mirror of
https://github.com/trustgraph-ai/trustgraph.git
synced 2026-04-28 01:46:22 +02:00
Features/vertex anthropic (#458)
* Added Anthropic support for VertexAI * Update tests to match code * Fixed private.json usage with Anthropic (I think). * Fixed test --------- Co-authored-by: Cyber MacGeddon <cybermaggedon@gmail.com>
This commit is contained in:
Jack Colquitt
GitHub
parent
e89a5b5d23
commit
244da4aec1
2 changed files with 187 additions and 81 deletions
|
|
@ -1,7 +1,7 @@
|
|||
|
||||
"""
|
||||
Simple LLM service, performs text prompt completion using VertexAI on
|
||||
Google Cloud. Input is prompt, output is response.
|
||||
Supports both Google's Gemini models and Anthropic's Claude models.
|
||||
"""
|
||||
|
||||
#
|
||||
|
|
@ -17,7 +17,7 @@ Google Cloud. Input is prompt, output is response.
|
|||
# This module's imports bring in a lot of libraries.
|
||||
|
||||
from google.oauth2 import service_account
|
||||
import google
|
||||
import google.auth
|
||||
import vertexai
|
||||
import logging
|
||||
|
||||
|
|
@ -27,6 +27,9 @@ from vertexai.generative_models import (
|
|||
HarmCategory, HarmBlockThreshold, Part, Tool, SafetySetting,
|
||||
)
|
||||
|
||||
# Added for Anthropic model support
|
||||
from anthropic import AnthropicVertex, RateLimitError
|
||||
|
||||
from .... exceptions import TooManyRequests
|
||||
from .... base import LlmService, LlmResult
|
||||
|
||||
|
|
@ -35,7 +38,7 @@ logger = logging.getLogger(__name__)
|
|||
|
||||
default_ident = "text-completion"
|
||||
|
||||
default_model = 'gemini-2.0-flash-001'
|
||||
default_model = 'gemini-1.5-flash-001'
|
||||
default_region = 'us-central1'
|
||||
default_temperature = 0.0
|
||||
default_max_output = 8192
|
||||
|
|
@ -52,111 +55,148 @@ class Processor(LlmService):
|
|||
max_output = params.get("max_output", default_max_output)
|
||||
|
||||
if private_key is None:
|
||||
raise RuntimeError("Private key file not specified")
|
||||
logger.warning("Private key file not specified, using Application Default Credentials")
|
||||
|
||||
super(Processor, self).__init__(**params)
|
||||
|
||||
self.parameters = {
|
||||
self.model = model
|
||||
self.is_anthropic = 'claude' in self.model.lower()
|
||||
|
||||
# Shared parameters for both model types
|
||||
self.api_params = {
|
||||
"temperature": temperature,
|
||||
"top_p": 1.0,
|
||||
"top_k": 32,
|
||||
"candidate_count": 1,
|
||||
"max_output_tokens": max_output,
|
||||
}
|
||||
|
||||
self.generation_config = GenerationConfig(
|
||||
temperature=temperature,
|
||||
top_p=1.0,
|
||||
top_k=10,
|
||||
candidate_count=1,
|
||||
max_output_tokens=max_output,
|
||||
)
|
||||
|
||||
# Block none doesn't seem to work
|
||||
block_level = HarmBlockThreshold.BLOCK_ONLY_HIGH
|
||||
# block_level = HarmBlockThreshold.BLOCK_NONE
|
||||
|
||||
self.safety_settings = [
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_HARASSMENT,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_HATE_SPEECH,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
|
||||
threshold = block_level,
|
||||
),
|
||||
]
|
||||
|
||||
logger.info("Initializing VertexAI...")
|
||||
|
||||
# Unified credential and project ID loading
|
||||
if private_key:
|
||||
credentials = (
|
||||
service_account.Credentials.from_service_account_file(
|
||||
private_key
|
||||
)
|
||||
)
|
||||
project_id = credentials.project_id
|
||||
else:
|
||||
credentials = None
|
||||
credentials, project_id = google.auth.default()
|
||||
|
||||
if credentials:
|
||||
vertexai.init(
|
||||
location=region,
|
||||
credentials=credentials,
|
||||
project=credentials.project_id,
|
||||
)
|
||||
else:
|
||||
vertexai.init(
|
||||
location=region
|
||||
if not project_id:
|
||||
raise RuntimeError(
|
||||
"Could not determine Google Cloud project ID. "
|
||||
"Ensure it's set in your environment or service account."
|
||||
)
|
||||
|
||||
logger.info(f"Initializing model {model}")
|
||||
self.llm = GenerativeModel(model)
|
||||
self.model = model
|
||||
# Initialize the appropriate client based on the model type
|
||||
if self.is_anthropic:
|
||||
logger.info(f"Initializing Anthropic model '{model}' via AnthropicVertex SDK")
|
||||
# Initialize AnthropicVertex with credentials if provided, otherwise use ADC
|
||||
anthropic_kwargs = {'region': region, 'project_id': project_id}
|
||||
if credentials and private_key: # Pass credentials only if from a file
|
||||
anthropic_kwargs['credentials'] = credentials
|
||||
logger.debug(f"Using service account credentials for Anthropic model")
|
||||
else:
|
||||
logger.debug(f"Using Application Default Credentials for Anthropic model")
|
||||
|
||||
self.llm = AnthropicVertex(**anthropic_kwargs)
|
||||
else:
|
||||
# For Gemini models, initialize the Vertex AI SDK
|
||||
logger.info(f"Initializing Google model '{model}' via Vertex AI SDK")
|
||||
init_kwargs = {'location': region, 'project': project_id}
|
||||
if credentials and private_key: # Pass credentials only if from a file
|
||||
init_kwargs['credentials'] = credentials
|
||||
|
||||
vertexai.init(**init_kwargs)
|
||||
|
||||
self.llm = GenerativeModel(model)
|
||||
|
||||
self.generation_config = GenerationConfig(
|
||||
temperature=temperature,
|
||||
top_p=1.0,
|
||||
top_k=10,
|
||||
candidate_count=1,
|
||||
max_output_tokens=max_output,
|
||||
)
|
||||
|
||||
# Block none doesn't seem to work
|
||||
block_level = HarmBlockThreshold.BLOCK_ONLY_HIGH
|
||||
# block_level = HarmBlockThreshold.BLOCK_NONE
|
||||
|
||||
self.safety_settings = [
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_HARASSMENT,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_HATE_SPEECH,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT,
|
||||
threshold = block_level,
|
||||
),
|
||||
SafetySetting(
|
||||
category = HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT,
|
||||
threshold = block_level,
|
||||
),
|
||||
]
|
||||
|
||||
|
||||
logger.info("VertexAI initialization complete")
|
||||
|
||||
async def generate_content(self, system, prompt):
|
||||
|
||||
try:
|
||||
if self.is_anthropic:
|
||||
# Anthropic API uses a dedicated system prompt
|
||||
logger.debug("Sending request to Anthropic model...")
|
||||
response = self.llm.messages.create(
|
||||
model=self.model,
|
||||
system=system,
|
||||
messages=[{"role": "user", "content": prompt}],
|
||||
max_tokens=self.api_params['max_output_tokens'],
|
||||
temperature=self.api_params['temperature'],
|
||||
top_p=self.api_params['top_p'],
|
||||
top_k=self.api_params['top_k'],
|
||||
)
|
||||
|
||||
prompt = system + "\n\n" + prompt
|
||||
resp = LlmResult(
|
||||
text=response.content[0].text,
|
||||
in_token=response.usage.input_tokens,
|
||||
out_token=response.usage.output_tokens,
|
||||
model=self.model
|
||||
)
|
||||
else:
|
||||
# Gemini API combines system and user prompts
|
||||
logger.debug("Sending request to Gemini model...")
|
||||
full_prompt = system + "\n\n" + prompt
|
||||
|
||||
response = self.llm.generate_content(
|
||||
prompt, generation_config = self.generation_config,
|
||||
safety_settings = self.safety_settings,
|
||||
)
|
||||
response = self.llm.generate_content(
|
||||
full_prompt, generation_config = self.generation_config,
|
||||
safety_settings = self.safety_settings,
|
||||
)
|
||||
|
||||
resp = LlmResult(
|
||||
text = response.text,
|
||||
in_token = response.usage_metadata.prompt_token_count,
|
||||
out_token = response.usage_metadata.candidates_token_count,
|
||||
model = self.model
|
||||
)
|
||||
resp = LlmResult(
|
||||
text = response.text,
|
||||
in_token = response.usage_metadata.prompt_token_count,
|
||||
out_token = response.usage_metadata.candidates_token_count,
|
||||
model = self.model
|
||||
)
|
||||
|
||||
logger.info(f"Input Tokens: {resp.in_token}")
|
||||
logger.info(f"Output Tokens: {resp.out_token}")
|
||||
|
||||
logger.debug("Send response...")
|
||||
|
||||
return resp
|
||||
|
||||
except google.api_core.exceptions.ResourceExhausted as e:
|
||||
|
||||
except (google.api_core.exceptions.ResourceExhausted, RateLimitError) as e:
|
||||
logger.warning(f"Hit rate limit: {e}")
|
||||
|
||||
# Leave rate limit retries to the base handler
|
||||
raise TooManyRequests()
|
||||
|
||||
except Exception as e:
|
||||
|
||||
# Apart from rate limits, treat all exceptions as unrecoverable
|
||||
logger.error(f"VertexAI LLM exception: {e}", exc_info=True)
|
||||
raise e
|
||||
|
|
@ -169,12 +209,12 @@ class Processor(LlmService):
|
|||
parser.add_argument(
|
||||
'-m', '--model',
|
||||
default=default_model,
|
||||
help=f'LLM model (default: {default_model})'
|
||||
help=f'LLM model (e.g., gemini-1.5-flash-001, claude-3-sonnet@20240229) (default: {default_model})'
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
'-k', '--private-key',
|
||||
help=f'Google Cloud private JSON file'
|
||||
help=f'Google Cloud private JSON file (optional, uses ADC if not provided)'
|
||||
)
|
||||
|
||||
parser.add_argument(
|
||||
|
|
@ -198,5 +238,4 @@ class Processor(LlmService):
|
|||
)
|
||||
|
||||
def run():
|
||||
Processor.launch(default_ident, __doc__)
|
||||
|
||||
Processor.launch(default_ident, __doc__)
|
||||
Loading…
Add table
Add a link
Reference in a new issue