file extension
   • A fixed 20–30% slowdown for every model
   • The number of labels in the dataset
   • Which operations are recomputed, checkpoint placement, graph structure, and hardware
5. Gudaha PyTorch, waa kuwee utility inta badan loo isticmaalo in lagu dabaqo isbaarada gradient module ah?
   • toosh.utils.koontarool
   • toosh.optim.SGD
   • toosh.nn.Dropout
   • toosh.cuda.cache_madhan
6. What happens to operations inside a checkpointed region during backward?
   • Their gradients are skipped entirely
   • Their forward computations are rerun to regenerate needed activations
   • Their weights are permanently discarded
   • They are executed only during inference
7. Which training techniques can be combined with activation checkpointing to reduce different memory costs?
   • Only increasing the number of labels
   • Turning off backpropagation
   • FSDP for sharded model state and mixed precision for lower-precision tensors
   • Replacing the model with a tokenizer
8. What is selective activation checkpointing intended to choose?
   • Which training labels are removed
   • Which model outputs are published
   • Which GPU becomes the tokenizer
   • Which operations to save and which to recompute based on their memory and runtime costs