Approximations for SIN/LOG2/EXP2 passing all tests. (#5187)
* [WIP] Added an approximated implementation of Sin(FP32, FP64) passing all tests on Clang runtime
* Map nan/-inf/inf as 1.0 in order to avoid doing as_const(math.inf)
* [WIP] Added a support for LLVM IR
* cleaned up the code for the mypy and linter
* [WIP] Updated fp64 supports (bitwise shift causes the compilation error), fixed linter issue.
* [Add] added fast=true mode which disables the payne-hanek reduction which is slow
* [Fix] fails to compute elements when shape includes zero
* [WIP] Added BinaryOps.ADD/BinaryOps.OR to assembly
* [wip] update the assembly for ptx
* Enables fast=True when device is one of PTX, NV, CUDA, to avoid slow bitwise ops (as lv3 reduction is not required).
* [WIP] Added an approximation of LOG2/EXP2 (FP32, FP64)
* [Fix] Cyclic dependencies existing in xlog2
* [Fix] Cycle dependency in the graph of exp2, and log2. (passing test_symbolic_ops.py)
* [Fix] keep using higher precision for exp2, but cycle graph issue remained to be fixed...
* [Refactor] removed is_metal option. xsin does not rely on fp64 when fp32 mode.
* [WIP] fp16 xsin implementation passing all tests. (still needs to be refactored)
* [WIP] Added fp16 exp2 implementation
* [WIP] Increased the precision of Log2 from 3.5 ULP to 1.0 ULP, and added FP16 Log2 approximation.
* stashed the changes for FP16 sin
* [Fix] Patch for FP16 Sin/Exp2. (updated the dtype_via, fp32_p, and lower)
* [Refactor] migration to fastmath.py, some code simplification, renamed apis in fastmath, et al.
* [Refactor] Added the function polyN to clean-up N-terms polynomial approximation.
* [Patch] Increase fp64 precision when ldexp3k if possible, and patch for fp16 exp2
* [Patch] added bitcast_forward option
* [Patch] resolved cycle graph
* patch fix cycle graph
* set bitcast_forward=True in ilogb2k
* bitcast_forward for multi.py
* E501
* Break into multiple small PRs
* [Patch] FP16 -> FP64 upcast is not anymore required since xlog2 use quad precision polyN
* [Patch] NV still required FP64 for xlog2
* updated schedule test
* updated the count of kernels
* [Update] Removed all bitwise ops (SHL/SHR), tweaked the nan manipulation of log2, passing all tests except for AMD.
* Bitcast: make them api-compatible
* [update] force to use bitcast
* updated the count of constant folding
* [Patch] Creating a mask for exp2 using x <= Inf satisfies True as long as x is a real value
* [Update] isNaN(x) Free log2 algorithm, passing PTX tests, METAL with fastmath enabled is able to handle nan well, amd backend will not crash.
* xsin is reluctant to call payne_hanek_reduction which is slow to compile, passing stable diffusion compilation in a realistic time
* some minor simplification to payne hanek reduction
* [refactor] refactored some rebundant parts existing in payne hanek
* [refactor] more readable payne hanek impl
* [refactor] improved the code consistency of payne hanek
* [experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)
* Revert "[experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)"
This reverts commit 0eee08b87c9e46da8aec0a8edec5316634031a49.
* use allow_buffer_view
* lets support multilazytensor
* updated the count of kernels
* [test] added the jit tests for approx ops
* keep failed constant folding tests tested, added expectedFailure
* explict the timeout deadline when testing approx jit timeout
* [WIP] Simplified the implementation of xsin, never timeouts
* [Refactor] Improved the consistency of approx sin implementation, passing time out tests
* integrated xexp2_base into xexp2
* Set switch_over=39800.0
* delete: is_buffer_fastmath_supported
* sin: compute against abs(x)
* some cleanups
* fix typo
* removed the space between param and dtype
* allow 514 kernels on CI for sd
* [refactor] no need to upcast ad ldexp3k
* [refactor] added some comments, references to help understanding the code.
* [Fix] 1.0 ULP Sine Approximation for FP16
* [update] assume e != 0
* use pow2if instead of ldexp3k to fuse payne_hanek reduction into one
* check if approximated sin/log2/exp are fused into one
* clean up changes
* test amd exp
* some code cleanup and test sigmoid
* fix: enabled payne_hanek for fp16 to achieve higher acc
* fix: payne_hanek always accumlates the value with uint64, and fp16 sin is fused to a single kernel
* [Refactor] Rename: fastmath -> transcendental
* [Refactor] Added TRANSCENDENTAL, Moved the gate function to function.py
* updated const folding tests
* TRANSCENDENTAL as a ContextVar, removed old test of cody waite reduction, added assertions, et al.
* Add: unittest.main()
* Import TRANSCENDENTAL instead of getenv
* Refactor: Added dtype check when TRANSCENDENTAL=2, more context var
* Patch: xlog2, break expt(2, 32) x 2 -> expt(2, 16) x 4 for fp16 math
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
Co-authored-by: chenyu <chenyu@fastmail.com>
2024-07-11 07:44:58 +08:00
|
|
|
import unittest
|
2024-07-13 13:54:52 +08:00
|
|
|
from tinygrad import Tensor, Device, dtypes
|
|
|
|
|
from tinygrad.tensor import _to_np_dtype
|
|
|
|
|
from tinygrad.helpers import Context, getenv
|
Approximations for SIN/LOG2/EXP2 passing all tests. (#5187)
* [WIP] Added an approximated implementation of Sin(FP32, FP64) passing all tests on Clang runtime
* Map nan/-inf/inf as 1.0 in order to avoid doing as_const(math.inf)
* [WIP] Added a support for LLVM IR
* cleaned up the code for the mypy and linter
* [WIP] Updated fp64 supports (bitwise shift causes the compilation error), fixed linter issue.
* [Add] added fast=true mode which disables the payne-hanek reduction which is slow
* [Fix] fails to compute elements when shape includes zero
* [WIP] Added BinaryOps.ADD/BinaryOps.OR to assembly
* [wip] update the assembly for ptx
* Enables fast=True when device is one of PTX, NV, CUDA, to avoid slow bitwise ops (as lv3 reduction is not required).
* [WIP] Added an approximation of LOG2/EXP2 (FP32, FP64)
* [Fix] Cyclic dependencies existing in xlog2
* [Fix] Cycle dependency in the graph of exp2, and log2. (passing test_symbolic_ops.py)
* [Fix] keep using higher precision for exp2, but cycle graph issue remained to be fixed...
* [Refactor] removed is_metal option. xsin does not rely on fp64 when fp32 mode.
* [WIP] fp16 xsin implementation passing all tests. (still needs to be refactored)
* [WIP] Added fp16 exp2 implementation
* [WIP] Increased the precision of Log2 from 3.5 ULP to 1.0 ULP, and added FP16 Log2 approximation.
* stashed the changes for FP16 sin
* [Fix] Patch for FP16 Sin/Exp2. (updated the dtype_via, fp32_p, and lower)
* [Refactor] migration to fastmath.py, some code simplification, renamed apis in fastmath, et al.
* [Refactor] Added the function polyN to clean-up N-terms polynomial approximation.
* [Patch] Increase fp64 precision when ldexp3k if possible, and patch for fp16 exp2
* [Patch] added bitcast_forward option
* [Patch] resolved cycle graph
* patch fix cycle graph
* set bitcast_forward=True in ilogb2k
* bitcast_forward for multi.py
* E501
* Break into multiple small PRs
* [Patch] FP16 -> FP64 upcast is not anymore required since xlog2 use quad precision polyN
* [Patch] NV still required FP64 for xlog2
* updated schedule test
* updated the count of kernels
* [Update] Removed all bitwise ops (SHL/SHR), tweaked the nan manipulation of log2, passing all tests except for AMD.
* Bitcast: make them api-compatible
* [update] force to use bitcast
* updated the count of constant folding
* [Patch] Creating a mask for exp2 using x <= Inf satisfies True as long as x is a real value
* [Update] isNaN(x) Free log2 algorithm, passing PTX tests, METAL with fastmath enabled is able to handle nan well, amd backend will not crash.
* xsin is reluctant to call payne_hanek_reduction which is slow to compile, passing stable diffusion compilation in a realistic time
* some minor simplification to payne hanek reduction
* [refactor] refactored some rebundant parts existing in payne hanek
* [refactor] more readable payne hanek impl
* [refactor] improved the code consistency of payne hanek
* [experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)
* Revert "[experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)"
This reverts commit 0eee08b87c9e46da8aec0a8edec5316634031a49.
* use allow_buffer_view
* lets support multilazytensor
* updated the count of kernels
* [test] added the jit tests for approx ops
* keep failed constant folding tests tested, added expectedFailure
* explict the timeout deadline when testing approx jit timeout
* [WIP] Simplified the implementation of xsin, never timeouts
* [Refactor] Improved the consistency of approx sin implementation, passing time out tests
* integrated xexp2_base into xexp2
* Set switch_over=39800.0
* delete: is_buffer_fastmath_supported
* sin: compute against abs(x)
* some cleanups
* fix typo
* removed the space between param and dtype
* allow 514 kernels on CI for sd
* [refactor] no need to upcast ad ldexp3k
* [refactor] added some comments, references to help understanding the code.
* [Fix] 1.0 ULP Sine Approximation for FP16
* [update] assume e != 0
* use pow2if instead of ldexp3k to fuse payne_hanek reduction into one
* check if approximated sin/log2/exp are fused into one
* clean up changes
* test amd exp
* some code cleanup and test sigmoid
* fix: enabled payne_hanek for fp16 to achieve higher acc
* fix: payne_hanek always accumlates the value with uint64, and fp16 sin is fused to a single kernel
* [Refactor] Rename: fastmath -> transcendental
* [Refactor] Added TRANSCENDENTAL, Moved the gate function to function.py
* updated const folding tests
* TRANSCENDENTAL as a ContextVar, removed old test of cody waite reduction, added assertions, et al.
* Add: unittest.main()
* Import TRANSCENDENTAL instead of getenv
* Refactor: Added dtype check when TRANSCENDENTAL=2, more context var
* Patch: xlog2, break expt(2, 32) x 2 -> expt(2, 16) x 4 for fp16 math
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
Co-authored-by: chenyu <chenyu@fastmail.com>
2024-07-11 07:44:58 +08:00
|
|
|
from test.test_schedule import check_schedule
|
2024-07-13 13:54:52 +08:00
|
|
|
from test.test_dtype_alu import ht
|
|
|
|
|
from test.helpers import is_dtype_supported
|
|
|
|
|
import numpy as np
|
|
|
|
|
from hypothesis import given, settings, strategies as strat
|
|
|
|
|
|
|
|
|
|
settings.register_profile("my_profile", max_examples=200, deadline=None, derandomize=getenv("DERANDOMIZE_CI", False))
|
|
|
|
|
settings.load_profile("my_profile")
|
|
|
|
|
|
|
|
|
|
class TestTranscendentalMath(unittest.TestCase):
|
|
|
|
|
@unittest.skipUnless(is_dtype_supported(dtypes.float64, Device.DEFAULT), f"no float64 on {Device.DEFAULT}")
|
2024-08-05 04:06:38 +08:00
|
|
|
@unittest.skipIf(getenv("MOCKGPU") and Device.DEFAULT == "NV", "crashed")
|
2024-07-13 13:54:52 +08:00
|
|
|
@given(ht.float64, strat.sampled_from([(Tensor.exp, np.exp), (Tensor.log, np.log), (Tensor.sin, np.sin)]))
|
|
|
|
|
def test_float64(self, x, op):
|
|
|
|
|
if op[0] == Tensor.sin:
|
2024-07-13 15:31:21 +08:00
|
|
|
# TODO: reduction does not work # 536870912.125 # 2914593.01171875 # 134217728.03125
|
|
|
|
|
if abs(x) > 536870912: return
|
2024-07-13 13:54:52 +08:00
|
|
|
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
np.testing.assert_allclose(op[0](Tensor([x], dtype=dtypes.float64)).numpy(),
|
|
|
|
|
op[1](np.array([x], dtype=_to_np_dtype(dtypes.float64))),
|
2024-07-13 15:31:21 +08:00
|
|
|
atol=3e-2, rtol=1e-5) # sin can have bigger atol for very big x
|
2024-07-13 13:54:52 +08:00
|
|
|
|
2024-08-05 04:06:38 +08:00
|
|
|
@unittest.skipIf(getenv("MOCKGPU") and Device.DEFAULT == "NV", "crashed")
|
2024-07-13 13:54:52 +08:00
|
|
|
@given(ht.float32, strat.sampled_from([(Tensor.exp, np.exp), (Tensor.log, np.log), (Tensor.sin, np.sin)]))
|
|
|
|
|
def test_float32(self, x, op):
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
np.testing.assert_allclose(op[0](Tensor([x], dtype=dtypes.float32)).numpy(),
|
|
|
|
|
op[1](np.array([x], dtype=_to_np_dtype(dtypes.float32))),
|
2024-07-13 15:31:21 +08:00
|
|
|
atol=2e-5, rtol=1e-5)
|
2024-07-13 13:54:52 +08:00
|
|
|
|
|
|
|
|
@unittest.skipUnless(is_dtype_supported(dtypes.float16, Device.DEFAULT), f"no float16 on {Device.DEFAULT}")
|
|
|
|
|
@given(ht.float16, strat.sampled_from([(Tensor.exp, np.exp), (Tensor.log, np.log), (Tensor.sin, np.sin)]))
|
|
|
|
|
def test_float16(self, x, op):
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
np.testing.assert_allclose(op[0](Tensor([x], dtype=dtypes.float16)).numpy(),
|
|
|
|
|
op[1](np.array([x], dtype=_to_np_dtype(dtypes.float16))),
|
2024-07-13 15:31:21 +08:00
|
|
|
atol=1e-2, rtol=4e-3) # exp can have bigger rtol
|
Approximations for SIN/LOG2/EXP2 passing all tests. (#5187)
* [WIP] Added an approximated implementation of Sin(FP32, FP64) passing all tests on Clang runtime
* Map nan/-inf/inf as 1.0 in order to avoid doing as_const(math.inf)
* [WIP] Added a support for LLVM IR
* cleaned up the code for the mypy and linter
* [WIP] Updated fp64 supports (bitwise shift causes the compilation error), fixed linter issue.
* [Add] added fast=true mode which disables the payne-hanek reduction which is slow
* [Fix] fails to compute elements when shape includes zero
* [WIP] Added BinaryOps.ADD/BinaryOps.OR to assembly
* [wip] update the assembly for ptx
* Enables fast=True when device is one of PTX, NV, CUDA, to avoid slow bitwise ops (as lv3 reduction is not required).
* [WIP] Added an approximation of LOG2/EXP2 (FP32, FP64)
* [Fix] Cyclic dependencies existing in xlog2
* [Fix] Cycle dependency in the graph of exp2, and log2. (passing test_symbolic_ops.py)
* [Fix] keep using higher precision for exp2, but cycle graph issue remained to be fixed...
* [Refactor] removed is_metal option. xsin does not rely on fp64 when fp32 mode.
* [WIP] fp16 xsin implementation passing all tests. (still needs to be refactored)
* [WIP] Added fp16 exp2 implementation
* [WIP] Increased the precision of Log2 from 3.5 ULP to 1.0 ULP, and added FP16 Log2 approximation.
* stashed the changes for FP16 sin
* [Fix] Patch for FP16 Sin/Exp2. (updated the dtype_via, fp32_p, and lower)
* [Refactor] migration to fastmath.py, some code simplification, renamed apis in fastmath, et al.
* [Refactor] Added the function polyN to clean-up N-terms polynomial approximation.
* [Patch] Increase fp64 precision when ldexp3k if possible, and patch for fp16 exp2
* [Patch] added bitcast_forward option
* [Patch] resolved cycle graph
* patch fix cycle graph
* set bitcast_forward=True in ilogb2k
* bitcast_forward for multi.py
* E501
* Break into multiple small PRs
* [Patch] FP16 -> FP64 upcast is not anymore required since xlog2 use quad precision polyN
* [Patch] NV still required FP64 for xlog2
* updated schedule test
* updated the count of kernels
* [Update] Removed all bitwise ops (SHL/SHR), tweaked the nan manipulation of log2, passing all tests except for AMD.
* Bitcast: make them api-compatible
* [update] force to use bitcast
* updated the count of constant folding
* [Patch] Creating a mask for exp2 using x <= Inf satisfies True as long as x is a real value
* [Update] isNaN(x) Free log2 algorithm, passing PTX tests, METAL with fastmath enabled is able to handle nan well, amd backend will not crash.
* xsin is reluctant to call payne_hanek_reduction which is slow to compile, passing stable diffusion compilation in a realistic time
* some minor simplification to payne hanek reduction
* [refactor] refactored some rebundant parts existing in payne hanek
* [refactor] more readable payne hanek impl
* [refactor] improved the code consistency of payne hanek
* [experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)
* Revert "[experiment] topological sort when doing _recursive_group (i dunno if this is good but at least it works.)"
This reverts commit 0eee08b87c9e46da8aec0a8edec5316634031a49.
* use allow_buffer_view
* lets support multilazytensor
* updated the count of kernels
* [test] added the jit tests for approx ops
* keep failed constant folding tests tested, added expectedFailure
* explict the timeout deadline when testing approx jit timeout
* [WIP] Simplified the implementation of xsin, never timeouts
* [Refactor] Improved the consistency of approx sin implementation, passing time out tests
* integrated xexp2_base into xexp2
* Set switch_over=39800.0
* delete: is_buffer_fastmath_supported
* sin: compute against abs(x)
* some cleanups
* fix typo
* removed the space between param and dtype
* allow 514 kernels on CI for sd
* [refactor] no need to upcast ad ldexp3k
* [refactor] added some comments, references to help understanding the code.
* [Fix] 1.0 ULP Sine Approximation for FP16
* [update] assume e != 0
* use pow2if instead of ldexp3k to fuse payne_hanek reduction into one
* check if approximated sin/log2/exp are fused into one
* clean up changes
* test amd exp
* some code cleanup and test sigmoid
* fix: enabled payne_hanek for fp16 to achieve higher acc
* fix: payne_hanek always accumlates the value with uint64, and fp16 sin is fused to a single kernel
* [Refactor] Rename: fastmath -> transcendental
* [Refactor] Added TRANSCENDENTAL, Moved the gate function to function.py
* updated const folding tests
* TRANSCENDENTAL as a ContextVar, removed old test of cody waite reduction, added assertions, et al.
* Add: unittest.main()
* Import TRANSCENDENTAL instead of getenv
* Refactor: Added dtype check when TRANSCENDENTAL=2, more context var
* Patch: xlog2, break expt(2, 32) x 2 -> expt(2, 16) x 4 for fp16 math
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
Co-authored-by: chenyu <chenyu@fastmail.com>
2024-07-11 07:44:58 +08:00
|
|
|
|
|
|
|
|
class TestTranscendentalSchedule(unittest.TestCase):
|
|
|
|
|
# w/ payne_hanek_reduction (fp32)
|
|
|
|
|
def test_transcendental_sin_fusion(self):
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
a = Tensor.empty(10)
|
|
|
|
|
b = Tensor.empty(10)
|
|
|
|
|
c = a.sin() + b.sin()
|
|
|
|
|
c = c.sin()
|
|
|
|
|
check_schedule(c, 1)
|
|
|
|
|
|
|
|
|
|
def test_transcendental_log2_fusion(self):
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
a = Tensor.empty(10)
|
|
|
|
|
b = Tensor.empty(10)
|
|
|
|
|
c = a.log2() + b.log2()
|
|
|
|
|
c = c.log2()
|
|
|
|
|
check_schedule(c, 1)
|
|
|
|
|
|
|
|
|
|
def test_transcendental_exp2_fusion(self):
|
|
|
|
|
with Context(TRANSCENDENTAL=2):
|
|
|
|
|
a = Tensor.empty(10)
|
|
|
|
|
b = Tensor.empty(10)
|
|
|
|
|
c = a.exp2() + b.exp2()
|
|
|
|
|
c = c.exp2()
|
|
|
|
|
check_schedule(c, 1)
|
|
|
|
|
|
|
|
|
|
if __name__ == '__main__':
|
|
|
|
|
unittest.main()
|
