Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

jump to next gen, three modes #1571

Merged
merged 3 commits into from
Nov 20, 2024
Merged

jump to next gen, three modes #1571

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6711434
jump to next gen, three modes
kali Nov 18, 2024
ae56a71
include embeddings variants
kali Nov 18, 2024
fb7b8f1
fix script for e variants
kali Nov 18, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/workflows/llm.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,7 @@ jobs:
matrix:
os: [ macos, ubuntu ]
model: [ OpenELM-270M, OpenELM-1_1B, TinyLlama_v1.1, phi-1_5 ]
q: [ f16f16, f32f32, q40f16, q40f32 ]
q: [ f16f16, f32f32, q40f16, q40f32, q40ef16, q40ef32 ]
fail-fast: false
permissions:
id-token: write
Expand Down
25 changes: 12 additions & 13 deletions .travis/test-llm.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,24 +30,23 @@ esac

case $q in
q40f16) approx=ultra;;
q40ef16) approx=ultra;;
f16f16) approx=ultra;;
q40f32) approx=very;;
q40ef32) approx=very;;
f32f32) approx=approximate;;
esac

nnef=llm/$generation/$id/$id.nnef.tgz
pp=llm/$generation/$id/$id.pp.io.npz
tg=llm/$generation/$id/$id.tg.io.npz

set -x
$CACHE_FILE $nnef $pp $tg

$TRACT_RUN -v --nnef-tract-core $MODELS/$nnef -O run \
--input-from-npz $MODELS/$pp \
--assert-output-bundle $MODELS/$pp \
--approx $approx --allow-float-casts

$TRACT_RUN -v --nnef-tract-core $MODELS/$nnef -O run \
--input-from-npz $MODELS/$tg \
--assert-output-bundle $MODELS/$tg \
--approx $approx --allow-float-casts
$CACHE_FILE $nnef
for t in p0s100 p50s50 p99s1
do
npz=llm/$generation/$id/$id.$t.io.npz
$CACHE_FILE $npz
$TRACT_RUN -v --nnef-tract-core $MODELS/$nnef -O run \
--input-from-npz $MODELS/$npz \
--assert-output-bundle $MODELS/$npz \
--approx $approx --allow-float-casts
done
Loading