Update chat example, prompt formats

turboderp-org · Nov 30, 2024 · 48e6306 · 48e6306
1 parent 1f685bd
commit 48e6306
Show file tree

Hide file tree

Showing 2 changed files with 103 additions and 58 deletions.
diff --git a/examples/chat.py b/examples/chat.py
@@ -40,6 +40,7 @@
 parser.add_argument("-un", "--username", type = str, default = "User", help = "Username when using raw chat mode")
 parser.add_argument("-bn", "--botname", type = str, default = "Chatbort", help = "Bot name when using raw chat mode")
 parser.add_argument("-sp", "--system_prompt", type = str, help = "Use custom system prompt")
+parser.add_argument("-nsp", "--no_system_prompt", action = "store_true", help = "Do not use any system prompt")
 
 parser.add_argument("-temp", "--temperature", type = float, default = 0.95, help = "Sampler temperature, default = 0.95 (1 to disable)")
 parser.add_argument("-smooth", "--smoothing_factor", type = float, default = 0.0, help = "Smoothing Factor, default = 0.0 (0 to disable")
@@ -90,6 +91,8 @@
 username = args.username
 botname = args.botname
 system_prompt = args.system_prompt
+if args.no_system_prompt:
+    system_prompt = ""
 
 if args.mode is None:
     print(" ## Error: No mode specified.")
@@ -185,7 +188,7 @@ def format_prompt(user_prompt, first):
     global system_prompt, prompt_format
 
     if first:
-        return prompt_format.first_prompt() \
+        return prompt_format.first_prompt(not system_prompt) \
             .replace("<|system_prompt|>", system_prompt) \
             .replace("<|user_prompt|>", user_prompt)
     else:
@@ -288,9 +291,10 @@ def get_tokenized_context(max_len):
 # Main loop
 
 print(f" -- Prompt format: {args.mode}")
-print(f" -- System prompt:")
-print()
-print(col_sysprompt + system_prompt.strip() + col_default)
+if system_prompt:
+    print(f" -- System prompt:")
+    print()
+    print(col_sysprompt + system_prompt.strip() + col_default)
 
 while True:
 

diff --git a/examples/chat_prompts.py b/examples/chat_prompts.py
@@ -12,7 +12,7 @@ def __init__(self):
     def default_system_prompt(self):
         raise NotImplementedError
 
-    def first_prompt(self):
+    def first_prompt(self, sysprompt: bool):
         raise NotImplementedError
 
     def subs_prompt(self):
@@ -44,9 +44,11 @@ def default_system_prompt(self):
             f"""This is a conversation between a helpful AI assistant named {self.botname} and a """ + \
             (f"""user named {self.username}.""" if self.username != "User" else """user.""")
 
-    def first_prompt(self):
-        return \
-            f"""<|system_prompt|>\n{self.username}: <|user_prompt|>\n{self.botname}:"""
+    def first_prompt(self, sysprompt):
+        if sysprompt:
+            return f"""<|system_prompt|>\n{self.username}: <|user_prompt|>\n{self.botname}:"""
+        else:
+            return f"""{self.username}: <|user_prompt|>\n{self.botname}:"""
 
     def subs_prompt(self):
         return \
@@ -61,7 +63,7 @@ def stop_conditions(self, tokenizer):
              tokenizer.eos_token_id]
 
     def encoding_options(self):
-        return False, False, False
+        return True, False, False
 
     def print_bot_name(self):
         return True
@@ -81,9 +83,11 @@ def default_system_prompt(self):
             """Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. """ + \
             """Please ensure that your responses are socially unbiased and positive in nature."""
 
-    def first_prompt(self):
-        return \
-            """[INST] <<SYS>>\n<|system_prompt|>\n<</SYS>>\n\n<|user_prompt|> [/INST]"""
+    def first_prompt(self, sysprompt):
+        if sysprompt:
+            return """[INST] <<SYS>>\n<|system_prompt|>\n<</SYS>>\n\n<|user_prompt|> [/INST]"""
+        else:
+            return """[INST] <|user_prompt|> [/INST]"""
 
     def subs_prompt(self):
         return \
@@ -115,19 +119,23 @@ def default_system_prompt(self):
             """to find the answer or suggest where to find it. Keep responses concise and relevant. Follow ethical """ + \
             """guidelines and promote a safe and respectful interaction."""
 
-    def first_prompt(self):
-        return \
-            """<|start_header_id|>system<|end_header_id|>\n\n""" + \
-            """<|system_prompt|><|eot_id|>""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += \
+                """<|start_header_id|>system<|end_header_id|>\n\n""" + \
+                """<|system_prompt|><|eot_id|>"""
+        r += \
             """<|start_header_id|>user<|end_header_id|>\n\n""" + \
             """<|user_prompt|><|eot_id|>""" + \
-            """<|start_header_id|>assistant<|end_header_id|>"""
+            """<|start_header_id|>assistant<|end_header_id|>\n\n"""
+        return r
 
     def subs_prompt(self):
         return \
             """<|eot_id|><|start_header_id|>user<|end_header_id|>\n\n""" + \
             """<|user_prompt|><|eot_id|>""" + \
-            """<|start_header_id|>assistant<|end_header_id|>"""
+            """<|start_header_id|>assistant<|end_header_id|>\n\n"""
 
     def stop_conditions(self, tokenizer):
         return \
@@ -136,7 +144,7 @@ def stop_conditions(self, tokenizer):
              tokenizer.single_id("<|start_header_id|>")]
 
     def encoding_options(self):
-        return False, False, True
+        return True, False, True
 
     def print_extra_newline(self):
         return True
@@ -154,14 +162,18 @@ def default_system_prompt(self):
         return \
             """You are a helpful AI assistant."""
 
-    def first_prompt(self):
-        return \
-            """<s><|system|>\n""" + \
-            """<|system_prompt|>""" + \
-            """<|end|>\n""" + \
+    def first_prompt(self, sysprompt):
+        r = """<s>"""
+        if sysprompt:
+            r += \
+                """<|system|>\n""" + \
+                """<|system_prompt|>""" + \
+                """<|end|>\n"""
+        r += \
             """<|user|>\n""" + \
             """<|user_prompt|><|end|>\n""" + \
             """<|assistant|>\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -210,14 +222,18 @@ def default_system_prompt(self):
         return \
             f"""You are {self.botname}, a large language model. Answer as concisely as possible."""
 
-    def first_prompt(self):
-        return \
-            """<|im_start|>system\n""" + \
-            """<|system_prompt|>\n""" + \
-            """<|im_end|>\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += \
+                """<|im_start|>system\n""" + \
+                """<|system_prompt|>\n""" + \
+                """<|im_end|>\n"""
+        r += \
             """<|im_start|>user\n""" + \
             """<|user_prompt|><|im_end|>\n""" + \
             """<|im_start|>assistant\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -259,14 +275,18 @@ def default_system_prompt(self):
         return \
             f"""You are {self.botname}, a large language model. Answer as concisely as possible."""
 
-    def first_prompt(self):
-        return \
-            """<|system|>\n""" + \
-            """<|system_prompt|>\n""" + \
-            """</s>\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += \
+                """<|system|>\n""" + \
+                """<|system_prompt|>\n""" + \
+                """</s>\n"""
+        r += \
             """<|user|>\n""" + \
             """<|user_prompt|></s>\n""" + \
             """<|assistant|>\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -299,12 +319,15 @@ def default_system_prompt(self):
         return \
             f"""You are an AI programming assistant, utilizing the Deepseek Coder model, developed by Deepseek Company, and you only answer questions related to computer science. For politically sensitive questions, security and privacy issues, and other non-computer science questions, you will refuse to answer."""
 
-    def first_prompt(self):
-        return \
-            """<|system_prompt|>\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += """<|system_prompt|>\n"""
+        r += \
             """### Instruction:\n""" + \
             """<|user_prompt|>\n""" + \
             """### Response:\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -335,13 +358,17 @@ def default_system_prompt(self):
         return \
             f"""You are an AI assistant."""
 
-    def first_prompt(self):
-        return \
-            """### System\n""" + \
-            """<|system_prompt|>\n\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += \
+                """### System\n""" + \
+                """<|system_prompt|>\n\n"""
+        r += \
             """### User:\n""" + \
             """<|user_prompt|>\n\n""" + \
             """### Assistant:\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -374,9 +401,13 @@ def default_system_prompt(self):
         return \
             f"""You are an AI assistant."""
 
-    def first_prompt(self):
-        return \
-            """<|system_prompt|><|end_of_turn|>GPT4 Correct User:<|user_prompt|><|end_of_turn|>GPT4 Correct Assistant:"""
+    def first_prompt(self, sysprompt):
+        if sysprompt:
+            return \
+                """<|system_prompt|><|end_of_turn|>GPT4 Correct User:<|user_prompt|><|end_of_turn|>GPT4 Correct Assistant:"""
+        else:
+            return \
+                """GPT4 Correct User:<|user_prompt|><|end_of_turn|>GPT4 Correct Assistant:"""
 
     def subs_prompt(self):
         return \
@@ -408,12 +439,15 @@ def default_system_prompt(self):
         return \
             f"""Perform the task to the best of your ability."""
 
-    def first_prompt(self):
-        return \
-            """<|system_prompt|>\n\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += """<|system_prompt|>\n\n"""
+        r += \
             """USER:\n""" + \
             """<|user_prompt|>\n\n""" + \
             """ASSISTANT:\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -444,7 +478,7 @@ def __init__(self):
     def default_system_prompt(self):
         return ""
 
-    def first_prompt(self):
+    def first_prompt(self, sysprompt):
         return \
             """<bos><start_of_turn>user\n""" + \
             """<|user_prompt|><end_of_turn>\n""" + \
@@ -481,13 +515,17 @@ def __init__(self):
     def default_system_prompt(self):
         return "You are an AI coding assistant."
 
-    def first_prompt(self):
-        return \
-            """System:\n""" + \
-            """<|system_prompt|>\n\n""" + \
+    def first_prompt(self, sysprompt):
+        r = ""
+        if sysprompt:
+            r += \
+                """System:\n""" + \
+                """<|system_prompt|>\n\n"""
+        r += \
             """Question:\n""" + \
             """<|user_prompt|>\n\n""" + \
             """Answer:\n"""
+        return r
 
     def subs_prompt(self):
         return \
@@ -520,16 +558,19 @@ def default_system_prompt(self):
         return \
             f"""You are a helpful AI assistant."""
 
-    def first_prompt(self):
-        return \
-            """<BOS_TOKEN>""" + \
-            """<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>""" + \
-            """<|system_prompt|>""" + \
-            """<|END_OF_TURN_TOKEN|>""" + \
+    def first_prompt(self, sysprompt):
+        r = """<BOS_TOKEN>"""
+        if sysprompt:
+            r += \
+                """<|START_OF_TURN_TOKEN|><|SYSTEM_TOKEN|>""" + \
+                """<|system_prompt|>""" + \
+                """<|END_OF_TURN_TOKEN|>"""
+        r += \
             """<|START_OF_TURN_TOKEN|><|USER_TOKEN|>""" + \
             """<|user_prompt|>""" + \
             """<|END_OF_TURN_TOKEN|>""" + \
             """<|START_OF_TURN_TOKEN|><|CHATBOT_TOKEN|>"""
+        return r
 
     def subs_prompt(self):
         return \