Josh Moore 1f9d2442b9 Add Tensor.scatter_reduce (#8947) ... * pytorch scatter -> scatter_reduce * WIP scatter_reduce implementation * _pre_scatter return type hint * split out src, mask to satisfy linter * Add src cast back in * dict of lambdas instead of ifs * sum and prod reduction ops with include_self * add reduce arg error message * add amax and amin reduction ops * Fix include_self for higher dims * Simplify * Simplify amax and amin too * Pull include_self logic out into _inv_mask function * reduce arg cannot be None for scatter_reduce * Fix self-mask issue * Add mean reduce op * Add tests * any() not needed here * remove comment * End support for Tensor src with reduce arg in tinygrad scatter * Process index, dim inside actual functions * Add scatter_reduce to onnx * Add excluded onnx ScatterElements reduction tests back in * Save 2 lines on the mask helpers * Update docs * Add include_self=False tests * cleanup * Remove unneeded helper function --------- Co-authored-by: chenyu <chenyu@fastmail.com>		2025-02-13 09:08:54 -05:00
..
mlperf_bert	delete SAVE_SCHEDULE=1 [pr] (#7087)	2024-10-16 07:13:20 +03:00
mlperf_resnet	PolynomialDecayWithWarmup + tests (#3649)	2024-03-07 18:53:36 -05:00
mlperf_unet3d	[MLPerf] UNet3D dataloader (#4343)	2024-04-28 22:34:18 -04:00
openpilot	default threefry (#6116)	2024-09-25 17:45:13 +08:00
process_replay	rename to KernelContext and move the linearize_sched comment [pr] (#8899)	2025-02-05 07:49:58 +01:00
external_benchmark_hcopt.py	move time_linearizer to extra.optimization.helpers [pr] (#9048)	2025-02-12 15:49:58 -05:00
external_benchmark_hip_compile.py	hip compile speed (#2606)	2023-12-04 13:47:40 -08:00
external_benchmark_kernel_launch.py	hotfix: GRAPH_ONE_KERNEL + fix timing	2025-02-06 17:52:20 +08:00
external_benchmark_load_stable_diffusion.py	gate METAL_FAST_LOAD	2023-12-01 15:28:40 -08:00
external_benchmark_multitensor_allreduce.py	allreduce: add flags (#8713)	2025-01-22 17:44:31 +03:00
external_benchmark_openpilot.py	make onnx runner a class (#8647)	2025-01-20 10:11:05 -08:00
external_benchmark_resnet.py	delete SAVE_SCHEDULE=1 [pr] (#7087)	2024-10-16 07:13:20 +03:00
external_benchmark_schedule.py	rename uopgraph to rewriter [pr] (#8682)	2025-01-19 17:03:12 -08:00
external_cl_half_max.py	use default dict for external_model_benchmark (#2592)	2023-12-03 15:25:43 -08:00
external_debug_metal_sd_conv.py	[TIP-9] rename Opt's amt to arg 2 (#8770)	2025-01-27 14:19:04 -05:00
external_fuzz_am_interrupts.py	am: recover from any boot interrupt (#8703)	2025-01-21 22:22:23 +03:00
external_fuzz_ampt.py	am: cleanup mm (#8730)	2025-01-23 15:49:37 +03:00
external_fuzz_hcq_signals.py	hcq pci signal fuzzer (#8854)	2025-02-01 23:42:27 +03:00
external_fuzz_tlsf.py	AM driver (#6923)	2024-12-31 23:06:17 +03:00
external_gpu_fail_osx.py	new openpilot compile (#6573)	2024-09-18 14:22:50 +08:00
external_hip_compiler_bug.py	CompiledASTRunner -> CompiledRunner (#4148)	2024-04-11 08:49:52 -07:00
external_jit_failure.py	fix jit realize issue (#3258)	2024-01-26 18:27:35 -08:00
external_llama_eval.py	ruff checks the max line length is 150 (#2734)	2023-12-12 17:34:47 -08:00
external_metal_compile_fail.py	Use MTLCompiler directly (v2) (#7920)	2024-12-04 16:36:48 +08:00
external_model_benchmark.py	make onnx runner a class (#8647)	2025-01-20 10:11:05 -08:00
external_multi_gpu.py	move disassemblers and openpilot (#4592)	2024-05-14 19:30:02 -07:00
external_osx_profiling.py	move dtypes to dtype.py (#2964)	2024-01-01 14:58:48 -08:00
external_test_am.py	am: remove libpciaccess dep (#8980)	2025-02-09 16:06:55 +03:00
external_test_amd.py	RUF018 assignment-in-assert [run_process_replay] (#6172)	2024-08-19 00:34:52 -04:00
external_test_datasets.py	Revert "make kits19 dataset samples have small sizes (#8591)" (#8610)	2025-01-14 12:24:27 -05:00
external_test_embedding.py	make embedding and GPT-2 fast (#1631)	2023-08-22 15:14:38 -07:00
external_test_example.py	dsp simulator (#8869)	2025-02-04 09:45:04 +08:00
external_test_hcq.py	create_schedule([x.lazydata]) -> x.schedule() in tests (#8449)	2024-12-31 03:15:52 +08:00
external_test_hcq_fuzz_failures.py	[TIP-9] rename Opt's amt to arg 2 (#8770)	2025-01-27 14:19:04 -05:00
external_test_hip_compile.py	lowerer is kernel [run_process_replay] (#5437)	2024-07-12 18:50:55 -07:00
external_test_hsa_driver.py	Rename tinygrad/runtime/driver to support (#5413)	2024-07-12 11:06:42 -07:00
external_test_image.py	fixes (#1893)	2023-09-22 07:20:27 +08:00
external_test_jit_on_models.py	Pulled CLIP and UNet into Seperate Files (#5253)	2024-07-01 22:33:01 -04:00
external_test_llama3_ff.py	work to make GEMV fast (#5824)	2024-07-30 17:41:40 -07:00
external_test_lm_head.py	isolate the 134ms kernel in train_gpt2.py (#4773)	2024-05-29 17:26:24 -04:00
external_test_losses.py	[MLPerf][UNet3D] Add DICE loss + metrics (#4204)	2024-04-17 20:09:33 -04:00
external_test_mamba.py	external that test	2024-03-29 19:35:50 -07:00
external_test_metrics.py	Convert BinaryOps.DIV to UnaryOps.RECIP and BinaryOps.IDIV (#4887)	2024-06-14 02:43:46 -07:00
external_test_mnist_data_select.py	add quick external data select test	2024-03-02 05:38:32 -08:00
external_test_nv.py	tc_select noop (#8801)	2025-01-29 13:53:23 -05:00
external_test_onnx_backend.py	Add Tensor.scatter_reduce (#8947)	2025-02-13 09:08:54 -05:00
external_test_opt.py	remove PUSH_PERMUTES from external_test_opt (#7232)	2024-10-23 00:11:34 -04:00
external_test_optim.py	improve test_dropout_on_shard (#4912)	2024-06-11 11:36:02 -04:00
external_test_speed_llama.py	rename allocator methods to not conflict [pr] (#7788)	2024-11-20 00:10:29 +08:00
external_test_speed_theoretical.py	test flops (and allow wide ALU in UOps) [run_process_replay] (#5749)	2024-07-26 21:07:28 -07:00
external_test_tlsf.py	AM driver (#6923)	2024-12-31 23:06:17 +03:00
external_test_train_gpt2.py	[TIP-9] rename Opt's amt to arg 2 (#8770)	2025-01-27 14:19:04 -05:00
external_test_valid_remove.py	[TIP-9] rename Opt's amt to arg 2 (#8770)	2025-01-27 14:19:04 -05:00
external_test_whisper_librispeech.py	names shadowing builtins (#5179)	2024-06-27 08:15:01 -04:00
external_test_yolo.py	move to new cached fetch (#2493)	2023-11-28 17:36:55 -08:00
external_test_yolov8.py	ruff checks the max line length is 150 (#2734)	2023-12-12 17:34:47 -08:00
fuzz_graph.py	create_schedule([x.lazydata]) -> x.schedule() in tests (#8449)	2024-12-31 03:15:52 +08:00
fuzz_kfd.py	unify to HWQueue [pr] (#7812)	2024-11-21 10:33:08 +08:00
fuzz_linearizer.py	more typing cleanups [pr] (#8376)	2024-12-22 05:21:03 -05:00
fuzz_schedule.py	rename LazyBuffer -> UOp [pr] (#8169)	2024-12-11 16:15:52 -08:00
fuzz_shapetracker.py	[bounty] Stride is flip (#8792)	2025-01-31 11:34:10 +09:00
fuzz_shapetracker_math.py	more typing cleanups [pr] (#8376)	2024-12-22 05:21:03 -05:00
fuzz_shapetracker_size.py	limit real_size to the size of first View of ShapeTracker (#8628)	2025-01-16 16:27:39 -05:00
fuzz_symbolic.py	Remove UnaryOps, BinaryOps, TernaryOps, MetaOps [pr] (#7725)	2024-11-16 20:56:56 +08:00
fuzz_uops.py	minor uop speedups [pr] (#8002)	2024-12-03 17:04:48 +08:00
graph_batchnorm.py	with Tensor.train() (#1935)	2023-09-28 18:02:31 -07:00
speed_beam_v_hcopt.py	move time_linearizer to extra.optimization.helpers [pr] (#9048)	2025-02-12 15:49:58 -05:00
speed_compare_cuda_nv.py	move colorize_float to helpers.py (#5490)	2024-07-15 11:29:03 -07:00
speed_compare_cuda_ptx.py	move colorize_float to helpers.py (#5490)	2024-07-15 11:29:03 -07:00
speed_v_theoretical.py	lower green test_gemm_8192 tflops to 125 [pr] (#8820)	2025-01-30 17:30:08 -05:00
verify_kernel.py	move time_linearizer to extra.optimization.helpers [pr] (#9048)	2025-02-12 15:49:58 -05:00