commaai
/
tinygrad
mirror of https://github.com/commaai/tinygrad.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
484 Commits 10 Branches 0 Tags 52 MiB
Commit Graph

484 Commits

Author SHA1 Message Date
Marcel Bischoff 5d46df638a
abs as non-first class operation using relu (#171) 
* abs (non-first class)

* whitespace
 2020-12-09 12:20:34 -08:00
George Hotz 4c55c7208f no pow if mul will do 2020-12-09 08:19:29 -08:00
George Hotz b85f17f247 more optim cleanup 2020-12-09 08:18:10 -08:00
George Hotz 9a64d13b94 add conv biases and max pool 2020-12-09 08:01:20 -08:00
George Hotz 99fa65f057 enable batchnorm in serious mnist 2020-12-09 03:29:40 -08:00
George Hotz ffb96b2d0b batchnorm by marcelbischoff 2020-12-09 03:23:04 -08:00
NeuralLink 00e376f36c
leaky relu as geohot suggested (#167) 2020-12-09 02:58:35 -08:00
George Hotz c225e62dd2 touchups 2020-12-09 02:52:28 -08:00
Liam 89d0ff6989
Consistent testing (#137) 
* Consistent GPU classes

Convert the existing GPU classes into one standard format.

Remove duplicated functions in `test_mnist` and create a TestMNISTGPU
class. This reduces line count and ensures consistency.

Use `@unittest.skipUnless(GPU, "Requires GPU")` instead of `if GPU:` to
skip GPU testing. This will ensure that skipped tests are displayed
accordingly in the pytest output.

* Optim Testing now supports GPU

* Tensor testing now supports GPU

jacobian and gradcheck auto skipped until GPU float64 support added.

* GPU support for custom constructor methods

* Remove GPU flag from Model constructors

It was requested that the `gpu` kwarg be removed from the model
constructor. GPU conversion is now handled in the train function.

This also required the conversion of Optimizer parameters as they are
constructed prior to execution of the `train` function and are dependant
on the model GPU state.

* Fix typo: float32->float64

* Clean `get_parameters` utility

Just a quick refactor w/ the new support for optimizers.

* Remove GPU kwarg from TinyNet

Remove `gpu` kwarg from tiny net to match test_mnist `train` function.
 2020-12-09 02:25:27 -08:00
Liam 34b38dd4d0
Extra install requirements. (#164) 
* Testing install requirements

* GPU install requirements
 2020-12-09 02:22:47 -08:00
George Hotz 0e02f394ee serious_mnist 2020-12-08 21:43:05 -08:00
Daulet 24d688c184
win more lines for core library (#158) 
...and sacrifice test speed
 2020-12-08 14:18:45 -08:00
NeuralLink 9f77fd6135
🔨 refactor optim (#156) 
* 🔨 refactor optim

* 🔨 refactor optim

* 🔨 more clean up
 2020-12-08 14:16:31 -08:00
George Hotz 4e1a0de392 fix rsub 2020-12-08 10:05:21 -08:00
George Hotz c4540f1b8c Support scalars by kartik4949 2020-12-08 09:52:07 -08:00
George Hotz 97fd9c1237 zero_grad there to match readme 2020-12-07 23:12:18 -08:00
George Hotz c63f950348 need zero grad now 2020-12-07 23:10:43 -08:00
George Hotz b355cd2571
Mean axis (doesn't work) (#154) 
* mean axis

* fixed
 2020-12-07 22:58:34 -08:00
George Hotz 38f97c8c80 prepare for ops_ane 2020-12-07 21:54:22 -08:00
George Hotz 7f249ec76d touch up 2020-12-07 21:51:32 -08:00
Marcel Bischoff 58ccebd7cd
Sum with axis (#153) 
* sum with axis and tests

* broken

* works again

* clean up

* Update test_ops.py
 2020-12-07 21:49:18 -08:00
George Hotz ac9fecb05d lots of notes 2020-12-07 21:40:31 -08:00
George Hotz 8d1500f497 conv neuron 2020-12-07 21:12:52 -08:00
George Hotz e4bb53b0e9 work out more 2020-12-07 20:32:50 -08:00
George Hotz 4927ad1897 float16 weights in min.weights 2020-12-07 20:15:15 -08:00
George Hotz 3aac9aefce fix GPU profiling 2020-12-07 20:03:28 -08:00
James Roberts b2eca6d45f
Format debug output (#152) 2020-12-07 14:07:14 -08:00
George Hotz c7973cb0a1 ugh buffer_np is bad 2020-12-07 08:07:00 -08:00
George Hotz 088f280dc3 touchups 2020-12-07 07:50:27 -08:00
George Hotz 0cf21881b7 hwx parse w/o macho mods 2020-12-06 23:13:28 -08:00
Josh Smith aa4161f63e
use classmethods for Tensor helper funcs (#146) 2020-12-06 22:35:43 -08:00
George Hotz e75a6d1b4b quadconv 2020-12-06 20:39:50 -08:00
George Hotz 23664c99bd double conv 2020-12-06 20:26:02 -08:00
George Hotz 1a0f826dc6 highlight the commands 2020-12-06 20:03:21 -08:00
George Hotz 2f1f006003 we have docs 2020-12-06 19:54:03 -08:00
George Hotz 3531e81f0e dumping ANE docs 2020-12-06 18:58:36 -08:00
George Hotz 6e793e96c3 deeebug 2020-12-06 17:49:17 -08:00
George Hotz dced0cb44b oops, path to weights 2020-12-06 16:33:42 -08:00
George Hotz c57dc61ea7 simple op examples 2020-12-06 16:32:26 -08:00
George Hotz 7babf38617 found concat 2020-12-06 16:27:12 -08:00
George Hotz ddd6778423 add neuron 2020-12-06 16:24:42 -08:00
George Hotz f2f2d6aea3 docs and noop 2020-12-06 16:10:44 -08:00
George Hotz 6ba25834ee found some plists 2020-12-06 15:57:28 -08:00
George Hotz e2184c20ad min weights, update golden 2020-12-06 15:29:15 -08:00
George Hotz d4d8bd0337 make minimal plist for compare 2020-12-06 15:10:15 -08:00
George Hotz 0845ec43c6 compile takes in plist 2020-12-06 14:51:33 -08:00
George Hotz 00312b8ad1 batchnorm work 2020-12-06 14:40:07 -08:00
George Hotz da514c2918 fix enet init 2020-12-06 13:52:07 -08:00
George Hotz 3b982f2f7a get_parameters 2020-12-06 13:47:28 -08:00
George Hotz 102e6356e9 replace layer_init_uniform with .uniform 2020-12-06 13:44:31 -08:00