bedrock-wrapper 1.0.13 → 1.0.15

package/README.md

@@ -1,5 +1,5 @@
 # 🪨 Bedrock Wrapper
-Bedrock Wrapper is an npm package that simplifies the integration of existing OpenAI-compatible API objects with AWS Bedrock's serverless inference LLMs.  Follow the steps below to integrate into your own application, or alternativly use the 🪨 [Bedrock Proxy Endpoint](https://github.com/jparkerweb/bedrock-proxy-endpoint) project to spin up your own custom OpenAI server endpoint for even easier inference (using the standard `baseUrl`, and `apiKey` params).
+Bedrock Wrapper is an npm package that simplifies the integration of existing OpenAI-compatible API objects with AWS Bedrock's serverless inference LLMs.  Follow the steps below to integrate into your own application, or alternativly use the 🔀 [Bedrock Proxy Endpoint](https://github.com/jparkerweb/bedrock-proxy-endpoint) project to spin up your own custom OpenAI server endpoint for even easier inference (using the standard `baseUrl`, and `apiKey` params).
 
 ---
 
@@ -98,6 +98,7 @@ Bedrock Wrapper is an npm package that simplifies the integration of existing Op
 |----------------|------------------------------------|
 | Llama-3-8b     | meta.llama3-8b-instruct-v1:0       |
 | Llama-3-70b    | meta.llama3-70b-instruct-v1:0      |
+| Mistral-7b     | mistral.mistral-7b-instruct-v0:2   |
 | Mixtral-8x7b   | mistral.mixtral-8x7b-instruct-v0:1 |
 | Mistral-Large  | mistral.mistral-large-2402-v1:0    |
 

package/bedrock-models.js

@@ -22,6 +22,7 @@ export const bedrock_models = [
         "eom_text":                      "<|eot_id|>",
         "display_role_names":            true,
         "max_tokens_param_name":         "max_gen_len",
+        "max_supported_response_tokens": 2048,
         "response_chunk_element":        "generation",
     },
     {
@@ -46,8 +47,34 @@ export const bedrock_models = [
         "eom_text":                      "<|eot_id|>",
         "display_role_names":            true,
         "max_tokens_param_name":         "max_gen_len",
+        "max_supported_response_tokens": 2048,
         "response_chunk_element":        "generation",
     },
+    {
+        // ================
+        // == Mistral-7b ==
+        // ================
+        "modelName":                     "Mistral-7b",
+        "modelId":                       "mistral.mistral-7b-instruct-v0:2",
+        "bos_text":                      "<s>",
+        "role_system_message_prefix":    "",
+        "role_system_message_suffix":    "",
+        "role_system_prefix":            "",
+        "role_system_suffix":            "",
+        "role_user_message_prefix":      "[INST]",
+        "role_user_message_suffix":      "[/INST]",
+        "role_user_prefix":              "",
+        "role_user_suffix":              "",
+        "role_assistant_message_prefix": "",
+        "role_assistant_message_suffix": "",
+        "role_assistant_prefix":         "",
+        "role_assistant_suffix":         "",
+        "eom_text":                      "</s>",
+        "display_role_names":            false,
+        "max_tokens_param_name":         "max_tokens",
+        "max_supported_response_tokens": 8192,
+        "response_chunk_element":        "outputs[0].text",
+    },
     {
         // ==================
         // == Mixtral-8x7b ==
@@ -70,6 +97,7 @@ export const bedrock_models = [
         "eom_text":                      "</s>",
         "display_role_names":            false,
         "max_tokens_param_name":         "max_tokens",
+        "max_supported_response_tokens": 4096,
         "response_chunk_element":        "outputs[0].text",
     },
     {
@@ -94,6 +122,7 @@ export const bedrock_models = [
         "eom_text":                      "</s>",
         "display_role_names":            false,
         "max_tokens_param_name":         "max_tokens",
+        "max_supported_response_tokens": 8192,
         "response_chunk_element":        "outputs[0].text",
     },
 ];

package/bedrock-wrapper.js

@@ -96,11 +96,13 @@ export async function* bedrockWrapper(awsCreds, openaiChatCompletionsCreateObjec
         console.log(`\nPrompt: ${prompt}\n`);
     }
 
+    const max_gen_tokens = max_tokens <= awsModel.max_supported_response_tokens ? max_tokens : awsModel.max_supported_response_tokens;
+
     // Format the request payload using the model's native structure.
     const request = {
         prompt,
         // Optional inference parameters:
-        [awsModel.max_tokens_param_name]: max_tokens,
+        [awsModel.max_tokens_param_name]: max_gen_tokens,
         temperature: temperature,
         top_p: top_p,
     };

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "bedrock-wrapper",
-  "version": "1.0.13",
+  "version": "1.0.15",
   "description": "🪨 Bedrock Wrapper is an npm package that simplifies the integration of existing OpenAI-compatible API objects with AWS Bedrock's serverless inference LLMs.",
   "repository": {
 		"type": "git",
@@ -21,7 +21,7 @@
   "author": "",
   "license": "ISC",
   "dependencies": {
-    "@aws-sdk/client-bedrock-runtime": "^3.567.0",
+    "@aws-sdk/client-bedrock-runtime": "^3.575.0",
     "dotenv": "^16.4.5"
   }
 }