Merge pull request #28 from iryna-kondr/hotfix_ml_classifier

Improved multi-label classifier compatibility with gpt4all

Author: iryna-kondr

skllm/models/gpt_zero_shot_clf.py

@@ -98,16 +98,11 @@ def _get_prompt(self, x) -> str:
     def _predict_single(self, x):
         completion = self._get_chat_completion(x)
         try:
-            if self.openai_model.startswith("gpt4all::"):
-                label = str(
-                    extract_json_key(
-                        completion["choices"][0]["message"]["content"], "label"
-                    )
-                )
-            else:
-                label = str(
-                    extract_json_key(completion.choices[0].message["content"], "label")
+            label = str(
+                extract_json_key(
+                    completion["choices"][0]["message"]["content"], "label"
                 )
+            )
         except Exception as e:
             print(completion)
             print(f"Could not extract the label from the completion: {str(e)}")
@@ -155,10 +150,13 @@ def _get_prompt(self, x) -> str:
     def _predict_single(self, x):
         completion = self._get_chat_completion(x)
         try:
-            labels = extract_json_key(completion.choices[0].message["content"], "label")
+            labels = extract_json_key(completion["choices"][0]["message"]["content"], "label")
             if not isinstance(labels, list):
-                raise RuntimeError("Invalid labels type, expected list")
-        except Exception:
+                labels = labels.split(",")
+                labels = [l.strip() for l in labels]
+        except Exception as e:
+            print(completion)
+            print(f"Could not extract the label from the completion: {str(e)}")
             labels = []
 
         labels = list(filter(lambda l: l in self.classes_, labels))

skllm/openai/chatgpt.py

@@ -34,12 +34,18 @@ def get_chat_completion(messages, key, org, model="gpt-3.5-turbo", max_retries=3
 
 
 def extract_json_key(json_, key):
-    try:
-        json_ = json_.replace("\n", "")
-        json_ = find_json_in_string(json_)
-        as_json = json.loads(json_)
-        if key not in as_json.keys():
-            raise KeyError("The required key was not found")
-        return as_json[key]
-    except Exception:
-        return None
+    original_json = json_
+    for i in range(2):
+        try:
+            json_ = original_json.replace("\n", "")
+            if i == 1:
+                json_ = json_.replace("'", "\"")
+            json_ = find_json_in_string(json_)
+            as_json = json.loads(json_)
+            if key not in as_json.keys():
+                raise KeyError("The required key was not found")
+            return as_json[key]
+        except Exception:
+            if i == 0:
+                continue
+            return None

skllm/prompts/templates.py

@@ -44,7 +44,7 @@
 Perform the following tasks:
 1. Identify to which categories the provided text belongs to with the highest probability.
 2. Assign the text sample to at least 1 but up to {max_cats} categories based on the probabilities.
-3. Provide your response in a JSON format containing a single key `label` and a value corresponding to the list of assigned categories. Do not provide any additional information except the JSON.
+3. Provide your response in a JSON format containing a single key `label` and a value corresponding to the array of assigned categories. Do not provide any additional information except the JSON.
 
 List of categories: {labels}