Awni Hannun 2419edd5b2 Faster indexing math in a few kernels (#1589) ...

* wip: faster compiled kernels

* faster general unary with uint specialization

* index type in compiled, unary, binary, ternary, copy

* fix jit

* jit fix

* specialize gather + scatter

* nit in docs

2024-11-18 19:52:00 -08:00

..

fft

Feature complete Metal FFT (#1102)

2024-06-06 12:57:25 -07:00

jit

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

metal_3_0

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

metal_3_1

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

reduction

Reductions update (#1351)

2024-11-04 22:25:16 -08:00

steel

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

arange.h

More jitting (#1132)

2024-05-23 16:23:44 -07:00

arange.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

arg_reduce.metal

More fixes for arrays with large sizes (#1405)

2024-09-17 12:46:31 -07:00

atomic.h

Refactor reductions and fix scatter atomics for large sizes (#1300)

2024-08-22 16:03:31 -07:00

bf16_math.h

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

binary_ops.h

Fix complex power on Metal (#1460)

2024-10-06 19:58:30 -07:00

binary_two.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

binary_two.metal

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

binary.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

binary.metal

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

CMakeLists.txt

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

complex.h

Refactor reductions and fix scatter atomics for large sizes (#1300)

2024-08-22 16:03:31 -07:00

conv.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

copy.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

copy.metal

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

defines.h

Refactor reductions and fix scatter atomics for large sizes (#1300)

2024-08-22 16:03:31 -07:00

erf.h

JIT compile option for binary minimization (#1091)

2024-05-22 12:57:13 -07:00

expm1f.h

Fix overflow / underflow handling for expm1f (#1278)

2024-07-23 07:29:06 -07:00

fft.h

Feature complete Metal FFT (#1102)

2024-06-06 12:57:25 -07:00

fft.metal

Add Quantized Ops to the JIT (#1204)

2024-06-12 09:47:12 -07:00

gather.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

gemv_masked.h

Add gemv masked to JIT plus some fixes (#1310)

2024-08-07 13:38:07 -07:00

gemv_masked.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

gemv.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

hadamard.h

Fix bfloat16 Hadamard (#1283)

2024-07-23 14:54:43 -07:00

indexing.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

layer_norm.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

quantized.h

OOB QMV fix (#1579)

2024-11-08 17:59:45 -08:00

quantized.metal

Add split_k qvm for long context (#1564)

2024-11-05 11:25:19 -08:00

random.metal

Faster bits and bernoulli (#1535)

2024-10-28 11:11:00 -07:00

reduce_utils.h

More jitting (#1132)

2024-05-23 16:23:44 -07:00

reduce.h

Fix JIT reductions (#1373)

2024-08-28 16:39:11 -07:00

reduce.metal

Reductions update (#1351)

2024-11-04 22:25:16 -08:00

rms_norm.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

rope.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

scaled_dot_product_attention_params.h

Metal shaders for memory efficient self attention on large sequences (#964)

2024-06-03 09:16:19 -07:00

scaled_dot_product_attention.metal

2-Pass Sdpa Inference Kernel (#1597)

2024-11-18 17:31:53 -08:00

scan.h

Working 64-bit scans (#1506)

2024-10-24 11:05:46 -07:00

scan.metal

Working 64-bit scans (#1506)

2024-10-24 11:05:46 -07:00

scatter.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

sdpa_vector.h

2-Pass Sdpa Inference Kernel (#1597)

2024-11-18 17:31:53 -08:00

softmax.h

consistently handle all -inf in softmax (#1470)

2024-10-08 09:54:02 -07:00

softmax.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

sort.h

More fixes for arrays with large sizes (#1405)

2024-09-17 12:46:31 -07:00

sort.metal

Dispatch bf16 at run time when using the JIT (#1584)

2024-11-15 16:54:36 -08:00

ternary_ops.h

JIT compile option for binary minimization (#1091)

2024-05-22 12:57:13 -07:00

ternary.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

ternary.metal

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

unary_ops.h

Real and Imag (#1490)

2024-10-15 16:23:15 -07:00

unary.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

unary.metal

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00

utils.h

Faster indexing math in a few kernels (#1589)

2024-11-18 19:52:00 -08:00