update scripts and doc

supercoderhawk · Jan 16, 2020 · ad46061 · ad46061
1 parent 9b3be83
commit ad46061
Show file tree

Hide file tree

Showing 4 changed files with 16 additions and 22 deletions.
diff --git a/ReadMe.md b/ReadMe.md
@@ -18,7 +18,7 @@ python3 -m spacy download en
 2. setting value `ES_BASE_URL` in constants.py with your  configured elastic search endpoint.
 
 ### Prepare Data
-1. unzip file and put all files under `data/` folder
+1. unzip file and put all files under `data/` folder, rename `test.csv` to `test_release.csv`
 2. execute `bash scripts/prepare_data.sh` in **project root folder** to build the data for next step
 
 ### Execute the retrieval process

diff --git a/scripts/run_end2end.sh b/scripts/run_end2end.sh
@@ -1,12 +1,18 @@
 #!/bin/bash
 
+DATA_DIR=${PWD}/data/
+ES_RESULT_FILE=$DATA_DIR/validation_es_result.jsonl
+FINAL_RESULT_FILENAME=$DATA_DIR/validation_final_result.jsonl
+MODEL_PATH=$MODEL_DIR/rerank_model.model
+TOPK=20
 
-# process raw data into jsonl
-echo 'starting build raw data...'
+# run elasticsearch (BM25)
+python3 wsdm_digg/benchmark/benchmarker.py -src_filename $DATA_DIR/validation.jsonl \
+      -dest_filename $ES_RESULT_FILE
 
-# build elasticsearch index
-echo 'starting building elasticsearch indexing...'
-
-# execute recall stage
-
-# execute reranking stage
+# run rerank by bert
+python3 wsdm_digg/reranking/predict.py -eval_search_filename $ES_RESULT_FILE \
+  -golden_filename $VALID_FILE \
+  -dest_filename $RESULT_DIR/$FINAL_RESULT_FILENAME \
+  -model_path $MODEL_PATH \
+  -eval_batch_size 10 -topk $TOPK
diff --git a/scripts/run_retrieval.sh b/scripts/run_retrieval.sh
diff --git a/wsdm_digg/benchmark/benchmarker.py b/wsdm_digg/benchmark/benchmarker.py
@@ -100,7 +100,7 @@ def main():
     parser.add_argument('-dest_filename', type=str, required=True, )
     parser.add_argument('-batch_size', type=int, default=100)
     parser.add_argument('-parallel_count', type=int, default=20)
-    parser.add_argument('-top_n', type=int, default=20)
+    parser.add_argument('-top_n', type=int, default=100)
     parser.add_argument('-is_submit', action='store_true')
     args = parser.parse_args()