thek0tyara/llama-cpp-turboquant
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions 4

Updated push const readme

Browse source
This commit is contained in:
Alejandro Saucedo 2021-03-04 08:40:24 +00:00
parent 647f2f1e62
commit 0558a1b6fc
1 changed files with 9 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

15
README.md
View file

@ -71,13 +71,14 @@ void kompute(const std::string& shader) {
auto algorithm = mgr.algorithm(params,
kp::Shader::compile_source(shader),
workgroup,
specConsts);
specConsts,
pushConstsA);
// 4. Run operation synchronously using sequence
mgr.sequence()
->record<kp::OpTensorSyncDevice>(params)
->record<kp::OpAlgoDispatch>(algorithm, pushConstsA)
->record<kp::OpAlgoDispatch>(algorithm, pushConstsB)
->record<kp::OpAlgoDispatch>(algorithm) // Binds default push consts
->record<kp::OpAlgoDispatch>(algorithm, pushConstsB) // Overrides push consts
->eval();
// 5. Sync results from the GPU asynchronously
@ -155,13 +156,15 @@ def kompute(shader):
push_consts_a = [2]
push_consts_b = [3]
algo = mgr.algorithm(params, kp.Shader.compile_source(shader), workgroup, spec_consts)
spirv = kp.Shader.compile_source(shader)
algo = mgr.algorithm(params, spirv, workgroup, spec_consts, push_consts_a)
# 4. Run operation synchronously using sequence
(mgr.sequence()
.record(kp.OpTensorSyncDevice(params))
.record(kp.OpAlgoDispatch(algo, push_consts_a))
.record(kp.OpAlgoDispatch(algo, push_consts_b))
.record(kp.OpAlgoDispatch(algo)) # Binds default push consts provided
.record(kp.OpAlgoDispatch(algo, push_consts_b)) # Overrides push consts
.eval())
# 5. Sync results from the GPU asynchronously