Update poetry lock, and fix run for template prompt format

private_gpt/components/llm/prompt/prompt_helper.py

@@ -178,8 +178,10 @@ class AbstractPromptStyle(abc.ABC):
 class AbstractPromptStyleWithSystemPrompt(AbstractPromptStyle, abc.ABC):
     _DEFAULT_SYSTEM_PROMPT = DEFAULT_SYSTEM_PROMPT
 
-    def __init__(self, default_system_prompt: str | None) -> None:
+    def __init__(
-        super().__init__()
+        self, default_system_prompt: str | None, *args: Any, **kwargs: Any
+    ) -> None:
+        super().__init__(*args, **kwargs)
         logger.debug("Got default_system_prompt='%s'", default_system_prompt)
         self.default_system_prompt = default_system_prompt
 
@@ -235,9 +237,13 @@ class LlamaIndexPromptStyle(AbstractPromptStyleWithSystemPrompt):
     ```
     """
 
-    def __init__(self, default_system_prompt: str | None = None) -> None:
+    def __init__(
+        self, default_system_prompt: str | None = None, *args: Any, **kwargs: Any
+    ) -> None:
         # If no system prompt is given, the default one of the implementation is used.
-        super().__init__(default_system_prompt=default_system_prompt)
+        # default_system_prompt can be None here
+        kwargs["default_system_prompt"] = default_system_prompt
+        super().__init__(*args, **kwargs)
 
     def _messages_to_prompt(self, messages: Sequence[ChatMessage]) -> str:
         return messages_to_prompt(messages, self.default_system_prompt)
@@ -264,12 +270,14 @@ class VigognePromptStyle(AbstractPromptStyleWithSystemPrompt):
         self,
         default_system_prompt: str | None = None,
         add_generation_prompt: bool = True,
+        *args: Any,
+        **kwargs: Any,
     ) -> None:
         # We have to define a default system prompt here as the LLM will not
         # use the default llama_utils functions.
         default_system_prompt = default_system_prompt or self._DEFAULT_SYSTEM_PROMPT
-        super().__init__(default_system_prompt)
+        kwargs["default_system_prompt"] = default_system_prompt
-        self.system_prompt: str = default_system_prompt
+        super().__init__(*args, **kwargs)
         self.add_generation_prompt = add_generation_prompt
 
     def _messages_to_prompt(self, messages: Sequence[ChatMessage]) -> str:
@@ -300,7 +308,11 @@ class VigognePromptStyle(AbstractPromptStyleWithSystemPrompt):
 
 class LlamaCppPromptStyle(AbstractPromptStyleWithSystemPrompt):
     def __init__(
-        self, prompt_style: str, default_system_prompt: str | None = None
+        self,
+        prompt_style: str,
+        default_system_prompt: str | None = None,
+        *args: Any,
+        **kwargs: Any,
     ) -> None:
         """Wrapper for llama_cpp_python defined prompt format.
 
@@ -309,7 +321,8 @@ class LlamaCppPromptStyle(AbstractPromptStyleWithSystemPrompt):
         """
         assert prompt_style.startswith("llama_cpp.")
         default_system_prompt = default_system_prompt or self._DEFAULT_SYSTEM_PROMPT
-        super().__init__(default_system_prompt)
+        kwargs["default_system_prompt"] = default_system_prompt
+        super().__init__(*args, **kwargs)
 
         self.prompt_style = prompt_style[len("llama_cpp.") :]
         if self.prompt_style is None:
@@ -339,6 +352,8 @@ class TemplatePromptStyle(AbstractPromptStyleWithSystemPrompt):
         template_dir: str | None = None,
         add_generation_prompt: bool = True,
         default_system_prompt: str | None = None,
+        *args: Any,
+        **kwargs: Any,
     ) -> None:
         """Prompt format using a Jinja template.
 
@@ -350,7 +365,8 @@ class TemplatePromptStyle(AbstractPromptStyleWithSystemPrompt):
             given in the messages.
         """
         default_system_prompt = default_system_prompt or DEFAULT_SYSTEM_PROMPT
-        super().__init__(default_system_prompt)
+        kwargs["default_system_prompt"] = default_system_prompt
+        super().__init__(*args, **kwargs)
 
         self._add_generation_prompt = add_generation_prompt
 

private_gpt/components/llm/prompt/template/vigogne-2-7b-chat.jinja

@@ -1,2 +1,2 @@
 {# This template is coming from: https://huggingface.co/bofenghuang/vigogne-2-7b-chat/blob/main/tokenizer_config.json  #}
-{{ bos_token }}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% elif true == true %}{% set loop_messages = messages %}{% set system_message = 'Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.' %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% if system_message != false %}{{ '<|system|>: ' + system_message + '\\n' }}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '<|user|>: ' + message['content'].strip() + '\\n' }}{% elif message['role'] == 'assistant' %}{{ '<|assistant|>: ' + message['content'].strip() + eos_token + '\\n' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>:' }}{% endif %}
+{{ bos_token }}{% if messages[0]['role'] == 'system' %}{% set loop_messages = messages[1:] %}{% set system_message = messages[0]['content'] %}{% elif true == true %}{% set loop_messages = messages %}{% set system_message = 'Vous êtes Vigogne, un assistant IA créé par Zaion Lab. Vous suivez extrêmement bien les instructions. Aidez autant que vous le pouvez.' %}{% else %}{% set loop_messages = messages %}{% set system_message = false %}{% endif %}{% if system_message != false %}{{ '<|system|>: ' + system_message + '\n' }}{% endif %}{% for message in loop_messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if message['role'] == 'user' %}{{ '<|user|>: ' + message['content'].strip() + '\n' }}{% elif message['role'] == 'assistant' %}{{ '<|assistant|>: ' + message['content'].strip() + eos_token + '\n' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|assistant|>:' }}{% endif %}

pyproject.toml

@@ -36,12 +36,13 @@ gradio = "^4.7.1"
 [tool.poetry.group.local]
 optional = true
 [tool.poetry.group.local.dependencies]
-llama-cpp-python = "^0.2.11"
+llama-cpp-python = "^0.2.20"
-numpy = "1.26.0"
+jinja2 = "^3.1.2"
+# numpy = "1.26.0"
 sentence-transformers = "^2.2.2"
 # https://stackoverflow.com/questions/76327419/valueerror-libcublas-so-0-9-not-found-in-the-system-path
 torch = ">=2.0.0, !=2.0.1, !=2.1.0"
-transformers = "^4.34.0"
+transformers = "^4.35.2"
 
 [tool.poetry.extras]
 chroma = ["chromadb"]