* Metal quantized modifications proposal.
- Add a device param, wherever needed.
- Create new QMetal storage thing that implements QuantizedType.
- Update everywhere needed.
Fix Python.
Fixing examples.
Fix: fmt + clippy + stub.
Moving everything around.
Only missing the actual implems.
Fixing everything + adding dequantized kernels.
More work.
Fixing matmul.
Fmt + Clippy
Some clippy fixes.
Working state.
Q2K Metal -> Bugged (also present in GGML).
Q4K CPU -> Bugged (present previously, new test catch it).
Q5K CPU -> Bugged (present previously).
Q8_1 Both -> Never really implemented it seems
Q8K metal -> Never implemented in metal
Fixing Q2K bug (present in ggml).
* Cleanup.
* Fix the rebase.
* Removing the fences speeds everything up and *is* correct this time...
* Cleanup the fence.
* After rebase.
* Bad code removal.
* Rebase after phi2 merge + fix replit default to CPU.
* Making the CI happy.
* More happy tests.
---------
Co-authored-by: Nicolas Patry <nicolas@Nicolass-MacBook-Pro.local>
* Only optimize float tensors.
* Use full tensors for zeros and ones.
* Add a benchmark for the matmul slowness.
* Add the convmixer model.
* Proper adaptive pooling.
* im2col implementation for conv2d.
* Fix for the im2col implementation to match the current conv2d.
* Small optimization.
* Add a cuda kernel.
* Handle arbitrary layouts.
* Im2Col cuda code.
* Add a custom softmax implementation.
* Add softmaxlastdim to the benchmarks.
* And add a test.
* Support more dtypes.
* Polish the code.
* Use the slow implementation on cuda.
* Add a todo for the cuda kernel.
* Add a couple functions required for yolo.
* Add the yolo-v3 example.
* Add minimum and maximum.
* Use the newly introduced maximum.
* Cuda support for min/max + add some testing.
* Allow for more tests to work with accelerate.
* Fix a typo.
