Merge pull request #5642 from BerriAI/litellm_otel_fixes

[Fix-Perf] OTEL use sensible default values for logging

litellm/integrations/opentelemetry.py

@@ -650,8 +650,6 @@ class OpenTelemetry(CustomLogger):
                 OTLPSpanExporterHTTP(
                     endpoint=self.OTEL_ENDPOINT, headers=_split_otel_headers
                 ),
-                max_queue_size=100,
-                max_export_batch_size=100,
             )
         elif self.OTEL_EXPORTER == "otlp_grpc":
             verbose_logger.debug(
@@ -662,8 +660,6 @@ class OpenTelemetry(CustomLogger):
                 OTLPSpanExporterGRPC(
                     endpoint=self.OTEL_ENDPOINT, headers=_split_otel_headers
                 ),
-                max_queue_size=100,
-                max_export_batch_size=100,
             )
         else:
             verbose_logger.debug(

litellm/model_prices_and_context_window_backup.json

@@ -2015,6 +2015,37 @@
         "supports_vision": true,
         "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models"
     },
+    "gemini-1.5-flash-exp-0827": {
+        "max_tokens": 8192,
+        "max_input_tokens": 1000000,
+        "max_output_tokens": 8192,
+        "max_images_per_prompt": 3000,
+        "max_videos_per_prompt": 10,
+        "max_video_length": 1,
+        "max_audio_length_hours": 8.4,
+        "max_audio_per_prompt": 1,
+        "max_pdf_size_mb": 30,
+        "input_cost_per_image": 0.0001315,
+        "input_cost_per_video_per_second": 0.0001315,
+        "input_cost_per_audio_per_second": 0.000125,
+        "input_cost_per_token": 0.0000005, 
+        "input_cost_per_character": 0.000000125, 
+        "input_cost_per_token_above_128k_tokens": 0.000001, 
+        "input_cost_per_character_above_128k_tokens": 0.00000025, 
+        "output_cost_per_token": 0.0000015,
+        "output_cost_per_character": 0.000000375,
+        "output_cost_per_token_above_128k_tokens": 0.000003,
+        "output_cost_per_character_above_128k_tokens": 0.00000075,
+        "output_cost_per_image": 0.000263,
+        "output_cost_per_video_per_second": 0.000263,
+        "output_cost_per_audio_per_second": 0.00025,
+        "litellm_provider": "vertex_ai-language-models",
+        "mode": "chat",
+        "supports_system_messages": true,
+        "supports_function_calling": true,
+        "supports_vision": true,
+        "source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models"
+    },
     "gemini-1.5-flash-001": {
         "max_tokens": 8192,
         "max_input_tokens": 1000000,
@@ -5586,4 +5617,4 @@
         "source": "https://www.databricks.com/product/pricing/foundation-model-serving",
         "metadata": {"notes": "Input/output cost per token is dbu cost * $0.070, based on databricks Llama 3.1 70B conversion. Number provided for reference, '*_dbu_cost_per_token' used in actual calculation."}
     }
-}
+}

litellm/proxy/proxy_config.yaml

@@ -16,6 +16,9 @@ general_settings:
  master_key: sk-1234 
 
 litellm_settings:
+
+  callbacks: ["otel"]
+
   success_callback: ["langsmith", "prometheus"]
   service_callback: ["prometheus_system"]
-  callbacks: ["otel"]
+