Spaces:
Paused
Paused
File size: 4,101 Bytes
f647629 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
import os
from typing import Any, Dict, List, Optional, Tuple
import anthropic
import weave
# Load .env file before anything else that might need environment variables
from dotenv import load_dotenv
from pydantic import BaseModel, Field
from wandb_mcp_server.utils import get_rich_logger
load_dotenv()
logger = get_rich_logger(__name__)
ANTHROPIC_API_KEY = os.getenv("ANTHROPIC_API_KEY")
client = anthropic.Anthropic(api_key=ANTHROPIC_API_KEY) if ANTHROPIC_API_KEY else None
# -----------------------------------------------------------------------------
# Test tool response correctness
# -----------------------------------------------------------------------------
class CheckCorrectness(BaseModel):
"""Check if the tool response is correct."""
reasoning: str = Field(
...,
description="Reasoning about the correctness of the tool response given \
the user query, the expected value of the output and the response data from the W&B Api. The \
expected outout should be clear to see in the response data.",
)
is_correct: bool = Field(
...,
description="Whether the tool response is correct given the \
expected output.",
)
check_correctness_schema = CheckCorrectness.model_json_schema()
check_correctness_tool = {
"name": "check_correctness_tool",
"description": "Check if the assistant's response is correct given an expected output.",
"input_schema": check_correctness_schema,
}
# -----------------------------------------------------------------------------
# Call Anthropic
# -----------------------------------------------------------------------------
@weave.op
def call_anthropic(
model_name: str,
messages: List[Dict[str, Any]],
tools: Optional[List[Dict[str, Any]]] = None,
check_correctness_tool: Optional[Dict[str, Any]] = None,
):
"""Send a chat completion request to the Anthropic client with the supplied tools."""
if client is None:
raise EnvironmentError(
"ANTHROPIC_API_KEY environment variable must be set for live Anthropic calls."
)
if tools:
return client.messages.create(
model=model_name, max_tokens=4000, tools=tools, messages=messages
)
elif check_correctness_tool:
return client.messages.create(
model=model_name,
max_tokens=4000,
tools=[check_correctness_tool],
messages=messages,
tool_choice={"type": "tool", "name": "check_correctness_tool"},
)
else:
return client.messages.create(
model=model_name, max_tokens=4000, messages=messages
)
@weave.op
def extract_anthropic_tool_use(
response,
) -> Tuple[Any, str | None, Dict[str, Any] | None, str | None]:
"""Grab the first tool_use block from an Anthropic response and return (tool_use, name, input, id)."""
for idx, content in enumerate(response.content):
logger.debug(f"LLM response content {idx}: {content}")
if content.type == "tool_use":
return content, content.name, content.input, content.id
return None, None, None, None
@weave.op
def extract_anthropic_text(
response,
) -> Tuple[Any, str | None, Dict[str, Any] | None, str | None]:
"""Grab the first text block from an Anthropic response and return (text, id)."""
for idx, content in enumerate(response.content):
logger.debug(f"LLM response content {idx}: {content}")
if content.type == "text":
return content.text
return None, None
@weave.op
def get_anthropic_tool_result_message(tool_result: Any, tool_id: str) -> Dict[str, Any]:
"""Helper for feeding a tool result back to Anthropic in the required format."""
return {
"role": "user",
"content": [
{
"type": "tool_result",
"tool_use_id": tool_id,
"content": str(tool_result),
}
],
}
# Export symbols for ease of import
__all__ = [
"call_anthropic",
"extract_anthropic_tool_use",
"get_anthropic_tool_result_message",
]
|