* start uop emu
* tiny_add passes
* more ops
* emulate the whole warp
* test_gemm passes
* metal gemm test pass
* works on big gemm
* works on big gemm
* more tests pass
* touch ups
* fix mypy
* cleanups
* exp2 mypy
* arch is where it belongs
* actually emulate tensor cores
* fix test
* new style
run on TORCH since it's the fastest one on CI.
caught a bug in multinomial, and update the behavior of fancy index and gather to move the indices Tensor to same device as self.
fix when correction is too big. it seems to only work when input size is 0 though.
torch can output -inf in var when correction is too big, which does not make sense.
* fix Tensor.mean to compute the mean correctly with 0-length axes are selected
* add a regression test
* rename sum variable to sum_t to avoid conflict with built it function
* refactor Tensor.mean to has less lines
* skip matacc opt if the all src buffers of mul op are const buffers
* add noqa directive for long test
* unskip MALACC opt
* ensure that a_axes at least includes summation axes in order to perform np.einsum correctly
* add regression test for mulacc op
* compute a_slices using a_axes
* refactor helper of function to retrieve axes and slices for nonzero strides as well as summation axes
* include a regression test that uses and to test the behaviour indirectly
- removed noop a=0
- fixed integer div test
- added test for both python expression and Tensor method call
- reordered for consistency and added some spaces
* init
* test: added dtype tests for maximum
* fix: seperate maximum const and maximum tensors
* fix: del useless line
* fix: some dtypes
* CODE GOLF: we golfing at mar-a-lago golf club tonight boyyyys
* fix: add lil helper function
* fix: some test refactoring
* done
* sike: not done yet lol
* wtf I missed an assert, am I drunk
* yeah idk
* fix: line save from redundant check
* revert: line save
* fix: simplify test_broadcast cuz I'm stumped
* change some test name
* fix: bool max bool works
* test: add a maximum bool test
* test: make sure minimum also works with bool
* fix: something like this? :s
* fix: maybe this?
* fix: how about this? tighter check
* fix: this.
* revert: nvm mul(0.5) and div(2) has the same kernel for backward
* fix: .is_floating_point() xD
* revert: maximum and minimum and add cast
* fix: cover negative const case in test
* fix: use eq because I don't understand clang :D
* WHOOOOPS
* try
* test: add logical_not tests
* gah im retarded, but this doesn't match types for const()
* fix: can't we jsut do this?
* big change: I don't actually know what I'm doing
* WOOO IM JUST CHANGING EVERYTHING WOW probably gon revert later
* BYE BYE noqa: E501
* fix: less lines and add test
* fix: rm 2 redundant tests
* fix: eq with False so we don't unintentionally implicit upcast, but it's bool anyways so w/e
- removed exact duplicated tests
- only kept one function if torch_fxn is the same as tinygrad_fxn
- used tensor method instead of class method style
- replaced unneeded `lamdba f: f(x)` with just `f`
- re-enabled commented tests that work now
- removed some forward_only now 0 shape tensor can backward
* move reduce over 0 len axis logic to lazy.py
this fixed uneven shard reduce case if the uneven one has length 0
* fix interpreted backends
* fix backwards for 0 shape tensors too
* init
* feat: add _to_const_val to getitem
* doc: changed docs
* docs: updated more docs
* merge: improved/fancy
* better error msg, minor cleanups
* feat: added index_put to test_indexing
* clean: test_indexing
* revert: gather changes lol
* refactor: use dict for tracking tensor indexing, also asserts for type
* oooooooooops
* ugh
* will revert this commit xD
* fix: removed asserts
* improvement: made in-line if statement clearer
* improved err message and improved slice_int tests
* fix: recover accidentally deleted line
* finishing touches
* reword some docs and del torch device tests in test_indexing
* del some redundant tests
* revert: gather asserts, do it in seperate pr
* fix some data_ptr stuff
* done
* done done
* fix no grad fn for < and ==
* remove 2 line breaks
* Remove deprecated autograd variable
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
* fix broadcasted logic if there's 0 in shapes
should always expand into 0, not the other way around. fixed matmul with 0 in input shapes.
for forwards for now though, backward is more involved and would need to change 0 size shortcuts
* fix tests
* onehot in Tensor.py
* one_hot tests
* works for all shapes, not just 1
* pylint
* not a static method
* moved around, num_classes mandatory
* pylint
* pylint
* space & moving
* formatting
* moved tests
* WebGL WIP
* 84% of ops passing test
* tests passing 100%
* Cleanup, refactor
* Shave off some lines
* Work on dtypes
* TestOps at 100% again
* Efficient net shaders compile in browser webgl2
* Compile all efficientnet shaders in browser
* Create empty textures for tensor buffers
* Run program. Up next weight loading
* Exported WebGL model working
* Add tests, refactor
* Explicit cast alu for GLSL
* Fix CI tests
* WebGL efficientnet demo
* Compile and run yolov8 in browser
* Fix imports
* Simplify yolo compile
* Fix bool*bool and cast cmplt to float
* More tests
* Do std tests pass on CI?
* Skip std tests on CI
* Remove explicit_cast_alu hack, and solve it in code_for_op
* Move to new dtype-less alloc api
* Remove local size hack: optimize local_size only if device has local
* Remove glsl.py, and move content to cstyle
* dont_use_locals in opts
* Fix dtype tests
* type_map in CStyleLanguage
* Make core changes smaller, cleaner, refactor export_model and demo
* Skip pad_slice
* Simplify: render_const, render_conditional
* solve bool alu for other binops, cleaner ops_webgl
* Fix noopt hack
* Remove some skipIfs
* WebGL image hack
* type_names is a better name
* global_max
* Fix dtype import
* Fix type_names -> type_map
* Fix lint
* Remove webgpu, back to 5k lines (#3040)
* remove webgpu
* max 5000 lines
* revert those to master
* retain that cstyle
---------
Co-authored-by: Ahmed Harmouche <ahmedharmouche92@gmail.com>
* add Tensor.split (#2677)
* fix mypy errors
* add list support for Tensor.split
* fix ruff comments
* match tensor.split api
* simplify split and test_split
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
* space removal in formula and a single test to cover it
* space in torch einsum as well
* replacing spaces in a var formula to support truncating all the spaces
* lazy rewrite, try 2
* min fix tests
* pass contig test
* put broken pads back
* move that to realize
* no contig child fixes array packing
* so wrong
* now that's correct
* base children
* fix bind issues
* disable to_image_idx
* fix tests
* that failure shouldn't break other tests
* more fixes
* fix torch
* skip failing tests in CI
* 1e-7
* half is broken
* 1e-6 margin of error
