Refactor names (#4)

- Downsize embeddings model to mini-lm in docker-compose files
- Rename for structure
- Default queues defined in schema file
- Standardize naming: graph embeddings, chunk embeddings, triples

trustgraph/model/text_completion/azure/__init__.py

@@ -0,0 +1,3 @@
+
+from . llm import *
+

trustgraph/model/text_completion/azure/__main__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+
+from . llm import run
+
+if __name__ == '__main__':
+    run()
+

trustgraph/model/text_completion/azure/llm.py

@@ -0,0 +1,128 @@
+
+"""
+Simple LLM service, performs text prompt completion using the Azure
+serverless endpoint service.  Input is prompt, output is response.
+"""
+
+import requests
+import json
+
+from .... schema import TextCompletionRequest, TextCompletionResponse
+from .... schema import text_completion_request_queue
+from .... schema import text_completion_response_queue
+from .... log_level import LogLevel
+from .... base import ConsumerProducer
+
+module = ".".join(__name__.split(".")[1:-1])
+
+default_input_queue = text_completion_request_queue
+default_output_queue = text_completion_response_queue
+default_subscriber = module
+
+class Processor(ConsumerProducer):
+
+    def __init__(self, **params):
+
+        input_queue = params.get("input_queue", default_input_queue)
+        output_queue = params.get("output_queue", default_output_queue)
+        subscriber = params.get("subscriber", default_subscriber)
+        endpoint = params.get("endpoint")
+        token = params.get("token")
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "input_schema": TextCompletionRequest,
+                "output_schema": TextCompletionResponse,
+            }
+        )
+
+        self.endpoint = endpoint
+        self.token = token
+
+    def build_prompt(self, system, content):
+
+        data =  {
+            "messages": [
+                {
+                    "role": "system", "content": system
+                },
+                {
+                    "role": "user", "content": content
+                }
+            ],
+            "max_tokens": 4192,
+            "temperature": 0.2,
+            "top_p": 1
+        }
+
+        body = json.dumps(data)
+
+        return body
+
+    def call_llm(self, body):
+
+        url = self.endpoint
+
+        # Replace this with the primary/secondary key, AMLToken, or
+        # Microsoft Entra ID token for the endpoint
+        api_key = self.token
+
+        headers = {
+            'Content-Type': 'application/json',
+            'Authorization': f'Bearer {api_key}'
+        }
+
+        resp = requests.post(url, data=body, headers=headers)
+        result = resp.json()
+
+        message_content = result['choices'][0]['message']['content']
+
+        return message_content
+
+    def handle(self, msg):
+
+        v = msg.value()
+
+        # Sender-produced ID
+
+        id = msg.properties()["id"]
+
+        print(f"Handling prompt {id}...", flush=True)
+
+        prompt = self.build_prompt(
+            "You are a helpful chatbot",
+            v.prompt
+        )
+
+        response = self.call_llm(prompt)
+
+        print("Send response...", flush=True)
+        r = TextCompletionResponse(response=response)
+        self.producer.send(r, properties={"id": id})
+
+        print("Done.", flush=True)
+
+    @staticmethod
+    def add_args(parser):
+
+        ConsumerProducer.add_args(
+            parser, default_input_queue, default_subscriber,
+            default_output_queue,
+        )
+
+        parser.add_argument(
+            '-e', '--endpoint',
+            help=f'LLM model endpoint'
+        )
+
+        parser.add_argument(
+            '-k', '--token',
+            help=f'LLM model token'
+        )
+
+def run():
+    
+    Processor.start(module, __doc__)

trustgraph/model/text_completion/claude/__init__.py

@@ -0,0 +1,3 @@
+
+from . llm import *
+

trustgraph/model/text_completion/claude/__main__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+
+from . llm import run
+
+if __name__ == '__main__':
+    run()
+

trustgraph/model/text_completion/claude/llm.py

@@ -0,0 +1,110 @@
+
+"""
+Simple LLM service, performs text prompt completion using Claude.
+Input is prompt, output is response.
+"""
+
+import anthropic
+
+from .... schema import TextCompletionRequest, TextCompletionResponse
+from .... schema import text_completion_request_queue
+from .... schema import text_completion_response_queue
+from .... log_level import LogLevel
+from .... base import ConsumerProducer
+
+module = ".".join(__name__.split(".")[1:-1])
+
+default_input_queue = text_completion_request_queue
+default_output_queue = text_completion_response_queue
+default_subscriber = module
+default_model = 'claude-3-5-sonnet-20240620'
+
+class Processor(ConsumerProducer):
+
+    def __init__(self, **params):
+    
+        input_queue = params.get("input_queue", default_input_queue)
+        output_queue = params.get("output_queue", default_output_queue)
+        subscriber = params.get("subscriber", default_subscriber)
+        model = params.get("model", default_model)
+        api_key = params.get("api_key")
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "input_schema": TextCompletionRequest,
+                "output_schema": TextCompletionResponse,
+                "model": model,
+            }
+        )
+
+        self.model = model
+
+        self.claude = anthropic.Anthropic(api_key=api_key)
+
+        print("Initialised", flush=True)
+
+    def handle(self, msg):
+
+        v = msg.value()
+
+        # Sender-produced ID
+
+        id = msg.properties()["id"]
+
+        print(f"Handling prompt {id}...", flush=True)
+
+        prompt = v.prompt
+        response = message = self.claude.messages.create(
+            model=self.model,
+            max_tokens=1000,
+            temperature=0.1,
+            system = "You are a helpful chatbot.",
+            messages=[
+                {
+                    "role": "user",
+                    "content": [
+                        {
+                            "type": "text",
+                            "text": prompt
+                        }
+                    ]
+                }
+            ]
+        )
+
+        resp = response.content[0].text
+        print(resp, flush=True)
+
+        print("Send response...", flush=True)
+        r = TextCompletionResponse(response=resp)
+        self.send(r, properties={"id": id})
+
+        print("Done.", flush=True)
+
+    @staticmethod
+    def add_args(parser):
+
+        ConsumerProducer.add_args(
+            parser, default_input_queue, default_subscriber,
+            default_output_queue,
+        )
+
+        parser.add_argument(
+            '-m', '--model',
+            default="claude-3-5-sonnet-20240620",
+            help=f'LLM model (default: claude-3-5-sonnet-20240620)'
+        )
+
+        parser.add_argument(
+            '-k', '--api-key',
+            help=f'Claude API key'
+        )
+
+def run():
+
+    Processor.start(module, __doc__)
+
+    

trustgraph/model/text_completion/ollama/__init__.py

@@ -0,0 +1,3 @@
+
+from . llm import *
+

trustgraph/model/text_completion/ollama/__main__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+
+from . llm import run
+
+if __name__ == '__main__':
+    run()
+

trustgraph/model/text_completion/ollama/llm.py

@@ -0,0 +1,102 @@
+
+"""
+Simple LLM service, performs text prompt completion using an Ollama service.
+Input is prompt, output is response.
+"""
+
+from langchain_community.llms import Ollama
+from prometheus_client import Histogram, Info, Counter
+
+from .... schema import TextCompletionRequest, TextCompletionResponse
+from .... schema import text_completion_request_queue
+from .... schema import text_completion_response_queue
+from .... log_level import LogLevel
+from .... base import ConsumerProducer
+
+module = ".".join(__name__.split(".")[1:-1])
+
+default_input_queue = text_completion_request_queue
+default_output_queue = text_completion_response_queue
+default_subscriber = module
+default_model = 'gemma2'
+default_ollama = 'http://localhost:11434'
+
+class Processor(ConsumerProducer):
+
+    def __init__(self, **params):
+
+        input_queue = params.get("input_queue", default_input_queue)
+        output_queue = params.get("output_queue", default_output_queue)
+        subscriber = params.get("subscriber", default_subscriber)
+        model = params.get("model", default_model)
+        ollama = params.get("ollama", default_ollama)
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "model": model,
+                "ollama": ollama,
+                "input_schema": TextCompletionRequest,
+                "output_schema": TextCompletionResponse,
+            }
+        )
+
+        if not hasattr(__class__, "model_metric"):
+            __class__.model_metric = Info(
+                'model', 'Model information'
+            )
+
+        __class__.model_metric.info({
+            "model": model,
+            "ollama": ollama,
+        })
+
+        self.llm = Ollama(base_url=ollama, model=model)
+
+    def handle(self, msg):
+
+        v = msg.value()
+
+        # Sender-produced ID
+        id = msg.properties()["id"]
+
+        print(f"Handling prompt {id}...", flush=True)
+
+        prompt = v.prompt
+        response = self.llm.invoke(prompt)
+
+        print("Send response...", flush=True)
+
+        r = TextCompletionResponse(response=response)
+
+        self.send(r, properties={"id": id})
+
+        print("Done.", flush=True)
+
+    @staticmethod
+    def add_args(parser):
+
+        ConsumerProducer.add_args(
+            parser, default_input_queue, default_subscriber,
+            default_output_queue,
+        )
+
+        parser.add_argument(
+            '-m', '--model',
+            default="gemma2",
+            help=f'LLM model (default: gemma2)'
+        )
+
+        parser.add_argument(
+            '-r', '--ollama',
+            default=default_ollama,
+            help=f'ollama (default: {default_ollama})'
+        )
+
+def run():
+
+    Processor.start(module, __doc__)
+
+    

trustgraph/model/text_completion/vertexai/__init__.py

@@ -0,0 +1,3 @@
+
+from . llm import *
+

trustgraph/model/text_completion/vertexai/__main__.py

@@ -0,0 +1,7 @@
+#!/usr/bin/env python3
+
+from . llm import run
+
+if __name__ == '__main__':
+    run()
+

trustgraph/model/text_completion/vertexai/llm.py

@@ -0,0 +1,177 @@
+
+"""
+Simple LLM service, performs text prompt completion using VertexAI on
+Google Cloud.   Input is prompt, output is response.
+"""
+
+import vertexai
+import time
+
+from google.oauth2 import service_account
+import google
+
+from vertexai.preview.generative_models import (
+    Content,
+    FunctionDeclaration,
+    GenerativeModel,
+    GenerationConfig,
+    HarmCategory,
+    HarmBlockThreshold,
+    Part,
+    Tool,
+)
+
+from .... schema import TextCompletionRequest, TextCompletionResponse
+from .... schema import text_completion_request_queue
+from .... schema import text_completion_response_queue
+from .... log_level import LogLevel
+from .... base import ConsumerProducer
+
+module = ".".join(__name__.split(".")[1:-1])
+
+default_input_queue = text_completion_request_queue
+default_output_queue = text_completion_response_queue
+default_subscriber = module
+
+class Processor(ConsumerProducer):
+
+    def __init__(self, **params):
+
+        input_queue = params.get("input_queue", default_input_queue)
+        output_queue = params.get("output_queue", default_output_queue)
+        subscriber = params.get("subscriber", default_subscriber)
+        region = params.get("region", "us-west1")
+        model = params.get("model", "gemini-1.0-pro-001")
+        private_key = params.get("private_key")
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "input_schema": TextCompletionRequest,
+                "output_schema": TextCompletionResponse,
+            }
+        )
+
+        self.parameters = {
+            "temperature": 0.2,
+            "top_p": 1.0,
+            "top_k": 32,
+            "candidate_count": 1,
+            "max_output_tokens": 8192,
+        }
+
+        self.generation_config = GenerationConfig(
+            temperature=0.2,
+            top_p=1.0,
+            top_k=10,
+            candidate_count=1,
+            max_output_tokens=8191,
+        )
+
+        # Block none doesn't seem to work
+        block_level = HarmBlockThreshold.BLOCK_ONLY_HIGH
+        #     block_level = HarmBlockThreshold.BLOCK_NONE
+
+        self.safety_settings = {
+            HarmCategory.HARM_CATEGORY_HARASSMENT: block_level,
+            HarmCategory.HARM_CATEGORY_HATE_SPEECH: block_level,
+            HarmCategory.HARM_CATEGORY_SEXUALLY_EXPLICIT: block_level,
+            HarmCategory.HARM_CATEGORY_DANGEROUS_CONTENT: block_level,
+        }
+
+        print("Initialise VertexAI...", flush=True)
+
+        if private_key:
+            credentials = service_account.Credentials.from_service_account_file(private_key)
+        else:
+            credentials = None
+
+        if credentials:
+            vertexai.init(
+                location=region,
+                credentials=credentials,
+                project=credentials.project_id,
+            )
+        else:
+            vertexai.init(
+                location=region
+            )
+
+        print(f"Initialise model {model}", flush=True)
+        self.llm = GenerativeModel(model)
+
+        print("Initialisation complete", flush=True)
+
+    def handle(self, msg):
+
+        try:
+
+            v = msg.value()
+
+            # Sender-produced ID
+
+            id = msg.properties()["id"]
+
+            print(f"Handling prompt {id}...", flush=True)
+
+            prompt = v.prompt
+
+            resp = self.llm.generate_content(
+                prompt, generation_config=self.generation_config,
+                safety_settings=self.safety_settings
+            )
+
+            resp = resp.text
+
+            resp = resp.replace("```json", "")
+            resp = resp.replace("```", "")
+
+            print("Send response...", flush=True)
+            r = TextCompletionResponse(response=resp)
+            self.producer.send(r, properties={"id": id})
+
+            print("Done.", flush=True)
+
+            # Acknowledge successful processing of the message
+            self.consumer.acknowledge(msg)
+
+        except google.api_core.exceptions.ResourceExhausted:
+
+            print("429, resource busy, sleeping", flush=True)
+            time.sleep(15)
+            self.consumer.negative_acknowledge(msg)
+
+        # Let other exceptions fall through
+
+    @staticmethod
+    def add_args(parser):
+
+        ConsumerProducer.add_args(
+            parser, default_input_queue, default_subscriber,
+            default_output_queue,
+        )
+
+        parser.add_argument(
+            '-m', '--model',
+            default="gemini-1.0-pro-001",
+            help=f'LLM model (default: gemini-1.0-pro-001)'
+        )
+        # Also: text-bison-32k
+
+        parser.add_argument(
+            '-k', '--private-key',
+            help=f'Google Cloud private JSON file'
+        )
+
+        parser.add_argument(
+            '-r', '--region',
+            default='us-west1',
+            help=f'Google Cloud region (default: us-west1)',
+        )
+
+def run():
+
+    Processor.start(module, __doc__)
+