github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-01-09 06:58:11 -05:00
Code Issues Packages Projects Releases Wiki Activity

llama + beam to mac benchmark, full cifar to nvidia benchmark (#3612)

Browse Source 
would merge if it's also ~1 minute. btw why is gpt2 beam not slower in the first beam run?
This commit is contained in:
chenyu
2024-03-04 21:35:57 -05:00
committed by GitHub
parent 282bbd5acb
commit 957e9800f1
1 changed files with 6 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
.github/workflows/benchmark.yml vendored
View File

@@ -42,6 +42,8 @@ jobs:
run: |
JIT=0 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_unjitted.txt
JIT=1 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_jitted.txt
- name: Run LLaMA with BEAM
run: JIT=1 BEAM=2 CACHELEVEL=0 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_beam.txt
- name: Run GPT2
run: |
JIT=0 python3 examples/gpt2.py --prompt "Hello." --count 10 --temperature 0 --timing | tee gpt2_unjitted.txt
@@ -65,6 +67,7 @@ jobs:
train_cifar_wino.txt
llama_unjitted.txt
llama_jitted.txt
llama_beam.txt
gpt2_unjitted.txt
gpt2_jitted.txt
gpt2_half.txt
@@ -101,6 +104,8 @@ jobs:
run: CUDA=1 JIT=1 HALF=1 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half.txt
- name: Run GPT2 w HALF/BEAM
run: CUDA=1 JIT=1 HALF=1 BEAM=2 CACHELEVEL=0 CAST_BEFORE_VIEW=0 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half_beam.txt
- name: Run full CIFAR training
run: time CUDA=1 HALF=1 LATEWINO=1 STEPS=1000 TARGET_EVAL_ACC_PCT=93 python3 examples/hlb_cifar10.py | tee train_cifar_one_gpu.txt
- uses: actions/upload-artifact@v4
with:
name: Speed (NVIDIA)
@@ -112,6 +117,7 @@ jobs:
gpt2_jitted.txt
gpt2_half.txt
gpt2_half_beam.txt
train_cifar_one_gpu.txt
testamdbenchmark:
name: tinybox Benchmark