George HotzandGitHub 38fe84d92b cleanup mlops (#1521) ...

* cleanup mlops

* that line belongs there

2023-08-10 19:53:28 -07:00

..

codegen

use ConstOp for valid.max == 0 (#1501)

2023-08-09 00:01:59 -07:00

nn

Add assert for wrong matmul/dot shapes (#1438)

2023-08-04 18:16:56 -04:00

renderer

llvmir support for bool <-> float casting (#1492)

2023-08-09 13:12:52 -04:00

runtime

just cmplt (#1493)

2023-08-08 13:58:10 -07:00

shape

simplify Node.sum aggregating (#1449)

2023-08-05 22:19:36 -07:00

graph.py

Add WHERE ternary (or trinary?) op (#1196)

2023-07-16 00:31:55 -07:00

helpers.py

simple bitcast 2 (#1445)

2023-08-06 00:30:50 -07:00

jit.py

Enable JIT tests for supported devices, skip METAL and WEBGPU (#1265)

2023-07-18 11:40:37 -07:00

lazy.py

cleanup mlops (#1521)

2023-08-10 19:53:28 -07:00

mlops.py

cleanup mlops (#1521)

2023-08-10 19:53:28 -07:00

ops.py

just cmplt (#1493)

2023-08-08 13:58:10 -07:00

state.py

add stable diffusion and llama (#1471)

2023-08-06 21:31:51 -07:00

tensor.py

Tensor.scaled_dot_product_attention to match torch, used in LLaMA, and tested (#1502)

2023-08-08 23:27:13 -07:00