update eval fn

shibing624 · Oct 13, 2024 · 32192a5 · 32192a5
1 parent 49fc55d
commit 32192a5
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 7 deletions.
diff --git a/examples/evaluate_models/evaluate_models.py b/examples/evaluate_models/evaluate_models.py
@@ -84,14 +84,16 @@ def main(args):
                          model_type='chatglm',
                          peft_name="shibing624/chatglm3-6b-csc-chinese-lora")
         if args.data == 'sighan':
-            eval_model_batch(m.correct_batch)
+            eval_model_batch(m.correct_batch, prompt_template_name='vicuna')
             # Sentence Level: acc:0.5564, precision:0.5574, recall:0.4917, f1:0.5225, cost time:1572.49 s, total num: 1100
             #
         elif args.data == 'ec_law':
-            eval_model_batch(m.correct_batch, input_tsv_file=os.path.join(pwd_path, "../data/ec_law_test.tsv"))
+            eval_model_batch(m.correct_batch, input_tsv_file=os.path.join(pwd_path, "../data/ec_law_test.tsv"),
+                             prompt_template_name='vicuna')
             #
         elif args.data == 'mcsc':
-            eval_model_batch(m.correct_batch, input_tsv_file=os.path.join(pwd_path, "../data/mcsc_test.tsv"))
+            eval_model_batch(m.correct_batch, input_tsv_file=os.path.join(pwd_path, "../data/mcsc_test.tsv"),
+                             prompt_template_name='vicuna')
             #
     elif args.model == 'qwen1.5b':
         from pycorrector.gpt.gpt_corrector import GptCorrector

diff --git a/pycorrector/utils/evaluate_utils.py b/pycorrector/utils/evaluate_utils.py
@@ -12,7 +12,7 @@
 sighan_2015_path = os.path.join(pwd_path, '../data/sighan2015_test.tsv')
 
 
-def eval_model_single(correct_fn, input_tsv_file=sighan_2015_path, verbose=True):
+def eval_model_single(correct_fn, input_tsv_file=sighan_2015_path, verbose=True, **kwargs):
     """
     SIGHAN句级评估结果，设定需要纠错为正样本，无需纠错为负样本
     Args:
@@ -40,7 +40,7 @@ def eval_model_single(correct_fn, input_tsv_file=sighan_2015_path, verbose=True)
             src = parts[0]
             tgt = parts[1]
 
-            r = correct_fn(src)
+            r = correct_fn(src, **kwargs)
             tgt_pred, pred_detail = r['target'], r['errors']
             if verbose:
                 print()
@@ -80,7 +80,7 @@ def eval_model_single(correct_fn, input_tsv_file=sighan_2015_path, verbose=True)
         return acc, precision, recall, f1
 
 
-def eval_model_batch(correct_fn, input_tsv_file=sighan_2015_path, verbose=True):
+def eval_model_batch(correct_fn, input_tsv_file=sighan_2015_path, verbose=True, **kwargs):
     """
     SIGHAN句级评估结果，设定需要纠错为正样本，无需纠错为负样本
     Args:
@@ -113,7 +113,7 @@ def eval_model_batch(correct_fn, input_tsv_file=sighan_2015_path, verbose=True):
             srcs.append(src)
             tgts.append(tgt)
 
-    res = correct_fn(srcs)
+    res = correct_fn(srcs, **kwargs)
     for each_res, src, tgt in zip(res, srcs, tgts):
         pred_detail = ''
         if isinstance(each_res, str):