nimlgenandGitHub 5c7d004da5 hcq: refactor int ptrs to hcqbuffers (#10105) ...

* hcq: refactor int ptrs to hcqbuffers

* more refactors

* linter

* use in allocator

* test fiz

* fx

* ops

* final?

* simpler

* keep this for now

2025-04-30 00:12:18 +03:00

..

codegen

fix ptx linearizer bug 2 [pr] (#9967)

2025-04-29 14:30:07 -03:00

engine

names for graph rewrites + null device supports exp and friends (#10106)

2025-04-29 14:28:20 -04:00

frontend

add onnx frontend stub [pr] (#9558)

2025-03-24 12:24:34 +08:00

nn

load fast in sdxl (#10072)

2025-04-27 11:58:51 -04:00

renderer

range has only one src now [pr] (#10100)

2025-04-29 10:31:05 -04:00

runtime

hcq: refactor int ptrs to hcqbuffers (#10105)

2025-04-30 00:12:18 +03:00

shape

names for graph rewrites + null device supports exp and friends (#10106)

2025-04-29 14:28:20 -04:00

viz

display viz rewrites with tabbing if they are subrewrites (#10097)

2025-04-29 17:57:21 +08:00

__init__.py

UOp.variable (#7010)

2024-10-12 18:20:44 +08:00

device.py

hotfix amd bf16 is supported case (#10039)

2025-04-24 21:29:27 -03:00

dtype.py

FP8s truncate (#9937)

2025-04-22 19:12:49 -04:00

gradient.py

toposort is a function [pr] (#10004)

2025-04-23 16:25:03 +01:00

helpers.py

[bounty] [pr] index validation with z3 (#9981)

2025-04-24 08:06:08 -04:00

ops.py

range has only one src now [pr] (#10100)

2025-04-29 10:31:05 -04:00

py.typed

add a single py.typed (#6083)

2024-08-14 17:31:46 -07:00

spec.py

range has only one src now [pr] (#10100)

2025-04-29 10:31:05 -04:00

tensor.py

test sdxl softmax (#10096)

2025-04-28 21:55:50 -04:00

upat.py

don't use ops name, just keep sink (#9922)

2025-04-18 08:59:18 +01:00