[Eval] Use exact matching for Y/N and multi-choice when OPENAI_API_KE…

…Y not set (open-compass#44)

* update

* handle GPT_API_KEY missing

vlmeval/evaluate/multiple_choice.py

@@ -241,8 +241,11 @@ def multiple_choice_eval(eval_file, dataset=None, model='chatgpt-0613', nproc=4,
         model_name = 'gpt-3.5-turbo-0613'
         if INTERNAL:
             model = OpenAIWrapperInternal(model_name, verbose=verbose, retry=10)
-        else:
+        elif gpt_key_set():
             model = OpenAIWrapper(model_name, verbose=verbose, retry=10)
+        else:
+            logger.error('OPENAI_API_KEY is not set properly, will use exact matching for evaluation')
+            model = None
 
     logger.info(f'Evaluating {eval_file}')
     result_file = eval_file.replace(f'.{suffix}', f'_{name_str}_result.pkl')

vlmeval/evaluate/yes_or_no.py

@@ -175,18 +175,21 @@ def YOrN_eval(eval_file, model='chatgpt-0613', nproc=4, verbose=False, dataset=N
 
         if INTERNAL:
             model = OpenAIWrapperInternal(model_name, verbose=verbose, retry=10)
-        else:
+        elif gpt_key_set():
             model = OpenAIWrapper(model_name, verbose=verbose, retry=10)
-
-        lt = len(unknown)
-        lines = [unknown.iloc[i] for i in range(lt)]
-        tups = [(model, line) for line in lines]
-        indices = list(unknown['index'])
-
-        if len(tups):
-            res = track_progress_rich(YOrN_auxeval, tups, nproc=nproc, chunksize=nproc, keys=indices, save=tmp_file)
-            for k, v in zip(indices, res):
-                ans_map[k] = v
+        else:
+            logger.error('OPENAI_API_KEY is not set properly, will use exact matching for evaluation')
+            model = None
+
+        if model is not None:
+            lt = len(unknown)
+            lines = [unknown.iloc[i] for i in range(lt)]
+            tups = [(model, line) for line in lines]
+            indices = list(unknown['index'])
+            if len(tups):
+                res = track_progress_rich(YOrN_auxeval, tups, nproc=nproc, chunksize=nproc, keys=indices, save=tmp_file)
+                for k, v in zip(indices, res):
+                    ans_map[k] = v
 
         data['extracted'] = [ans_map[x] for x in data['index']]
         dump(data, storage)

vlmeval/smp.py

@@ -1,4 +1,4 @@
-    # flake8: noqa: F401, F403
+# flake8: noqa: F401, F403
 import abc
 import argparse
 import csv
@@ -29,6 +29,14 @@
 from huggingface_hub import scan_cache_dir
 import logging
 
+def gpt_key_set():
+    openai_key = os.environ.get('OPENAI_API_KEY', None)
+    return isinstance(openai_key, str) and openai_key.startswith('sk-')
+
+def apiok(wrapper):
+    s = wrapper.generate("Hello!")
+    return wrapper.fail_msg not in s
+
 def isimg(s):
     return osp.exists(s) or s.startswith('http')
 
@@ -219,24 +227,17 @@ def last_modified(pth):
 def mmqa_display(question):
     question = {k.lower(): v for k, v in question.items()}
     keys = list(question.keys())
-    if 'index' in keys:
-        keys.remove('index')
-    keys.remove('image')
+    keys = [k for k in keys if k not in ['index', 'image']]
 
     images = question['image']
     if isinstance(images, str):
         images = [images]
 
-    idx = 'XXX'
-    if 'index' in question:
-        idx = question.pop('index')
+    idx = question.pop('index', 'XXX')
     print(f'INDEX: {idx}')
 
     for im in images:
-        image = decode_base64_to_image(im)
-        w, h = image.size
-        ratio = 500 / h
-        image = image.resize((int(ratio * w), int(ratio * h)))
+        image = decode_base64_to_image(im, target_size=512)
         display(image)
 
     for k in keys:
@@ -289,13 +290,6 @@ def mwlines(lines, fname):
     with open(fname, 'w') as fout:
         fout.write('\n'.join(lines))
 
-def default_set(self, args, name, default):
-    if hasattr(args, name):
-        val = getattr(args, name)
-        setattr(self, name, val)
-    else:
-        setattr(self, name, default)
-
 def dict_merge(dct, merge_dct):
     for k, _ in merge_dct.items():
         if (k in dct and isinstance(dct[k], dict) and isinstance(merge_dct[k], dict)):  #noqa