sirhcm/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests 2 Actions Activity
11,106 Commits 336 Branches 15 Tags
python_speed
Commit Graph
11106 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
Matthew DaiterandGitHub b545338e59 isin_Tensor_out added (#10018) 2025-04-24 07:26:51 -04:00
chenyuandGitHub a25abf55e3 retinanet only call postprocess_detections with RUNMLPERF (#10017) 
during setup only need to compile `_eval_step().numpy()`
 2025-04-23 20:45:38 -04:00
nimlgenandGitHub 7f53e80db9 hotfix: amd mmio on mi300 (#10016) 
* hotfix: amd mmio on mi300

* fix

* ops
 2025-04-24 01:08:18 +03:00
nimlgenandGitHub 1c5e353249 am: use mmio iface (#10012) 
* am: use mmio iface

* linters

* fixes

* fixes + cleanups

* mute

* mypy

* style
 2025-04-24 00:27:04 +03:00
chenyuandGitHub 65faa1d94b explicit device in mlperf scripts (#10015) 2025-04-23 17:11:52 -04:00
chenyuandGitHub a3f938dbee remove retinanet INITMLPERF from beam script (#10011) 
it only controls logging, loading real data or not is solely controlled by RUNMLPERF
 2025-04-23 14:32:54 -04:00
nimlgenandGitHub cc52b9c528 am: add entry() to PT (#10010) 2025-04-23 20:45:52 +03:00
nimlgenandGitHub c952cb965e amd: use mmio iface (#9997) 
* amd: use mmio iface

* mypy

* rename
 2025-04-23 20:13:00 +03:00
Francis LataandGitHub 5542aeb0e4 RetinaNet MLPerf flag updates (#10009) 
* add RUNMLPERF and update INITMLPERF usage

* update scripts to use RUNMLPERF
 2025-04-23 13:00:34 -04:00
George HotzandGitHub de0504276b pop 0 is slow [pr] (#10007) 2025-04-23 17:00:59 +01:00
chenyuandGitHub d3a8d5c128 print postprocess_detections time in retinanet eval (#10005) 
`BS=96 BASEDIR="/raid/datasets/openimages" MODEL=retinanet python examples/mlperf/model_eval.py`

```
...
loaded dataset             @  8.64s
loaded initial data        @ 12.57s
******  619.97 ms to enqueue, 46042.13 ms to realize ( 116.22 ms fetching, 45399.58 ms postprocess_detections).     0.09 examples/sec.  0.83 TFLOPS  @ 59.23s
******  147.49 ms to enqueue, 37362.16 ms to realize ( 146.96 ms fetching, 36618.84 ms postprocess_detections).     0.11 examples/sec.  1.03 TFLOPS  @ 96.74s
******  152.85 ms to enqueue, 37244.08 ms to realize ( 120.67 ms fetching, 36235.19 ms postprocess_detections).     0.11 examples/sec.  1.04 TFLOPS  @ 134.14s
******  146.39 ms to enqueue, 37279.85 ms to realize (  65.07 ms fetching, 36233.56 ms postprocess_detections).     0.11 examples/sec.  1.04 TFLOPS  @ 171.56s
******  152.41 ms to enqueue, 37264.04 ms to realize ( 127.08 ms fetching, 36196.10 ms postprocess_detections).     0.11 examples/sec.  1.04 TFLOPS  @ 208.98s
******  151.29 ms to enqueue, 36868.08 ms to realize ( 142.73 ms fetching, 36153.07 ms postprocess_detections).     0.11 examples/sec.  1.05 TFLOPS  @ 246.00s
******  136.41 ms to enqueue, 37325.04 ms to realize (  90.29 ms fetching, 36573.38 ms postprocess_detections).     0.11 examples/sec.  1.04 TFLOPS  @ 283.46s
```
 2025-04-23 11:39:56 -04:00
George HotzandGitHub 2ed3acd767 toposort is a function [pr] (#10004) 2025-04-23 16:25:03 +01:00
uuuvnandGitHub 0730ff0e50 Skip test that requires lru if device's allocator isn't lru (#10003) 2025-04-23 16:12:56 +01:00
George HotzandGitHub 954cb06957 deepwalk without recursion [pr] (#10002) 
* deepwalk without recursion [pr]

* comment and remove that test
 2025-04-23 15:57:50 +01:00
uuuvnandGitHub 9de73ccc22 Skip test that requires python 3.12 on older versions (#10001) 
`out.cast(it.dtype.fmt).tolist()` fails with `ValueError: memoryview: destination format must be a native single character format prefixed with an optional '@'`
 2025-04-23 10:09:26 -04:00
George HotzandGitHub 71ecc7fa1a use a pattern matcher for upcast [pr] (#10000) 2025-04-23 14:24:23 +01:00
George HotzandGitHub cc1087d2ec move simplify into views_to_indexed_uops (#9999) 
* move simplify into views_to_indexed_uops

* cache that
 2025-04-23 13:50:27 +01:00
chenyuandGitHub c39128133c retinanet green scripts (#9996) 
also removed realize in data_get and used empty for fake data. slightly bigger lr. https://wandb.ai/chenyuxyz/MLPerf-RetinaNet/runs/8skid0e8?nw=nwuserchenyuxyz
 2025-04-23 08:28:03 -04:00
George HotzandGitHub a4a5f2d54a faster block order [pr] (#9998) 
* faster block reorder [pr]

* ahh, that's even faster
 2025-04-23 13:11:30 +01:00
chenyuandGitHub 61bfd23881 update mlperf-logging version (#9995) 2025-04-22 19:32:39 -04:00
pkotzbachGitHubpkotzbach
 		dbbd755cba FP8s truncate (#9937) 
* truncate fp8

* fix

* maybe like that?

* fix linters

* ruff

* move from extra and add ml_types to tests

* minor changes

* str to dtypes and nan support

---------

Co-authored-by: pkotzbach <[email protected]>
 2025-04-22 19:12:49 -04:00
qazalandGitHub 58180caad3 schedule linearize small cleanups [pr] (#9994) 2025-04-23 05:42:29 +08:00
qazalandGitHub f4ec57baff new schedule linearizer enqueues KERNEL UOps [pr] (#9993) 
* new schedule linearizer enqueues kernels [pr]

* no defaultdict

* diff

* minor
 2025-04-23 05:17:58 +08:00
geohot d1f6701eb7 hotfix: lower amd threshold + improve block reorder test 2025-04-22 20:44:29 +01:00
nimlgenandGitHub db51133537 rename HWInterface -> FileIOInterface (#9989) 
* rename HWInterface -> FileIOInterface

* ugh
 2025-04-22 22:18:57 +03:00
George HotzandGitHub c1539b0319 putting add first orders loads as expected (#9991) 2025-04-22 20:12:05 +01:00
nimlgenandGitHub bd580d8ea4 hcq: use mmio interface in nv (#9986) 
* hcq: start mmio interface

* allow double cast

* revert

* faster?

* simpler, not needed more now

* dd

* types

* fix
 2025-04-22 21:58:12 +03:00
George HotzandGitHub feee6986c9 faster block reorder (#9990) 
* faster block reorder [pr]

* that shouldn't change order

* key just in sorted

* ind
 2025-04-22 19:18:57 +01:00
qazalandGitHub 6cb2d18c03 refactor schedule linearize to defaultdict [pr] (#9984) 
* refactor schedule linearize to defaultdict [pr]

* skip that

* don't need .get
 2025-04-23 00:00:23 +08:00
chenyuandGitHub 9e5e371999 make DISABLE_COMPILER_CACHE a ContextVar [pr] (#9983) 2025-04-22 10:32:54 -04:00
qazalandGitHub bbc324f5dc remove CAST_AFTER_EXPAND (#9980) 2025-04-22 21:06:11 +08:00
George HotzandGitHub c519b553db non recursive toposort is 2x+ faster (#9979) 
* non recursive toposort is 2x+ faster

* don't change the order
 2025-04-22 13:59:38 +01:00
qazalandGitHub 0d9014d021 place create_ast last, type_verify in the end (once) [pr] (#9977) 2025-04-22 20:15:23 +08:00
chenyuandGitHub fb89d9a584 retinanet eval combine output on GPUS[0] (#9966) 
eval 35 sec -> 20 sec. it was spending 13 seconds assembling output tensor on CPU backend. GPUS[0] seems to have enough memory, otherwise we can lower EVAL_BS
 2025-04-22 07:43:51 -04:00
qazalandGitHub 7b55846e08 prep STORE UOp creation for multi output [pr] (#9975) 
* prep STORE UOp creation for multi output [pr]

* test_multioutput_ast
 2025-04-22 19:34:52 +08:00
George HotzandGitHub e358e0a0c6 move metadata set to tensor [pr] (#9976) 
* move metadata set to tensor [pr]

* only track that in tensor.py
 2025-04-22 12:30:35 +01:00
qazalandGitHub f6271515fe refactor UOp.st [pr] (#9973) 2025-04-22 18:46:56 +08:00
George HotzandGitHub f5dc70c624 microbenchmarks + micro speed ups (#9972) 
* microbenchmarks

* forgot the ubenchs

* clean up type verify
 2025-04-22 11:30:46 +01:00
qazalandGitHub 1cf4e24ca5 fix kernelize usage with pm_gradient (#9953) 
* fix kernelize usage with pm_gradient

* remove that
 2025-04-22 17:26:05 +08:00
deftdawgGitHubnimlgen
 		32bbff942c amd: add nbio 7.2.0 for some rdna2 (#9964) 
* - Updated of #9700 which fixes #9665 but for the Steam Deck which was erroring on NBIO 7.2.0

* unrelated change

---------

Co-authored-by: nimlgen <[email protected]>
 2025-04-22 12:10:48 +03:00
Ignacio SicaandGitHub 0e79aee706 use_tensor_cores bugfix (#9969) 2025-04-21 22:58:17 -03:00
chenyuandGitHub 5294c32279 dev scripts for retinanet (#9968) 
also BASE_DIR -> BASEDIR for consistency, and move wandb up a bit for more accurate timing
 2025-04-21 17:54:56 -04:00
nimlgenandGitHub 4340197132 am: download fw from web (#9956) 
* am: download fw from web

* tested

* link works

* default to web

* this is default

* not used
 2025-04-21 23:26:33 +03:00
nimlgenandGitHub 7244ca863c am: fix double read of sdma fw (#9965) 2025-04-21 23:04:34 +03:00
uuuvnandGitHub b35f94b6ec Don't hardcode default CLOUDDEV (#9935) 2025-04-21 18:46:55 +01:00
Francis LataandGitHub defa1e77f6 get the proper dataset count (#9962) 2025-04-21 12:11:37 -04:00
qazalandGitHub 36ed3c3253 fix kernelize with VIEW children (#9961) 2025-04-21 23:38:46 +08:00
uuuvnandGitHub 757533cbe6 Less verbose cloud multiprocessing start (#9960) 
The set name before starting part used to be required for #9935 when
CLOUDDEV was a global variable, now just readability improvement
 2025-04-21 16:19:54 +01:00
Francis LataandGitHub d7e247f329 RetinaNet INITMLPERF support (#9950) 
* fixes to make fake data work

* fix eval beam

* fix merge issue
 2025-04-21 10:32:05 -04:00
kamilisjonGitHubKamilisJonkus
 		014f870733 rm (#9959) 
Co-authored-by: KamilisJonkus <[email protected]>
 2025-04-21 15:23:45 +01:00