Awni Hannun b0cc71ae71 Faster triu, tril, where with scalar (#2644)

2025-10-02 12:21:27 -07:00

..

fft

Fix fft for integer overflow (#2161)

2025-05-09 14:25:12 -07:00

indexing

Faster contiguous gather for indices in the first axis (#2552)

2025-08-28 21:26:30 -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

Align mlx::core::min op nan propagation with NumPy (#2346)

2025-07-10 06:20:43 -07:00

steel

Add sdpa with sinks (#2558)

2025-09-10 14:53:00 -07:00

arange.h

More jitting (#1132)

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

arange.metal

Custom logsumexp (#2028)

2025-03-31 07:36:55 -07:00

arg_reduce.metal

fix large arg reduce (#2206)

2025-05-19 13:10:44 -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 power (#2523)

2025-08-21 06:46:01 -07:00

binary_two.h

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

binary_two.metal

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

binary.h

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

binary.metal

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

cexpf.h

[CUDA] Implement Scan kernel (#2347)

2025-07-10 16:54:12 -07:00

CMakeLists.txt

Add mode parameter for quantization (#2499)

2025-08-28 06:45:26 -07:00

complex.h

Add more complex unary ops (#2101)

2025-04-21 13:04:54 -07:00

conv.metal

faster depthwise 1D conv (#2567)

2025-09-08 11:37:23 -07:00

copy.h

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

copy.metal

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

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

fence.metal

Faster synchronization Fence primitive (#1773)

2025-01-17 18:42:19 -08:00

fft.h

fix fft bug (#2062)

2025-04-10 19:41:27 -07:00

fft.metal

Add Quantized Ops to the JIT (#1204)

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

fp4_quantized.h

fix METAL quantization in JIT (#2553)

2025-08-28 18:26:25 -07:00

fp4_quantized.metal

fix METAL quantization in JIT (#2553)

2025-08-28 18:26:25 -07:00

gemv_masked.h

Ensure no oob read in gemv_masked (#2508)

2025-08-17 08:42:33 -07:00

gemv_masked.metal

Use int64 stride everywhere (#1671)

2024-12-09 11:09:02 -08:00

gemv.metal

New tuning for small K gemv (#2620)

2025-09-23 12:28:35 -07:00

hadamard.h

GPU Hadamard for large N (#1879)

2025-05-01 17:19:17 -07:00

layer_norm.metal

Fix layernorm race condition (#2340)

2025-07-07 06:06:01 -07:00

logsumexp.h

Fix out-of-bounds default value in logsumexp/softmax (#2213)

2025-05-21 07:25:16 -07:00

logsumexp.metal

Custom logsumexp (#2028)

2025-03-31 07:36:55 -07:00

quantized_utils.h

Add mode parameter for quantization (#2499)

2025-08-28 06:45:26 -07:00

quantized.h

Add mode parameter for quantization (#2499)

2025-08-28 06:45:26 -07:00

quantized.metal

Add mode parameter for quantization (#2499)

2025-08-28 06:45:26 -07:00

random.metal

Use int64 stride everywhere (#1671)

2024-12-09 11:09:02 -08:00

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

Fix reduce sum/prod overflow (#2477)

2025-08-12 00:05:33 -07:00

rms_norm.metal

Custom logsumexp (#2028)

2025-03-31 07:36:55 -07:00

rope.metal

no copy batch rope (#2595)

2025-09-15 14:23:48 -07:00

scaled_dot_product_attention.metal

Add float mask to sdpa vector (#2068)

2025-04-11 17:29:40 -07:00

scan.h

fix metal scan (#2591)

2025-09-15 11:01:57 -07:00

scan.metal

Complex scan (#2094)

2025-04-22 18:56:28 -07:00

sdpa_vector.h

avoid producing NaN in attention (#2608)

2025-09-22 13:10:43 -07:00

softmax.h

Fix out-of-bounds default value in logsumexp/softmax (#2213)

2025-05-21 07:25:16 -07:00

softmax.metal

Custom logsumexp (#2028)

2025-03-31 07:36:55 -07:00

sort.h

Ensure small sort doesn't use indices if not argsort (#2506)

2025-08-17 08:42:20 -07:00

sort.metal

faster sort (#1831)

2025-02-05 06:10:22 -08:00

ternary_ops.h

JIT compile option for binary minimization (#1091)

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

ternary.h

Faster triu, tril, where with scalar (#2644)

2025-10-02 12:21:27 -07:00

ternary.metal

Faster triu, tril, where with scalar (#2644)

2025-10-02 12:21:27 -07:00

unary_ops.h

[CUDA] Implement Scan kernel (#2347)

2025-07-10 16:54:12 -07:00

unary.h

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

unary.metal

Improve metal elementwise kernels (#2247)

2025-06-06 11:37:40 -07:00

utils.h

fix bw for elementwise ops (#2151)

2025-05-05 06:15:04 -07:00