docs: add mistral instrumentation to mistral example

Arize-ai · Mar 22, 2024 · eaae4a8 · eaae4a8
1 parent f282778
commit eaae4a8
Showing 1 changed file with 52 additions and 12 deletions.
diff --git a/tutorials/mistral/evaluate_rag--mistral.ipynb b/tutorials/mistral/evaluate_rag--mistral.ipynb
@@ -77,7 +77,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "!pip install -qq \"arize-phoenix\" \"arize-phoenix-evals>=0.5.0\" \"llama-index==0.10.19\" \"llama-index-llms-mistralai\" \"llama-index-embeddings-mistralai\"  \"openinference-instrumentation-mistralai\" \"openinference-instrumentation-llama-index>=1.0.0\" \"llama-index-callbacks-arize-phoenix>=0.1.2\" gcsfs nest_asyncio"
+    "!pip install -qq \"arize-phoenix\" \"arize-phoenix-evals>=0.5.0\" \"llama-index==0.10.19\" \"llama-index-llms-mistralai\" \"llama-index-embeddings-mistralai\"  \"openinference-instrumentation-mistralai>=0.0.2\" \"openinference-instrumentation-llama-index>=1.0.0\" \"llama-index-callbacks-arize-phoenix>=0.1.2\" gcsfs nest_asyncio"
    ]
   },
   {
@@ -130,6 +130,48 @@
     "nest_asyncio.apply()"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import phoenix as px\n",
+    "from llama_index.core import set_global_handler\n",
+    "\n",
+    "session = px.launch_app()\n",
+    "set_global_handler(\"arize_phoenix\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Let's use the OpenInference Mistral Instrumentation as well."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from openinference.instrumentation.mistralai import MistralAIInstrumentor\n",
+    "from opentelemetry import trace as trace_api\n",
+    "from opentelemetry.exporter.otlp.proto.http.trace_exporter import OTLPSpanExporter\n",
+    "from opentelemetry.sdk import trace as trace_sdk\n",
+    "from opentelemetry.sdk.trace.export import ConsoleSpanExporter, SimpleSpanProcessor\n",
+    "\n",
+    "tracer_provider = trace_sdk.TracerProvider()\n",
+    "tracer_provider.add_span_processor(SimpleSpanProcessor(OTLPSpanExporter(\"http://127.0.0.1:6006/v1/traces\")))\n",
+    "# Optionally, you can also print the spans to the console.\n",
+    "tracer_provider.add_span_processor(SimpleSpanProcessor(ConsoleSpanExporter()))\n",
+    "trace_api.set_tracer_provider(tracer_provider)\n",
+    "\n",
+    "mistral_instrumentor = MistralAIInstrumentor()\n",
+    "mistral_instrumentor.instrument()"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": null,
@@ -138,7 +180,7 @@
    "source": [
     "import pandas as pd\n",
     "import phoenix as px\n",
-    "from llama_index.core import Settings, SimpleDirectoryReader, VectorStoreIndex, set_global_handler\n",
+    "from llama_index.core import Settings, SimpleDirectoryReader, VectorStoreIndex\n",
     "from llama_index.core.node_parser import SimpleNodeParser\n",
     "from llama_index.embeddings.mistralai import MistralAIEmbedding\n",
     "from llama_index.llms.mistralai import MistralAI\n",
@@ -440,7 +482,6 @@
    "source": [
     "import json\n",
     "\n",
-    "from openinference.instrumentation.mistralai import MistralAIInstrumentor\n",
     "from phoenix.evals import MistralAIModel, llm_generate\n",
     "\n",
     "\n",
@@ -452,16 +493,15 @@
     "\n",
     "\n",
     "with using_project(TESTSET_PROJECT):  # Collect traces under the project \"testset\"\n",
-    "    mistral_instrumentor = MistralAIInstrumentor()\n",
-    "    mistral_instrumentor.instrument() # Instrument MistralAI to see the synthetic testset generation\n",
     "    questions_df = llm_generate(\n",
     "        dataframe=document_chunks_df,\n",
     "        template=generate_questions_template,\n",
-    "        model=MistralAIModel(model=\"mistral-large-latest\", response_format={\"type\": \"json_object\"}),\n",
+    "        model=MistralAIModel(model=\"mistral-large-latest\"),\n",
     "        output_parser=output_parser,\n",
     "        concurrency=20,\n",
     "    )\n",
-    "    mistral_instrumentor.uninstrument() # No longer needed"
+    "\n",
+    "mistral_instrumentor.uninstrument() # No longer needed"
    ]
   },
   {
@@ -1057,11 +1097,11 @@
     "    )\n",
     ")\n",
     "# relaunch phoenix with a primary and corpus dataset to view embeddings\n",
-    "px.close_app()\n",
-    "session = px.launch_app(\n",
-    "    primary=px.Dataset(query_df, query_schema, \"query\"),\n",
-    "    corpus=px.Dataset(corpus_df.reset_index(drop=True), corpus_schema, \"corpus\"),\n",
-    ")"
+    "# px.close_app()\n",
+    "# session = px.launch_app(\n",
+    "#     primary=px.Dataset(query_df, query_schema, \"query\"),\n",
+    "#     corpus=px.Dataset(corpus_df.reset_index(drop=True), corpus_schema, \"corpus\"),\n",
+    "# )"
    ]
   },
   {