trustgraph-flow 0.21.9__tar.gz → 0.21.10__tar.gz

{trustgraph-flow-0.21.9 → trustgraph-flow-0.21.10}/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: trustgraph-flow
-Version: 0.21.9
+Version: 0.21.10
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Home-page: https://github.com/trustgraph-ai/trustgraph
-Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.21.9.tar.gz
+Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.21.10.tar.gz
 Author: trustgraph.ai
 Author-email: security@trustgraph.ai
 Classifier: Programming Language :: Python :: 3

trustgraph-flow-0.21.10/scripts/text-completion-lmstudio

@@ -0,0 +1,6 @@
+#!/usr/bin/env python3
+
+from trustgraph.model.text_completion.lmstudio import run
+
+run()
+

{trustgraph-flow-0.21.9 → trustgraph-flow-0.21.10}/setup.py

@@ -108,6 +108,7 @@ setuptools.setup(
         "scripts/text-completion-cohere",
         "scripts/text-completion-googleaistudio",
         "scripts/text-completion-llamafile",
+        "scripts/text-completion-lmstudio",
         "scripts/text-completion-mistral",
         "scripts/text-completion-ollama",
         "scripts/text-completion-openai",

trustgraph-flow-0.21.10/trustgraph/flow_version.py

@@ -0,0 +1 @@
+__version__ = "0.21.10"

trustgraph-flow-0.21.10/trustgraph/model/text_completion/lmstudio/llm.py

@@ -0,0 +1,193 @@
+
+"""
+Simple LLM service, performs text prompt completion using OpenAI.
+Input is prompt, output is response.
+"""
+
+from openai import OpenAI
+from prometheus_client import Histogram
+import os
+
+from .... schema import TextCompletionRequest, TextCompletionResponse, Error
+from .... schema import text_completion_request_queue
+from .... schema import text_completion_response_queue
+from .... log_level import LogLevel
+from .... base import ConsumerProducer
+from .... exceptions import TooManyRequests
+
+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 = 'gemma3:9b'
+default_url = os.getenv("LMSTUDIO_URL", "http://localhost:1234/")
+default_temperature = 0.0
+default_max_output = 4096
+
+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)
+        url = params.get("url", default_url)
+        temperature = params.get("temperature", default_temperature)
+        max_output = params.get("max_output", default_max_output)
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "input_schema": TextCompletionRequest,
+                "output_schema": TextCompletionResponse,
+                "model": model,
+                "temperature": temperature,
+                "max_output": max_output,
+                "url" : url,
+            }
+        )
+
+        if not hasattr(__class__, "text_completion_metric"):
+            __class__.text_completion_metric = Histogram(
+                'text_completion_duration',
+                'Text completion duration (seconds)',
+                buckets=[
+                    0.25, 0.5, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0,
+                    8.0, 9.0, 10.0, 11.0, 12.0, 13.0, 14.0, 15.0, 16.0,
+                    17.0, 18.0, 19.0, 20.0, 21.0, 22.0, 23.0, 24.0, 25.0,
+                    30.0, 35.0, 40.0, 45.0, 50.0, 60.0, 80.0, 100.0,
+                    120.0
+                ]
+            )
+
+        self.model = model
+        self.url = url + "v1/"
+        self.temperature = temperature
+        self.max_output = max_output
+        self.openai = OpenAI(
+            base_url=self.url, 
+            api_key = "sk-no-key-required",
+        )
+
+        print("Initialised", flush=True)
+
+    async def handle(self, msg):
+
+        v = msg.value()
+
+        # Sender-produced ID
+
+        id = msg.properties()["id"]
+
+        print(f"Handling prompt {id}...", flush=True)
+
+        prompt = v.system + "\n\n" + v.prompt
+
+        try:
+
+            # FIXME: Rate limits
+
+            with __class__.text_completion_metric.time():
+
+                print(prompt)
+
+                resp = self.openai.chat.completions.create(
+                    model=self.model,
+                    messages=[
+                        {"role": "user", "content": prompt}
+                    ]
+                    #temperature=self.temperature,
+                    #max_tokens=self.max_output,
+                    #top_p=1,
+                    #frequency_penalty=0,
+                    #presence_penalty=0,
+                    #response_format={
+                    #    "type": "text"
+                    #}
+                )
+
+                print(resp)
+
+            inputtokens = resp.usage.prompt_tokens
+            outputtokens = resp.usage.completion_tokens
+
+            print(resp.choices[0].message.content, flush=True)
+            print(f"Input Tokens: {inputtokens}", flush=True)
+            print(f"Output Tokens: {outputtokens}", flush=True)
+
+            print("Send response...", flush=True)
+            r = TextCompletionResponse(
+                response=resp.choices[0].message.content,
+                error=None,
+                in_token=inputtokens,
+                out_token=outputtokens,
+                model=self.model,
+            )
+            await self.send(r, properties={"id": id})
+
+            print("Done.", flush=True)
+
+            # SLM, presumably there aren't rate limits
+
+        except Exception as e:
+
+            print(f"Exception: {e}")
+
+            print("Send error response...", flush=True)
+
+            r = TextCompletionResponse(
+                error=Error(
+                    type = "llm-error",
+                    message = str(e),
+                ),
+                response=None,
+                in_token=None,
+                out_token=None,
+                model=None,
+            )
+
+            await self.send(r, properties={"id": id})
+
+            self.consumer.acknowledge(msg)
+
+    @staticmethod
+    def add_args(parser):
+
+        ConsumerProducer.add_args(
+            parser, default_input_queue, default_subscriber,
+            default_output_queue,
+        )
+
+        parser.add_argument(
+            '-m', '--model',
+            default=default_model,
+            help=f'LLM model (default: gemma3:9b)'
+        )
+
+        parser.add_argument(
+            '-u', '--url',
+            default=default_url,
+            help=f'LMStudio URL (default: {default_url})'
+        )
+
+        parser.add_argument(
+            '-t', '--temperature',
+            type=float,
+            default=default_temperature,
+            help=f'LLM temperature parameter (default: {default_temperature})'
+        )
+
+        parser.add_argument(
+            '-x', '--max-output',
+            type=int,
+            default=default_max_output,
+            help=f'LLM max output tokens (default: {default_max_output})'
+        )
+
+def run():
+    Processor.launch(module, __doc__)
+

trustgraph-flow-0.21.10/trustgraph/model/text_completion/openai/__init__.py

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

trustgraph-flow-0.21.10/trustgraph/model/text_completion/openai/__main__.py

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

{trustgraph-flow-0.21.9 → trustgraph-flow-0.21.10}/trustgraph_flow.egg-info/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: trustgraph-flow
-Version: 0.21.9
+Version: 0.21.10
 Summary: TrustGraph provides a means to run a pipeline of flexible AI processing components in a flexible means to achieve a processing pipeline.
 Home-page: https://github.com/trustgraph-ai/trustgraph
-Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.21.9.tar.gz
+Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.21.10.tar.gz
 Author: trustgraph.ai
 Author-email: security@trustgraph.ai
 Classifier: Programming Language :: Python :: 3

{trustgraph-flow-0.21.9 → trustgraph-flow-0.21.10}/trustgraph_flow.egg-info/SOURCES.txt

@@ -40,6 +40,7 @@ scripts/text-completion-claude
 scripts/text-completion-cohere
 scripts/text-completion-googleaistudio
 scripts/text-completion-llamafile
+scripts/text-completion-lmstudio
 scripts/text-completion-mistral
 scripts/text-completion-ollama
 scripts/text-completion-openai
@@ -182,6 +183,9 @@ trustgraph/model/text_completion/googleaistudio/llm.py
 trustgraph/model/text_completion/llamafile/__init__.py
 trustgraph/model/text_completion/llamafile/__main__.py
 trustgraph/model/text_completion/llamafile/llm.py
+trustgraph/model/text_completion/lmstudio/__init__.py
+trustgraph/model/text_completion/lmstudio/__main__.py
+trustgraph/model/text_completion/lmstudio/llm.py
 trustgraph/model/text_completion/mistral/__init__.py
 trustgraph/model/text_completion/mistral/__main__.py
 trustgraph/model/text_completion/mistral/llm.py

trustgraph-flow-0.21.9/trustgraph/flow_version.py

@@ -1 +0,0 @@
-__version__ = "0.21.9"