Update README and some examples.

andrewyng · Sep 26, 2024 · 5951c0c · 5951c0c
1 parent 2bca4de
commit 5951c0c
Show file tree

Hide file tree

Showing 2 changed files with 53 additions and 906 deletions.
diff --git a/examples/client.ipynb b/examples/client.ipynb
@@ -11,7 +11,7 @@
    "source": [
     "# Client Examples\n",
     "\n",
-    "Client provides a uniform interface for interacting with LLMs from various providers. It adapts the official python libraries from providers such as Mistral, OpenAI, Groq, Anthropic, Fireworks, Replicate, etc. to conform to the OpenAI chat completion interface.\n",
+    "Client provides a uniform interface for interacting with LLMs from various providers. It adapts the official python libraries from providers such as Mistral, OpenAI, Groq, Anthropic, AWS, etc to conform to the OpenAI chat completion interface. It directly calls the REST endpoints in some cases.\n",
     "\n",
     "Below are some examples of how to use Client to interact with different LLMs."
    ]
@@ -55,15 +55,12 @@
     "        for key, value in additional_env_vars.items():\n",
     "            os.environ[key] = value\n",
     "\n",
-    "# Define additional API keys and AWS credentials\n",
+    "# Define additional API keys and credentials\n",
     "additional_keys = {\n",
     "    'GROQ_API_KEY': 'xxx',\n",
-    "    'FIREWORKS_API_KEY': 'xxx', \n",
-    "    'REPLICATE_API_KEY': 'xxx', \n",
-    "    'TOGETHER_API_KEY': 'xxx', \n",
-    "    'OCTO_API_KEY': 'xxx',\n",
     "    'AWS_ACCESS_KEY_ID': 'xxx',\n",
     "    'AWS_SECRET_ACCESS_KEY': 'xxx',\n",
+    "    'ANTHROPIC_API_KEY': 'xxx',\n",
     "}\n",
     "\n",
     "# Configure environment\n",
@@ -111,9 +108,6 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# print(os.environ['AWS_SECRET_ACCESS_KEY'])\n",
-    "# print(os.environ['AWS_ACCESS_KEY_ID'])\n",
-    "# print(os.environ['AWS_REGION'])\n",
     "aws_bedrock_llama3_8b = \"aws-bedrock:meta.llama3-1-8b-instruct-v1:0\"\n",
     "response = client.chat.completions.create(model=aws_bedrock_llama3_8b, messages=messages)\n",
     "print(response.choices[0].message.content)"
@@ -126,15 +120,16 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# client2 = ai.Client({\"azure\" : {\n",
-    "#   \"api_key\": os.environ[\"AZURE_API_KEY\"],\n",
-    "# }});\n",
+    "# IMP NOTE: Azure expects model endpoint to be passed in the format of \"azure:<model_name>\".\n",
+    "# The model name is the deployment name in Project/Deployments.\n",
+    "# In the exmaple below, the model is \"mistral-large-2407\", but the name given to the\n",
+    "# deployment is \"aisuite-mistral-large-2407\" under the deployments section in Azure.\n",
     "client2 = ai.Client()\n",
     "client2.configure({\"azure\" : {\n",
     "  \"api_key\": os.environ[\"AZURE_API_KEY\"],\n",
-    "  \"base_url\": \"https://mistral-large-2407.westus3.models.ai.azure.com/v1/\",\n",
+    "  \"base_url\": \"https://aisuite-mistral-large-2407.westus3.models.ai.azure.com/v1/\",\n",
     "}});\n",
-    "azure_model = \"azure:mistral-large-2407\"\n",
+    "azure_model = \"azure:aisuite-mistral-large-2407\"\n",
     "response = client2.chat.completions.create(model=azure_model, messages=messages)\n",
     "print(response.choices[0].message.content)"
    ]
@@ -146,6 +141,10 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "# HuggingFace expects the model to be passed in the format of \"huggingface:<model_name>\".\n",
+    "# The model name is the full name of the model in HuggingFace.\n",
+    "# In the exmaple below, the model is \"mistralai/Mistral-7B-Instruct-v0.3\".\n",
+    "# The model is deployed as serverless inference endpoint in HuggingFace.\n",
     "client3 = ai.Client()\n",
     "hf_model = \"huggingface:mistralai/Mistral-7B-Instruct-v0.3\"\n",
     "response = client3.chat.completions.create(model=hf_model, messages=messages)\n",
@@ -159,11 +158,14 @@
    "metadata": {},
    "outputs": [],
    "source": [
+    "\n",
+    "# Groq expects the model to be passed in the format of \"groq:<model_name>\".\n",
+    "# The model name is the full name of the model in Groq.\n",
+    "# In the exmaple below, the model is \"llama3-8b-8192\".\n",
     "groq_llama3_8b = \"groq:llama3-8b-8192\"\n",
     "# groq_llama3_70b = \"groq:llama3-70b-8192\"\n",
     "\n",
     "response = client.chat.completions.create(model=groq_llama3_8b, messages=messages)\n",
-    "\n",
     "print(response.choices[0].message.content)"
    ]
   },
@@ -193,9 +195,7 @@
    "outputs": [],
    "source": [
     "mistral_7b = \"mistral:open-mistral-7b\"\n",
-    "\n",
     "response = client.chat.completions.create(model=mistral_7b, messages=messages, temperature=0.2)\n",
-    "\n",
     "print(response.choices[0].message.content)"
    ]
   },
@@ -207,9 +207,7 @@
    "outputs": [],
    "source": [
     "openai_gpt35 = \"openai:gpt-3.5-turbo\"\n",
-    "\n",
     "response = client.chat.completions.create(model=openai_gpt35, messages=messages, temperature=0.75)\n",
-    "\n",
     "print(response.choices[0].message.content)"
    ]
   }
@@ -230,7 +228,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.12.4"
+   "version": "3.12.6"
   }
  },
  "nbformat": 4,