mirrors/tinygrad
2
0
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-06-24 02:14:17 +00:00
Code Issues Projects Releases Packages Wiki Activity

valid

Browse source
This commit is contained in:
George Hotz 2026-06-23 17:26:11 -07:00
commit f2c93fd0d4
1 changed files with 2 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

4
extra/llama_kernels/quantize_fp8_delayed/__init__.py
View file

@ -42,8 +42,8 @@ def _custom_quantize_fp8_with_amax(fp8_out:UOp, amax_partial:UOp, x:UOp, amax_st
step = THREADS_PER_WG // 2
while step:
active = tid < step
other = lds[active.where(tid + step, UOp.invalid())].load()
lds = lds.after(lds[active.where(tid, UOp.invalid())].store(lds[tid].maximum(other)).barrier())
other = lds[(tid + step).valid(active)].load()
lds = lds.after(lds[tid.valid(active)].store(lds[tid].maximum(other)).barrier())
step //= 2
amax_store = amax_partial[tid.eq(0).where(wg, UOp.invalid())].store(lds[0])