Project 12 · Deep-learning foundations0/5 completeTensor/Autograd Engine
Shape contracts, matmul and broadcasting, a computation graph, backward pass, and gradient checks — with Python lists only.
- L1Tensor Shape Contractdefine the contract · ~16 minstart →
- L2Matmul + Broadcastingpreserve an invariant · ~22 minlocked
- L3Computation Graphmeasure behavior · ~20 minlocked
- L4Backward Passbound a failure · ~24 minlocked
- L5Gradient Check + Optimizermake an operational decision · ~28 minlocked
Artifacts this project producesL1 · Tensor Shape Contractoutputs: tensor-autograd:l1-tensor-shape:artifact
L2 · Matmul + Broadcastingoutputs: tensor-autograd:l2-matmul-broadcast:artifact
L3 · Computation Graphoutputs: tensor-autograd:l3-computation-graph:artifact
L4 · Backward Passoutputs: tensor-autograd:l4-backward-pass:artifact
L5 · Gradient Check + Optimizeroutputs: tensor-autograd:l5-gradient-check:artifact