WIP: add DAMO-NLP-MT/polylm-13b

Author: fujiki-1emon

models/DAMO-NLP-MT/polylm-13b/harness.sh

@@ -0,0 +1,16 @@
+#!/bin/bash
+
+PRETRAINED="DAMO-NLP-MT/polylm-13b"
+TOKENIZER="DAMO-NLP-MT/polylm-13b,use_fast=False"
+MODEL_ARGS="pretrained=${PRETRAINED},tokenizer=${TOKENIZER}"
+TASK="jcommonsenseqa-1.1-0.2,jnli-1.1-0.2,marc_ja-1.1-0.2,jsquad-1.1-0.2,jaqket_v2-0.1-0.2,xlsum_ja,xwinograd_ja,mgsm"
+NUM_FEWSHOT="3,3,3,2,1,1,0,5"
+OUTPUT_PATH="models/DAMO-NLP-MT/polylm-13b/result.json"
+python main.py \
+    --model hf-causal \
+    --model_args $MODEL_ARGS \
+    --tasks $TASK \
+    --num_fewshot $NUM_FEWSHOT \
+    --device "cuda" \
+    --no_cache \
+    --output_path $OUTPUT_PATH