9,700 commits

This branch

This branch

All branches

Author	SHA1	Message	Date
George Hotz	29262a7543	fine for ptx	2025-08-05 18:26:06 -07:00
George Hotz	dcc6ddf0eb	that hack broke things	2025-08-05 18:24:58 -07:00
George Hotz	d1d935242b	Revert "fix tests" ... This reverts commit a27019383d.	2025-08-05 18:06:27 -07:00
George Hotz	6b330f302d	remote metal was flaky	2025-08-05 17:50:14 -07:00
George Hotz	a27019383d	fix tests	2025-08-05 17:45:35 -07:00
George Hotz	8b285e193a	move those to fix_kernel_ops	2025-08-05 17:07:12 -07:00
George Hotz	f0c9b11b9e	early meta ops	2025-08-05 17:00:49 -07:00
George Hotz	1902a85ac1	early load buffer	2025-08-05 16:56:10 -07:00
George Hotz	b2fc111e3f	cleanup fix_kernel	2025-08-05 16:46:13 -07:00
George Hotz	067daee5be	pin torch to 2.7.1 (#11519)	2025-08-05 15:58:57 -07:00
George Hotz	b39f43c46a	optimize in rewrite, try 2 (#11518) ... * changes * fix test uops * optimize in rewrite, try 2	2025-08-05 15:52:53 -07:00
George Hotz	07b0df0d86	hotfix: test tensor dims start at 1	2025-08-05 15:40:24 -07:00
George Hotz	4dabdf7c6d	Revert "optimize in rewrite (#11516)" (#11517) ... This reverts commit 3b777a9e05.	2025-08-05 15:39:07 -07:00
George Hotz	3b777a9e05	optimize in rewrite (#11516) ... * changes * fix test uops * dim shouldn't be 0 * huh, why did that one not save	2025-08-05 15:33:26 -07:00
nimlgen	ec676eddfa	nv: move base address higher (#11514)	2025-08-05 22:42:53 +03:00
qazal	7703f8b805	viz: skip flops info if estimates is symbolic (#11513)	2025-08-05 22:12:52 +03:00
nimlgen	fc4e713d1c	jit graph split tests (#11507) ... * jit graph split tests * fix * one more test * more tests * fix * xm * rmeote	2025-08-05 21:32:37 +03:00
George Hotz	c57fde51f9	move swizzler to opt (#11509)	2025-08-05 11:31:30 -07:00
chenyu	ace8e9a706	fix test_conv2d_winograd (#11511)	2025-08-05 12:15:46 -04:00
chenyu	223aaa0492	clean up more conv tests (#11510)	2025-08-05 12:15:30 -04:00
Garret Castro	76e62a1c23	extract conv layer test logic (#11488) ... * refactor: extract conv layer test logic * tuple is unnecessary * integrate _test_conv logic into all conv tests * fix linter, forgot dilation * undo winograd extraction adds too many if statements for a single case	2025-08-05 11:15:54 -04:00
b1tg	8b8bd6c534	make einsum generate same kernels (#11508) ... Co-authored-by: b1tg <b1tg@users.noreply.github.com>	2025-08-05 11:12:52 -04:00
uuuvn	011ef8fa9d	Fix incorrect jit current batch devs reset (#11505) ... `current_batch_devs = []` (in `flush_batch()`) happens between `new_batched_devs = ...` and `current_batch_devs = new_batched_devs` => doesn't actually reset anything leading to things not jitting properly which 2xs remote bert step time (should have similar effects on any non-hcq backend)	2025-08-05 08:16:16 +03:00
chenyu	f02720ca2d	fix fuse gate_contiguous unique (#11504)	2025-08-04 23:43:31 -04:00
George Hotz	7f6acfb0d5	give define global and friends a shape (#11502) ... * give define global and friends a shape * ignore negative size * ptx fix	2025-08-04 19:09:39 -07:00
chenyu	83385e7abc	update gradient src in ramp.py (#11499) ... that's simplified now	2025-08-04 18:58:03 -04:00
qazal	846a2826ab	viz: remove TracingKey.fmt (#11482) ... * viz: remove TracingKey.fmt * remove from test too	2025-08-05 00:00:03 +03:00
chenyu	01d44e8f16	tiny reduce_gradient cleanup [pr] (#11498)	2025-08-04 16:12:53 -04:00
chenyu	8a11af01ed	remove broken paperswithcode links in doc (#11497)	2025-08-04 13:12:33 -04:00
leopf	4f0ee4e982	BPE tokenizer (#11415) ... * BPE works * refactor tok * oops * basic tests * fix eval * smaller diff * fix error * proper vocab decoding * use regex for splitting * escape ucatrange * full compat --------- Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>	2025-08-04 09:52:38 -07:00
b1tg	06af9f9236	fix double exception + add name,loc in error msg (#11487) ... Co-authored-by: b1tg <b1tg@users.noreply.github.com>	2025-08-04 13:41:23 +03:00
nimlgen	4877aa965a	ast seems to probe nv as well (#11494)	2025-08-04 11:47:07 +03:00
chenyu	e0106b6b25	1/(x*c) -> (1/c)*(1/x) (#11491) ... example: 2*(2*a).reciprocal() -> a.reciprocal() # TODO: bounds for reciprocal # TODO: should z3 work?	2025-08-03 23:35:46 -04:00
qazal	5870352fe1	viz: factorize llvm-mca call (#11490)	2025-08-04 00:31:23 +03:00
chenyu	dbc7807c61	enable WEBGPU tests with buffer limit (#11489) ... TestSample still fails?	2025-08-03 13:02:44 -07:00
nimlgen	8f374ee1f7	nv: print devfmr in gsp logs (#11484)	2025-08-03 15:12:53 +03:00
chenyu	823f1a01db	move cast around expand backward to tensor.py (#11483)	2025-08-02 23:03:54 -04:00
chenyu	0ce0f51010	generic double cast folding (#11481) ... b.cast(a).cast(b) -> b if a preserves all values in b	2025-08-02 19:26:37 -04:00
qazal	72e0d1d0dc	viz: profile the compiler in TINY device (#11457) ... * viz: profile the compiler in TINY device * leanup	2025-08-03 02:03:20 +03:00
chenyu	66be747908	few more dtype cast convinience methods (#11480)	2025-08-02 15:47:09 -04:00
chenyu	e22e5da9a5	move some test_dtype tests to unit (#11479)	2025-08-02 15:25:00 -04:00
nimlgen	da0b955be4	hcq: cpu can be graphed (#11474) ... * hcq: cpu can be graphed * ops * new jit decisions * fix test * fix remote * cleaner * fix	2025-08-02 21:01:19 +03:00
chenyu	f7965f85aa	Revert "feat: faster index building (#11462)" (#11478) ... This reverts commit 3a4deb08d2.	2025-08-02 12:50:48 -04:00
kevvz	ef7e01cadf	Fix SVD shape bug + Fix batched SVD bug (#11477) ... * failing test case * fix * better test * space	2025-08-02 09:47:41 -07:00
b1tg	6ecaf8e7b2	refactor: use less index and simplify reduce axes check [pr] (#11476) ... * use output_shape/full_shape * simple final_reduces check --------- Co-authored-by: b1tg <b1tg@users.noreply.github.com>	2025-08-02 09:44:51 -07:00
wozeparrot	3a4deb08d2	feat: faster index building (#11462) ... * feat: faster index building * feat: correct training samples	2025-08-02 11:50:18 -04:00
nimlgen	8cc2d64edb	amd: reuse create_queues for usb iface (#11473)	2025-08-02 14:40:46 +03:00
chenyu	9e8e6b45ab	grad acc train llama (#11467) ... * grad acc train llama * log step time	2025-08-01 15:54:50 -04:00
chenyu	7ad7329257	data parallel train llama (#11466)	2025-08-01 12:13:51 -04:00
nimlgen	9f2182f92f	cpu: start threading (#11324) ... * cpu: threading * syncs * llvm * fix * opt * fx * fix * missed sync * one line less * cleaner * fix	2025-08-01 15:35:07 +03:00