Merge pull request #104 from athewsey/feat/py310

Support Python 3.10 for SageMaker Studio

.gitignore

@@ -1,8 +1,22 @@
+# Results and artifacts:
+fmbench.log
+executed_notebooks/
+src/fmbench/results/
+analytics/results/
+results/
+results-*/
+
+# Python:
 __pycache__/
+.venv
+dist/*
+
+# Operating Systems:
+.DS_Store
+
 *.zip
 *.zip
 e.json
-.DS_Store
 .locks/*
 ./scripts/model.tar.gz
 /data/models/*/endpoints.json
@@ -18,22 +32,14 @@ llama2_tokenizer/*
 **/endpoint_path.txt
 **/results_path.txt
 metadata/*
-**/DS_Store
-.DS_Store
 .fmbt/*
-dist/*
 tokenizer/*
-src/fmbench/results/
 src/fmbench/metadata/
 src/fmbench/distilbert-base-uncased-v1-SageMaker/
 distilbert-base-uncased-v1-SageMaker/
-results/
 llama2-7b-v1-SageMaker/
-fmbench.log
 llama2-13b-inf2-g5-p4d-v1-SageMaker/
 llama2-70b-g5-p4d-trt-v1-SageMaker/
 Llama2-7b-g4dn-SageMaker/
 Llama2-7b-g4dn-g5-SageMaker/
 llama2-70b-g5-inf2-tgi-v1-SageMaker/
-results-*/
-analytics/results/

copy_s3_content.sh

@@ -1,23 +1,24 @@
+#!/bin/bash
+set -e
+
 # this scripts creates a local directory for running FMBench
 # without any s3 dependency and copies all relevant files
 # for public FMBench content
 FMBENCH_READ_DIR=/tmp/fmbench-read
 FMBENCH_WRITE_DIR=/tmp/fmbench-write
 BUCKET=aws-blogs-artifacts-public
 
-mkdir $FMBENCH_READ_DIR
+mkdir -p $FMBENCH_WRITE_DIR
 mkdir -p $FMBENCH_READ_DIR/tokenizer
 mkdir -p $FMBENCH_READ_DIR/llama2_tokenizer
 mkdir -p $FMBENCH_READ_DIR/llama3_tokenizer
 mkdir -p $FMBENCH_READ_DIR/mistral_tokenizer
-wget https://${BUCKET}.s3.amazonaws.com/artifacts/ML-FMBT/manifest.txt -P ${FMBENCH_READ_DIR}/
+curl --output-dir ${FMBENCH_READ_DIR}/ -O https://${BUCKET}.s3.amazonaws.com/artifacts/ML-FMBT/manifest.txt
 
 # copy each file of the public content for FMBench
 for i in `cat ${FMBENCH_READ_DIR}/manifest.txt`
 do
   dir_path=`dirname $i`
   mkdir -p ${FMBENCH_READ_DIR}/$dir_path
-  wget https://${BUCKET}.s3.amazonaws.com/artifacts/ML-FMBT/$i -P ${FMBENCH_READ_DIR}/$dir_path
+  curl --output-dir ${FMBENCH_READ_DIR}/$dir_path -O https://${BUCKET}.s3.amazonaws.com/artifacts/ML-FMBT/$i
 done
-
-

debug.sh

@@ -1,18 +1,22 @@
+#!/bin/bash
 # script for a debug/developer workflow
 # 1. Deletes the existing pfmbench package from the conda env
 # 2. Builds and installs a new one
 # 3. Runs fmbench as usual
 
+set -e
+
 CONDA_ENV_PATH=~/anaconda3/envs/fmbench_python311/lib/python3.11/site-packages
-CONFIG_FILE_PATH=src/fmbench/configs/llama2/7b/config-llama2-7b-g5-quick.yml
+CONFIG_FILE_PATH=src/fmbench/configs/llama2/7b/config-llama2-7b-g5-no-s3-quick.yml
 LOGFILE=fmbench.log
 
-# delete existing install
+# delete existing install and built artifacts
 rm -rf $CONDA_ENV_PATH/fmbench*
+rm -rf dist
 
 # build a new version
 poetry build
-pip install -U dist/*.whl
+pip install -U --force-reinstall dist/*.whl
 
 # run the newly installed version
 echo "going to run fmbench now"