Merge pull request #10 from da03/command_r_plus

add script for command-r-plus

scripts/cohere-command-r-plus.sh

@@ -0,0 +1,25 @@
+# export COHERE_API_KEY=your_cohere_api_key
+model_name="cohere/command-r-plus"
+model_pretty_name="command-r-plus"
+output_dir="result_dirs/wild_bench/"
+TEMP=0; TOP_P=1.0; MAX_TOKENS=2048;
+
+# shard_size should be 1024 // n_shards
+n_shards=8
+shard_size=128
+start_gpu=0
+shards_dir="${output_dir}/tmp_${model_pretty_name}"
+for ((start = 0, end = (($shard_size)), gpu = $start_gpu; gpu < $n_shards+$start_gpu; start += $shard_size, end += $shard_size, gpu++)); do
+    python src/unified_infer.py \
+        --data_name wild_bench \
+        --start_index $start --end_index $end \
+        --engine cohere \
+        --model_name $model_name \
+        --top_p $TOP_P --temperature $TEMP \
+        --max_tokens $MAX_TOKENS \
+        --output_folder $shards_dir/ \
+        --overwrite &
+done 
+wait 
+python src/merge_results.py $shards_dir/ $model_pretty_name
+cp $shards_dir/${model_pretty_name}.json $output_dir/${model_pretty_name}.json