Azure env vars can provide endpoint/token

trustgraph-flow/trustgraph/model/text_completion/azure/llm.py

@@ -7,6 +7,7 @@ serverless endpoint service.  Input is prompt, output is response.
 import requests
 import json
 from prometheus_client import Histogram
+import os
 
 from .... schema import TextCompletionRequest, TextCompletionResponse, Error
 from .... schema import text_completion_request_queue
@@ -23,6 +24,8 @@ default_subscriber = module
 default_temperature = 0.0
 default_max_output = 4192
 default_model = "AzureAI"
+default_endpoint = os.getenv("AZURE_ENDPOINT")
+default_token = os.getenv("AZURE_TOKEN")
 
 class Processor(ConsumerProducer):
 
@@ -31,12 +34,18 @@ class Processor(ConsumerProducer):
         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")
+        endpoint = params.get("endpoint", default_endpoint)
-        token = params.get("token")
+        token = params.get("token", default_token)
         temperature = params.get("temperature", default_temperature)
         max_output = params.get("max_output", default_max_output)
         model = default_model
 
+        if endpoint is None:
+            raise RuntimeError("Azure endpoint not specified")
+
+        if token is None:
+            raise RuntimeError("Azure token not specified")
+
         super(Processor, self).__init__(
             **params | {
                 "input_queue": input_queue,

trustgraph-flow/trustgraph/model/text_completion/azure_openai/llm.py

@@ -8,6 +8,7 @@ import requests
 import json
 from prometheus_client import Histogram
 from openai import AzureOpenAI
+import os
 
 from .... schema import TextCompletionRequest, TextCompletionResponse, Error
 from .... schema import text_completion_request_queue
@@ -24,6 +25,8 @@ default_subscriber = module
 default_temperature = 0.0
 default_max_output = 4192
 default_api = "2024-02-15-preview"
+default_endpoint = os.getenv("AZURE_ENDPOINT")
+default_token = os.getenv("AZURE_TOKEN")
 
 class Processor(ConsumerProducer):
 
@@ -32,13 +35,19 @@ class Processor(ConsumerProducer):
         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")
+        endpoint = params.get("endpoint", default_endpoint)
-        token = params.get("token")
+        token = params.get("token", default_token)
         temperature = params.get("temperature", default_temperature)
         max_output = params.get("max_output", default_max_output)
         model = params.get("model")
         api = params.get("api_version", default_api)
 
+        if endpoint is None:
+            raise RuntimeError("Azure endpoint not specified")
+
+        if token is None:
+            raise RuntimeError("Azure token not specified")
+
         super(Processor, self).__init__(
             **params | {
                 "input_queue": input_queue,

trustgraph-flow/trustgraph/model/text_completion/claude/llm.py

@@ -6,6 +6,7 @@ Input is prompt, output is response.
 
 import anthropic
 from prometheus_client import Histogram
+import os
 
 from .... schema import TextCompletionRequest, TextCompletionResponse, Error
 from .... schema import text_completion_request_queue
@@ -22,6 +23,7 @@ default_subscriber = module
 default_model = 'claude-3-5-sonnet-20240620'
 default_temperature = 0.0
 default_max_output = 8192
+default_api_key = os.getenv("CLAUDE_KEY")
 
 class Processor(ConsumerProducer):
 
@@ -31,7 +33,7 @@ class Processor(ConsumerProducer):
         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")
+        api_key = params.get("api_key", default_api_key)
         temperature = params.get("temperature", default_temperature)
         max_output = params.get("max_output", default_max_output)