DouDou commited on
Upload data3/download_and_process.sh with huggingface_hub
Browse files
data3/download_and_process.sh
ADDED
|
@@ -0,0 +1,64 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/bin/bash
|
| 2 |
+
# 下载和处理批处理结果
|
| 3 |
+
|
| 4 |
+
set -e
|
| 5 |
+
|
| 6 |
+
BATCH_ID_FILE="batch_id.txt"
|
| 7 |
+
BATCH_RESULTS_RAW="batch_results_raw.jsonl"
|
| 8 |
+
FINAL_OUTPUT="programming_problems_batch.jsonl"
|
| 9 |
+
MODEL="gpt-4o-mini"
|
| 10 |
+
|
| 11 |
+
# 获取 Batch ID
|
| 12 |
+
if [ -n "$1" ]; then
|
| 13 |
+
BATCH_ID=$1
|
| 14 |
+
elif [ -f "$BATCH_ID_FILE" ]; then
|
| 15 |
+
BATCH_ID=$(cat $BATCH_ID_FILE)
|
| 16 |
+
else
|
| 17 |
+
echo "❌ 错误: 请提供 Batch ID"
|
| 18 |
+
echo "用法: $0 <batch_id>"
|
| 19 |
+
exit 1
|
| 20 |
+
fi
|
| 21 |
+
|
| 22 |
+
echo "⬇️ 下载和处理批处理结果"
|
| 23 |
+
echo "========================================"
|
| 24 |
+
echo "Batch ID: $BATCH_ID"
|
| 25 |
+
echo ""
|
| 26 |
+
|
| 27 |
+
# 检查状态
|
| 28 |
+
echo "检查批处理状态..."
|
| 29 |
+
python3 generate_problems_batch.py status $BATCH_ID
|
| 30 |
+
|
| 31 |
+
echo ""
|
| 32 |
+
read -p "👉 确认下载? (y/n) " -n 1 -r
|
| 33 |
+
echo ""
|
| 34 |
+
if [[ ! $REPLY =~ ^[Yy]$ ]]; then
|
| 35 |
+
exit 0
|
| 36 |
+
fi
|
| 37 |
+
|
| 38 |
+
# 下载
|
| 39 |
+
echo ""
|
| 40 |
+
echo "⬇️ 下载结果..."
|
| 41 |
+
python3 generate_problems_batch.py download $BATCH_ID \
|
| 42 |
+
--output $BATCH_RESULTS_RAW
|
| 43 |
+
|
| 44 |
+
echo ""
|
| 45 |
+
echo "✅ 原始结果已下载: $BATCH_RESULTS_RAW"
|
| 46 |
+
|
| 47 |
+
# 处理
|
| 48 |
+
echo ""
|
| 49 |
+
echo "📊 处理结果..."
|
| 50 |
+
python3 generate_problems_batch.py process \
|
| 51 |
+
--input $BATCH_RESULTS_RAW \
|
| 52 |
+
--output $FINAL_OUTPUT \
|
| 53 |
+
--model $MODEL \
|
| 54 |
+
--requests batch_requests_full.jsonl
|
| 55 |
+
|
| 56 |
+
echo ""
|
| 57 |
+
echo "========================================"
|
| 58 |
+
echo "✅ 完成!"
|
| 59 |
+
echo "========================================"
|
| 60 |
+
echo "最终结果: $FINAL_OUTPUT"
|
| 61 |
+
echo ""
|
| 62 |
+
echo "查看结果:"
|
| 63 |
+
echo " wc -l $FINAL_OUTPUT"
|
| 64 |
+
echo " head -1 $FINAL_OUTPUT | python3 -m json.tool"
|