github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-01-10 07:28:15 -05:00
Code Issues Packages Projects Releases Wiki Activity
3,312 Commits 203 Branches 13 Tags
8cbcd1b342e9d65eff682e9384d71af54f79b097
Commit Graph

3312 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
George Hotz
8cbcd1b342 Remove webgpu, back to 5k lines (#3040) 
* remove webgpu

* max 5000 lines
 2024-01-08 09:10:07 -08:00
George Hotz
cf2eea961c more beautiful_cartpole with exposed hparams 2024-01-07 17:41:09 -08:00
Yixiang Gao
44618427f1 add bf16 type_map for both cuda and hip (#3036) 
* add typemap bfloat16 for cuda and hip

* add render_dtype

* add def in CStyleLanguage

* fix def

* save one line

* add header file for cuda bf16
 2024-01-07 14:26:55 -08:00
chenyu
ef5f545fd8 add more Tensor.clip test cases (#3034) 
* add more Tensor.clip test cases

add cases for same low/high and both negative etc

* case min > max
 2024-01-07 13:08:59 -05:00
chenyu
c9371f0d31 hotfix llama conversation mode (#3031) 
without contiguous on keys and values, it runs but the update is incorrect
 2024-01-06 16:57:07 -05:00
chenyu
fa707c81e5 move beautiful cartpole action sampling inside jit (#3028) 
tested by getting 3 full scores in a row
 2024-01-06 00:39:55 -05:00
George Hotz
ebb81e8f11 hotfix: st.size() -> st.size in llama 2024-01-05 20:18:52 -08:00
George Hotz
a8ba1ac08f track size in shapetracker (#3026) 
* track size in shapetracker

* shapetracker adapter

* size is an int

* create Buffer with st.size

* only compare the views for the jit

* fix webgpu
 2024-01-05 20:15:53 -08:00
chenyu
138c17c094 enable argmax tests for METAL/WEBGPU in CI (#3027) 
not sure why it was skipped but works now in CI
 2024-01-05 21:43:00 -05:00
George Hotz
2a2d3233d2 add test that the compiler isn't used (#3025) 
* add test that the compiler isn't used

* one print_tree

* improve speed with st size cache

* switch to gpt-2
 2024-01-05 17:24:01 -08:00
chenyu
520406cf3a add Tensor.unflatten and Tensor.flatten(end_dim) (#3023) 
simplified cases when splitting a dim, or merge dims in predix
 2024-01-05 17:55:29 -05:00
George Hotz
f432ec9c33 Bitcast hip fix + fix mixtral (#3022) 
* fix bitcast in hip

* wrong dtype for precast, double COPY
 2024-01-05 14:51:25 -08:00
chenyu
eda43767de use Scalar = Union[float, int, bool] in tensor.py (#3021) 
unify the type spec for Tensor creation functions and broadcasted elementwise ops that take python scalar
 2024-01-05 13:56:26 -05:00
George Hotz
60abc62a3f fast hip read (#3014) 
* fast hip read

* hip read faster

* fix tests

* to_mv

* simplify

* bump to 6k lines
 2024-01-05 10:33:13 -08:00
chenyu
4465ef28c5 add test_softmax to test_ops (#3020) 
* add test_softmax to test_ops

somehow it was not tested

* too many buffers in softmax backward for WEBGPU
 2024-01-05 11:19:49 -05:00
chenyu
7c80b78be9 cleanup gpt2 build function (#3018) 2024-01-04 23:14:53 -05:00
chenyu
55e52abeba minor cleanup of matvec in hand_coded_optimizations (#3015) 
remove noop isinstance check and fix long lines
 2024-01-04 19:43:49 -05:00
chenyu
f88506e630 move gpt2/llama sampling inside the model call (#3013) 
* move gpt2/llama sampling inside the model call

* argmax uses one more kernel
 2024-01-04 17:01:50 -05:00
George Hotz
c2a044ed83 disk_read_speed example 2024-01-04 13:59:43 -08:00
Yixiang Gao
8a63f26a0f make LR scheduler work with multigpu (#3011) 
* add a failing test for LR scheduler when using multigpu

* fix calculation order and unnecessary tensor created for float

* min_lr is no longer tensor
 2024-01-04 12:10:56 -08:00
chenyu
8524493748 minor gpt2 cleanup (#3012) 2024-01-04 13:53:18 -05:00
chenyu
2b6670d2ea separate entry for HALF hlb_cifar10 in benchmark (#3010) 2024-01-04 13:24:10 -05:00
chenyu
5337211058 llvm CMPEQ 2024-01-04 13:12:22 -05:00
chenyu
b8c30eb358 no midcast MULACC for llvm 2024-01-04 13:12:22 -05:00
chenyu
91665ef143 rewrite MUL CAST SUM to CAST MULACC 2024-01-04 13:12:22 -05:00
chenyu
ab7dfd637b use float for acc dtype for half tensor sum 
we previously only upcast uint and int, and half was using half for acc.
change to acc in float for precision. but cast the result back to half to match torch/jax output dtype
 2024-01-04 13:12:22 -05:00
chenyu
6fa285b943 touchup onnx xor and not (#3008) 2024-01-04 02:02:42 -05:00
geohotstan
57817028bb removed redundant dtype hacks in onnx_ops (#2939) 
* updated most dtype hacks in onnx_ops

* temporarily revert dequantizelinear change

* I think this is right...

* MORE FIXES WOOOO NEW DTYPE IS AWESOME

* ok

* oops missed a print

* half -> float32 for CI

* is npdtype

* some more

* fix if ordering

* more clean ups

* final cleanups

* casting to half not allowed

* k nvm

* revert ArgMax change

* only GPU

* llvm begone

* teeny tiny change

* fix: attempt to add cast tests

* try this

* fix dequantizelinear

* revert some stuff

* tests pass pls

* less lines in onnx_tests

* oops missed string tensor tests

* clean up

* try: revert default behavior changes

* fix: disabled Cast and Castlike tests

* docs: small changes

* fix: fixed isNaN op and enabled associated tests

* fix: forgot about float16

* done

* update disabled test

* gah missed another float16

* disable rest of failing tests

* rm extra line

* try...

---------

Co-authored-by: chenyu <chenyu@fastmail.com>
 2024-01-04 01:45:24 -05:00
chenyu
9f39165188 correct (dtype, device) in test_dtype.is_dtype_supported (#3007) 
corrected dtypes for TORCH and float64 support
 2024-01-04 00:25:37 -05:00
chenyu
ae112c9dbe fix some long lines in tests (#3006) 
* fix some long lines in tests

* better
 2024-01-03 23:53:33 -05:00
George Hotz
7e191fbb86 hotfix: don't jitcache with 1 kernel. improvements to hip sniffer 2024-01-03 19:17:08 -08:00
George Hotz
bcc1aa21ac make disk simpler (#3002) 
* make disk simpler

* upd ops_disk

* works on osx too

* revert ops_hip
 2024-01-03 17:46:21 -08:00
George Hotz
9699c8c90b don't alloc for InterpretedASTRunner (#2999) 2024-01-03 17:05:53 -08:00
chenyu
bca0b95ee3 bump shapetracker simplify message to DEBUG >= 5 (#2998) 2024-01-03 20:00:36 -05:00
chenyu
74a30431b4 replace d[a] if a in d else b with d.get(a, b) (#2997) 2024-01-03 18:10:25 -05:00
chenyu
74cc6fd3c2 remove AndNode.__floordiv__ special case (#2996) 
* remove AndNode.__floordiv__

AndNode produces a Node that min/max is bounded by [0, 1] so `//` on top of that is almost always 0.
we don't really use that either

* keep the test
 2024-01-03 17:44:55 -05:00
George Hotz
a0c7cb2564 hotfix: create weights dir in local tg checkout 2024-01-03 14:14:33 -08:00
George Hotz
fc36a7d669 tinygrad weights 2024-01-03 14:09:28 -08:00
chenyu
1ac4d27869 remove VariableOrNum from Node.substitute arg (#2995) 
having NumNode in var_vals does not change the substitute output
 2024-01-03 17:02:25 -05:00
George Hotz
65dc3700b7 hip device is default on supported platforms (#2993) 2024-01-03 13:42:13 -08:00
George Hotz
77c98a1543 hotfix: remove weights directory 2024-01-03 13:40:39 -08:00
George Hotz
0be0f2f745 remove stable diffusion test on tinymac 2024-01-03 13:18:24 -08:00
George Hotz
a354ec9dad Revert "hotfix: HIP is the default device on HIP platforms" 
This reverts commit b748b569f5.
 2024-01-03 13:16:54 -08:00
George Hotz
b748b569f5 hotfix: HIP is the default device on HIP platforms 2024-01-03 13:13:52 -08:00
George Hotz
753a7ecc05 Hip driver (#2992) 
* start hip driver

* fix hip llama

* make HIP default if we can

* don't change those
 2024-01-03 12:53:47 -08:00
George Hotz
f290ca3924 hotfix: save lines in graph 2024-01-03 12:03:42 -08:00
Yixiang Gao
bc4b6e758b Merge pull request #2981 from g1y5x3/cifar_fp16 
adjsut div factor to avoid underflow for cifar in fp16
 2024-01-03 11:15:42 -08:00
George Hotz
d7d5a487ad hotfix: all device canonicalize should be done in Tensor 2024-01-03 10:48:04 -08:00
Yixiang Gao
ea3bc2f509 remove wino benchmark for now 2024-01-03 10:46:43 -08:00
Yixiang Gao
5663dd46b6 Merge branch 'master' of github.com:tinygrad/tinygrad into cifar_fp16 2024-01-03 10:11:46 -08:00