trustgraph-flow 0.20.4__tar.gz → 0.20.6__tar.gz

{trustgraph-flow-0.20.4 → trustgraph-flow-0.20.6}/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: trustgraph-flow
-Version: 0.20.4
+Version: 0.20.6
 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/v0.20.4.tar.gz
+Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.20.6.tar.gz
 Author: trustgraph.ai
 Author-email: security@trustgraph.ai
 Classifier: Programming Language :: Python :: 3

trustgraph-flow-0.20.6/scripts/embeddings-fastembed

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

{trustgraph-flow-0.20.4 → trustgraph-flow-0.20.6}/setup.py

@@ -41,6 +41,7 @@ setuptools.setup(
         "cohere",
         "cryptography",
         "falkordb",
+        "fastembed",
         "google-generativeai",
         "ibis",
         "jsonschema",
@@ -49,6 +50,7 @@ setuptools.setup(
         "langchain-core",
         "langchain-text-splitters",
         "neo4j",
+        "ollama",
         "openai",
         "pinecone[grpc]",
         "prometheus-client",
@@ -77,6 +79,7 @@ setuptools.setup(
         "scripts/document-embeddings",
         "scripts/document-rag",
         "scripts/embeddings-ollama",
+        "scripts/embeddings-fastembed",
         "scripts/ge-query-milvus",
         "scripts/ge-query-pinecone",
         "scripts/ge-query-qdrant",

{trustgraph-flow-0.20.4/trustgraph/embeddings/ollama → trustgraph-flow-0.20.6/trustgraph/embeddings/fastembed}/processor.py

@@ -3,20 +3,20 @@
 Embeddings service, applies an embeddings model selected from HuggingFace.
 Input is text, output is embeddings vector.
 """
-from langchain_community.embeddings import OllamaEmbeddings
 
 from ... schema import EmbeddingsRequest, EmbeddingsResponse
 from ... schema import embeddings_request_queue, embeddings_response_queue
 from ... log_level import LogLevel
 from ... base import ConsumerProducer
+from fastembed import TextEmbedding
+import os
 
 module = ".".join(__name__.split(".")[1:-1])
 
 default_input_queue = embeddings_request_queue
 default_output_queue = embeddings_response_queue
 default_subscriber = module
-default_model="mxbai-embed-large"
-default_ollama = 'http://localhost:11434'
+default_model="sentence-transformers/all-MiniLM-L6-v2"
 
 class Processor(ConsumerProducer):
 
@@ -26,6 +26,8 @@ class Processor(ConsumerProducer):
         output_queue = params.get("output_queue", default_output_queue)
         subscriber = params.get("subscriber", default_subscriber)
 
+        model = params.get("model", default_model)
+
         super(Processor, self).__init__(
             **params | {
                 "input_queue": input_queue,
@@ -33,10 +35,11 @@ class Processor(ConsumerProducer):
                 "subscriber": subscriber,
                 "input_schema": EmbeddingsRequest,
                 "output_schema": EmbeddingsResponse,
+                "model": model,
             }
         )
 
-        self.embeddings = OllamaEmbeddings(base_url=ollama, model=model)
+        self.embeddings = TextEmbedding(model_name = model)
 
     def handle(self, msg):
 
@@ -49,10 +52,18 @@ class Processor(ConsumerProducer):
         print(f"Handling input {id}...", flush=True)
 
         text = v.text
-        embeds = self.embeddings.embed_query([text])
+        vecs = self.embeddings.embed([text])
+
+        vecs = [
+            v.tolist()
+            for v in vecs
+        ]
 
         print("Send response...", flush=True)
-        r = EmbeddingsResponse(vectors=[embeds])
+        r = EmbeddingsResponse(
+            vectors=list(vecs),
+            error=None,
+        )
 
         self.producer.send(r, properties={"id": id})
 
@@ -72,12 +83,6 @@ class Processor(ConsumerProducer):
             help=f'Embeddings model (default: {default_model})'
         )
 
-        parser.add_argument(
-            '-r', '--ollama',
-            default=default_ollama,
-            help=f'ollama (default: {default_ollama})'
-        )
-
 def run():
 
     Processor.start(module, __doc__)

trustgraph-flow-0.20.6/trustgraph/embeddings/ollama/__init__.py

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

trustgraph-flow-0.20.6/trustgraph/embeddings/ollama/__main__.py

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

trustgraph-flow-0.20.6/trustgraph/embeddings/ollama/processor.py

@@ -0,0 +1,97 @@
+
+"""
+Embeddings service, applies an embeddings model hosted on a local Ollama.
+Input is text, output is embeddings vector.
+"""
+
+from ... schema import EmbeddingsRequest, EmbeddingsResponse
+from ... schema import embeddings_request_queue, embeddings_response_queue
+from ... log_level import LogLevel
+from ... base import ConsumerProducer
+from ollama import Client
+import os
+
+module = ".".join(__name__.split(".")[1:-1])
+
+default_input_queue = embeddings_request_queue
+default_output_queue = embeddings_response_queue
+default_subscriber = module
+default_model="mxbai-embed-large"
+default_ollama = os.getenv("OLLAMA_HOST", '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)
+
+        ollama = params.get("ollama", default_ollama)
+        model = params.get("model", default_model)
+
+        super(Processor, self).__init__(
+            **params | {
+                "input_queue": input_queue,
+                "output_queue": output_queue,
+                "subscriber": subscriber,
+                "input_schema": EmbeddingsRequest,
+                "output_schema": EmbeddingsResponse,
+                "ollama": ollama,
+                "model": model,
+            }
+        )
+
+        self.client = Client(host=ollama)
+        self.model = model
+
+    def handle(self, msg):
+
+        v = msg.value()
+
+        # Sender-produced ID
+
+        id = msg.properties()["id"]
+
+        print(f"Handling input {id}...", flush=True)
+
+        text = v.text
+        embeds = self.client.embed(
+            model = self.model,
+            input = text
+        )
+
+        print("Send response...", flush=True)
+        r = EmbeddingsResponse(
+            vectors=embeds.embeddings,
+            error=None,
+        )
+
+        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(
+            '-m', '--model',
+            default=default_model,
+            help=f'Embeddings model (default: {default_model})'
+        )
+
+        parser.add_argument(
+            '-r', '--ollama',
+            default=default_ollama,
+            help=f'ollama (default: {default_ollama})'
+        )
+
+def run():
+
+    Processor.start(module, __doc__)
+

trustgraph-flow-0.20.6/trustgraph/flow_version.py

@@ -0,0 +1 @@
+__version__ = "0.20.6"

{trustgraph-flow-0.20.4 → trustgraph-flow-0.20.6}/trustgraph_flow.egg-info/PKG-INFO

@@ -1,9 +1,9 @@
 Metadata-Version: 2.1
 Name: trustgraph-flow
-Version: 0.20.4
+Version: 0.20.6
 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/v0.20.4.tar.gz
+Download-URL: https://github.com/trustgraph-ai/trustgraph/archive/refs/tags/v0.20.6.tar.gz
 Author: trustgraph.ai
 Author-email: security@trustgraph.ai
 Classifier: Programming Language :: Python :: 3

{trustgraph-flow-0.20.4 → trustgraph-flow-0.20.6}/trustgraph_flow.egg-info/SOURCES.txt

@@ -12,6 +12,7 @@ scripts/de-write-pinecone
 scripts/de-write-qdrant
 scripts/document-embeddings
 scripts/document-rag
+scripts/embeddings-fastembed
 scripts/embeddings-ollama
 scripts/ge-query-milvus
 scripts/ge-query-pinecone
@@ -80,6 +81,9 @@ trustgraph/embeddings/__init__.py
 trustgraph/embeddings/document_embeddings/__init__.py
 trustgraph/embeddings/document_embeddings/__main__.py
 trustgraph/embeddings/document_embeddings/embeddings.py
+trustgraph/embeddings/fastembed/__init__.py
+trustgraph/embeddings/fastembed/__main__.py
+trustgraph/embeddings/fastembed/processor.py
 trustgraph/embeddings/graph_embeddings/__init__.py
 trustgraph/embeddings/graph_embeddings/__main__.py
 trustgraph/embeddings/graph_embeddings/embeddings.py

{trustgraph-flow-0.20.4 → trustgraph-flow-0.20.6}/trustgraph_flow.egg-info/requires.txt

@@ -4,6 +4,7 @@ cassandra-driver
 cohere
 cryptography
 falkordb
+fastembed
 google-generativeai
 ibis
 jsonschema
@@ -12,6 +13,7 @@ langchain-community
 langchain-core
 langchain-text-splitters
 neo4j
+ollama
 openai
 pinecone[grpc]
 prometheus-client

trustgraph-flow-0.20.4/trustgraph/flow_version.py

@@ -1 +0,0 @@
-__version__ = "0.20.4"