Repository navigation
Add suppor for running MNIST sample app #2
Description
Support MNIST
To implement MNIST sample, new types of layers in skainet is required
sk-ai-net/skainet-samples/issues/1
To implement MNIST sample, new types of layers in skainet is required
sk-ai-net/skainet-samples/issues/1
Activity
Merge pull request #2 from skainet-dev/codex/add-edge-test-cases-for-…
Adds multiple changes:
Initial version of MLP based MNIST is working
The first working version of MNIST classier using MLP pretrained model.
Use skaneit release version.
Add initial implementation of compute graph.
Add shepe ops for yolo8
1 remaining item
Add shepe ops for yolo8
Add initial implementaion of addition in Cpu TensorOps
Improve test coverage
Fix failing tests
Add initial implementaion of addition in Cpu TensorOps
Improve on matmul implementaion.
Improve on ExecuteCOntext implementaion with nn.
Add DSL support to conv2d and maxpool.
Add for maxpool.
The first working version of MNIST classier using MLP pretrained model.
Use skaneit release version.
Update roadmap: mark Phase 1 as completed
michalharakal commentedon Jun 10, 2026
Closing as implemented. MNIST sample app ships via
skainet-lang-models/.../dnn/mlp/MNIST.ktplus the MNIST data loaders in skainet-data.