ce9ef3ca79
* Add support for non-ASCII characters in LLM prompts - Add ensure_ascii parameter to Graphiti class (default: True) - Create to_prompt_json helper function for consistent JSON serialization - Update all prompt files to use new helper function - Preserve Korean/Japanese/Chinese characters when ensure_ascii=False - Maintain backward compatibility with existing behavior Fixes issue where non-ASCII characters were escaped as unicode sequences in prompts, making them unreadable in LLM logs and potentially affecting model understanding. * Remove unused json imports after replacing with to_prompt_json helper - Fix ruff lint errors (F401) for unused json imports - All prompt files now use to_prompt_json helper instead of json.dumps - Maintains clean code style and passes lint checks * Fix ensure_ascii propagation to all LLM calls - Add ensure_ascii parameter to maintenance operation functions that were missing it - Update function signatures in node_operations, community_operations, temporal_operations, and edge_operations - Ensure all llm_client.generate_response calls receive proper ensure_ascii context - Fix hardcoded ensure_ascii: True values that prevented non-ASCII character preservation - Maintain backward compatibility with default ensure_ascii=True - Complete the fix for issue #804 ensuring Korean/Japanese/Chinese characters are properly handled in LLM prompts
208 lines
8 KiB
Python
208 lines
8 KiB
Python
"""
|
||
Copyright 2024, Zep Software, Inc.
|
||
|
||
Licensed under the Apache License, Version 2.0 (the "License");
|
||
you may not use this file except in compliance with the License.
|
||
You may obtain a copy of the License at
|
||
|
||
http://www.apache.org/licenses/LICENSE-2.0
|
||
|
||
Unless required by applicable law or agreed to in writing, software
|
||
distributed under the License is distributed on an "AS IS" BASIS,
|
||
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||
See the License for the specific language governing permissions and
|
||
limitations under the License.
|
||
"""
|
||
|
||
from typing import Any, Protocol, TypedDict
|
||
|
||
from pydantic import BaseModel, Field
|
||
|
||
from .models import Message, PromptFunction, PromptVersion
|
||
from .prompt_helpers import to_prompt_json
|
||
|
||
|
||
class NodeDuplicate(BaseModel):
|
||
id: int = Field(..., description='integer id of the entity')
|
||
duplicate_idx: int = Field(
|
||
...,
|
||
description='idx of the duplicate entity. If no duplicate entities are found, default to -1.',
|
||
)
|
||
name: str = Field(
|
||
...,
|
||
description='Name of the entity. Should be the most complete and descriptive name of the entity. Do not include any JSON formatting in the Entity name such as {}.',
|
||
)
|
||
duplicates: list[int] = Field(
|
||
...,
|
||
description='idx of all entities that are a duplicate of the entity with the above id.',
|
||
)
|
||
|
||
|
||
class NodeResolutions(BaseModel):
|
||
entity_resolutions: list[NodeDuplicate] = Field(..., description='List of resolved nodes')
|
||
|
||
|
||
class Prompt(Protocol):
|
||
node: PromptVersion
|
||
node_list: PromptVersion
|
||
nodes: PromptVersion
|
||
|
||
|
||
class Versions(TypedDict):
|
||
node: PromptFunction
|
||
node_list: PromptFunction
|
||
nodes: PromptFunction
|
||
|
||
|
||
def node(context: dict[str, Any]) -> list[Message]:
|
||
return [
|
||
Message(
|
||
role='system',
|
||
content='You are a helpful assistant that determines whether or not a NEW ENTITY is a duplicate of any EXISTING ENTITIES.',
|
||
),
|
||
Message(
|
||
role='user',
|
||
content=f"""
|
||
<PREVIOUS MESSAGES>
|
||
{to_prompt_json([ep for ep in context['previous_episodes']], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</PREVIOUS MESSAGES>
|
||
<CURRENT MESSAGE>
|
||
{context['episode_content']}
|
||
</CURRENT MESSAGE>
|
||
<NEW ENTITY>
|
||
{to_prompt_json(context['extracted_node'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</NEW ENTITY>
|
||
<ENTITY TYPE DESCRIPTION>
|
||
{to_prompt_json(context['entity_type_description'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</ENTITY TYPE DESCRIPTION>
|
||
|
||
<EXISTING ENTITIES>
|
||
{to_prompt_json(context['existing_nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</EXISTING ENTITIES>
|
||
|
||
Given the above EXISTING ENTITIES and their attributes, MESSAGE, and PREVIOUS MESSAGES; Determine if the NEW ENTITY extracted from the conversation
|
||
is a duplicate entity of one of the EXISTING ENTITIES.
|
||
|
||
Entities should only be considered duplicates if they refer to the *same real-world object or concept*.
|
||
Semantic Equivalence: if a descriptive label in existing_entities clearly refers to a named entity in context, treat them as duplicates.
|
||
|
||
Do NOT mark entities as duplicates if:
|
||
- They are related but distinct.
|
||
- They have similar names or purposes but refer to separate instances or concepts.
|
||
|
||
TASK:
|
||
1. Compare `new_entity` against each item in `existing_entities`.
|
||
2. If it refers to the same real‐world object or concept, collect its index.
|
||
3. Let `duplicate_idx` = the *first* collected index, or –1 if none.
|
||
4. Let `duplicates` = the list of *all* collected indices (empty list if none).
|
||
|
||
Also return the full name of the NEW ENTITY (whether it is the name of the NEW ENTITY, a node it
|
||
is a duplicate of, or a combination of the two).
|
||
""",
|
||
),
|
||
]
|
||
|
||
|
||
def nodes(context: dict[str, Any]) -> list[Message]:
|
||
return [
|
||
Message(
|
||
role='system',
|
||
content='You are a helpful assistant that determines whether or not ENTITIES extracted from a conversation are duplicates'
|
||
'of existing entities.',
|
||
),
|
||
Message(
|
||
role='user',
|
||
content=f"""
|
||
<PREVIOUS MESSAGES>
|
||
{to_prompt_json([ep for ep in context['previous_episodes']], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</PREVIOUS MESSAGES>
|
||
<CURRENT MESSAGE>
|
||
{context['episode_content']}
|
||
</CURRENT MESSAGE>
|
||
|
||
|
||
Each of the following ENTITIES were extracted from the CURRENT MESSAGE.
|
||
Each entity in ENTITIES is represented as a JSON object with the following structure:
|
||
{{
|
||
id: integer id of the entity,
|
||
name: "name of the entity",
|
||
entity_type: "ontological classification of the entity",
|
||
entity_type_description: "Description of what the entity type represents",
|
||
duplication_candidates: [
|
||
{{
|
||
idx: integer index of the candidate entity,
|
||
name: "name of the candidate entity",
|
||
entity_type: "ontological classification of the candidate entity",
|
||
...<additional attributes>
|
||
}}
|
||
]
|
||
}}
|
||
|
||
<ENTITIES>
|
||
{to_prompt_json(context['extracted_nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</ENTITIES>
|
||
|
||
<EXISTING ENTITIES>
|
||
{to_prompt_json(context['existing_nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
</EXISTING ENTITIES>
|
||
|
||
For each of the above ENTITIES, determine if the entity is a duplicate of any of the EXISTING ENTITIES.
|
||
|
||
Entities should only be considered duplicates if they refer to the *same real-world object or concept*.
|
||
|
||
Do NOT mark entities as duplicates if:
|
||
- They are related but distinct.
|
||
- They have similar names or purposes but refer to separate instances or concepts.
|
||
|
||
Task:
|
||
Your response will be a list called entity_resolutions which contains one entry for each entity.
|
||
|
||
For each entity, return the id of the entity as id, the name of the entity as name, and the duplicate_idx
|
||
as an integer.
|
||
|
||
- If an entity is a duplicate of one of the EXISTING ENTITIES, return the idx of the candidate it is a
|
||
duplicate of.
|
||
- If an entity is not a duplicate of one of the EXISTING ENTITIES, return the -1 as the duplication_idx
|
||
""",
|
||
),
|
||
]
|
||
|
||
|
||
def node_list(context: dict[str, Any]) -> list[Message]:
|
||
return [
|
||
Message(
|
||
role='system',
|
||
content='You are a helpful assistant that de-duplicates nodes from node lists.',
|
||
),
|
||
Message(
|
||
role='user',
|
||
content=f"""
|
||
Given the following context, deduplicate a list of nodes:
|
||
|
||
Nodes:
|
||
{to_prompt_json(context['nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
|
||
|
||
Task:
|
||
1. Group nodes together such that all duplicate nodes are in the same list of uuids
|
||
2. All duplicate uuids should be grouped together in the same list
|
||
3. Also return a new summary that synthesizes the summary into a new short summary
|
||
|
||
Guidelines:
|
||
1. Each uuid from the list of nodes should appear EXACTLY once in your response
|
||
2. If a node has no duplicates, it should appear in the response in a list of only one uuid
|
||
|
||
Respond with a JSON object in the following format:
|
||
{{
|
||
"nodes": [
|
||
{{
|
||
"uuids": ["5d643020624c42fa9de13f97b1b3fa39", "node that is a duplicate of 5d643020624c42fa9de13f97b1b3fa39"],
|
||
"summary": "Brief summary of the node summaries that appear in the list of names."
|
||
}}
|
||
]
|
||
}}
|
||
""",
|
||
),
|
||
]
|
||
|
||
|
||
versions: Versions = {'node': node, 'node_list': node_list, 'nodes': nodes}
|