George Hotz 5e07d4669d the speedy chonker is going to replace the old chonker (#432) ...

* bringing back reshape and permute

* done with E701

* 4x4 works in generic way

* max and sum not vectorizing...

* special case single float

* support comparing to MPS

* improve matmul speed, consider generic principles

* GlobalCounter

* fix op tracking

* faster

* comment that out for now

* err, it needs that

* fix minor issues

* fix global_mem

2022-11-11 18:34:24 -08:00

..

llops

the speedy chonker is going to replace the old chonker (#432)

2022-11-11 18:34:24 -08:00

nn

Fix issue where batch_invstd not being set (#421)

2022-11-08 09:24:53 -08:00

__init__.py

tinygrad.nn (#367)

2022-08-18 07:41:00 -07:00

graph.py

hasattr and DeviceBuffer type fixups

2022-10-28 09:05:45 -07:00

helpers.py

updates from the chonker branch

2022-11-07 21:12:08 -08:00

lazy.py

s/contiguous_op/contiguous

2022-11-11 00:07:05 -08:00

mlops.py

s/contiguous_op/contiguous

2022-11-11 00:07:05 -08:00

ops.py

the speedy chonker is going to replace the old chonker (#432)

2022-11-11 18:34:24 -08:00

shapetracker.py

factorizing shapetracker from chonker

2022-11-09 16:36:38 -08:00

tensor.py

Reduce line count (#424)

2022-11-09 10:07:22 -08:00