pytorch/benchmark

Pull Requests Commits

FindHao committed 2 years ago

27ae2389

Merge branch 'main' into findhao/add-api-coverage-test

FindHao committed 2 years ago

8344b780

FIX SAM for bfloat16 (#1764)

msaroufim committed 2 years ago

745644f3

remove wrong reference to __init__.py (#1746)

janeyx99 committed 2 years ago

2ea018e9

export chrome trace when profiler is enabled (#1760)

dshi7 committed 2 years ago

bfb6d156

fix broken torchbench mimo_cmf_30x

dshi7 committed 2 years ago

48a8ef6d

Move lit_llama and variants to canary_models (#1761)

ezyang committed 2 years ago

16e235ea

Add test runner on ibm cloud (#1755)

spzala committed 2 years ago

3df1cc93

Add lit-llama benchmarks (logits, autoregressive generation, lora fine tuning) (#1730)

ezyang committed 2 years ago

02ff72b4

Add special device mapping (#1752)

FindHao committed 2 years ago

9df52151

Increase the worker timeout to 1800 seconds

xuzhao9 committed 2 years ago

ae3534ab

Fix broken unit test (#1743)

xuzhao9 committed 2 years ago

4e9d86ec

Add boto3 to requirements.txt (#1745)

Thiago Crepaldi committed 2 years ago

5b2a70ea

Refactor Execution Graph (EG) to Execution Trace (ET) (#1566)

louisfeng committed 2 years ago

6d84522e

Add dependency boto3 to readme.md (#1741)

FindHao committed 2 years ago

eadfa982

Add back video_watch

xuzhao9 committed 2 years ago

b90583ac

Add GCP A100 bisection workflow (#1742)

xuzhao9 committed 2 years ago

81508866

Minor fix of the nvidia driver install (#1740)

xuzhao9 committed 2 years ago

84bac2a9

Upgrade the NVidia kernel driver (#1739)

xuzhao9 committed 2 years ago

2b9de6b0

Fix leftover __init__ to run rename mishap in userbenchmark/optim (#1738)

janeyx99 committed 2 years ago

678b7f29

Print complete traceback if unit test model import error. (#1737)

xuzhao9 committed 2 years ago

d67252c2

Fix the warnings from deprecating APIs or missing environment variables (#1735)

xuzhao9 committed 2 years ago

d3eab4bb

Download the input data ondemand (#1734)

xuzhao9 committed 2 years ago

e5bf5f0c

Add PT2 compilation time metric and remove eager-speedup metric

xuzhao9 committed 2 years ago

ebb6784d

Update python_utils.py (#1718)

johnnynunez committed 2 years ago

efc8d060

Bump transformers from 4.27.4 to 4.30.0 (#1731)

dependabot[bot] committed 2 years ago

790e646b

Add throughputs metric and enable in cpu userbenchmark (#1727)

chuanqi129 committed 2 years ago

7619bb8c

replace np.int with int [2]

igorsugak committed 2 years ago

b50703d5

Add model kwargs (#1729)

anijain2305 committed 2 years ago

72dcbd91

Fix bisection failures (#1722)

xuzhao9 committed 2 years ago

b0f75ea8

Older