Prepare implementation IBM WatsonxAI to langchain community package. (#…

…10)

Prepare implementation IBM WatsonxAI to langchain community package.

Provide solutions for:

- LLMs
- embeddings
- tests

docs/core_docs/docs/integrations/llms/ibm.mdx

@@ -0,0 +1,166 @@
+# @langchain/community/llm/ibm
+
+This is an intergation for the LangChain.js community with Watsonx AI by IBM through their SDK.
+
+## Installation
+
+```bash npm2yarn
+npm install @langchain/community @langchain/core
+```
+
+## Chat Models
+
+This package contains the `WatsonxLLM` class, which is the recommended way to interface with the Watsonx series of models.
+
+To use, install the requirements, and configure your environment depending on what type od authentication you will be using.
+
+## IAM authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=iam
+export WATSONX_AI_APIKEY=<YOUR-APIKEY>
+```
+
+## Bearer token authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=bearertoken
+export WATSONX_AI_BEARER_TOKEN=<YOUR-BEARER-TOKEN>
+```
+
+### CP4D authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=cp4d
+export WATSONX_AI_USERNAME=<YOUR_USERNAME>
+export WATSONX_AI_PASSWORD=<YOUR_PASSWORD>
+export WATSONX_AI_URL=<URL>
+```
+
+Once these are places in your enviromental variables and object is initialized authentication will proceed automatically.
+
+Authentication can also be accomplished by passing these values as parameters to a new instance.
+
+## IAM authentication
+
+```typescript
+import { WatsonxLLM } from "@langchain/community/llms/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "iam",
+  watsonxAIApikey: "<YOUR-APIKEY>",
+};
+const instance = new WatsonxLLM(props);
+```
+
+## Bearer token authentication
+
+```typescript
+import { WatsonxLLM } from "@langchain/community/llms/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "bearertoken",
+  watsonxAIBearerToken: "<YOUR-BEARERTOKEN>",
+};
+const instance = new WatsonxLLM(props);
+```
+
+### CP4D authentication
+
+```typescript
+import { WatsonxLLM } from "@langchain/community/llms/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "cp4d",
+  watsonxAIUsername: "<YOUR-USERNAME>",
+  watsonxAIPassword: "<YOUR-PASSWORD>",
+  watsonxAIUrl: "<url>",
+};
+const instance = new WatsonxLLM(props);
+```
+
+## Loading the model
+
+You might need to adjust model parameters for different models or tasks. For more details on the parameters, refer to IBM's documentation.
+
+```typescript
+import { WatsonxLLM } from "@langchain/community/llms/ibm";
+
+const props = {
+  decoding_method: "sample",
+  max_new_tokens: 100,
+  min_new_tokens: 1,
+  temperature: 0.5,
+  top_k: 50,
+  top_p: 1,
+};
+const instance = new WatsonxLLM({
+  version: "YYYY-MM-DD",
+  serviceUrl: process.env.API_URL,
+  projectId: "<PROJECT_ID>",
+  spaceId: "<SPACE_ID>",
+  idOrName: "<DEPLOYMENT_ID>",
+  modelId: "<MODEL_ID>",
+  ...props,
+});
+```
+
+Note:
+
+- You must provide spaceId, projectId or idOrName(deployment id) in order to proceed.
+- Depending on the region of your provisioned service instance, use correct serviceUrl.
+- You need to specify the model you want to use for inferencing through model_id.
+
+## Props overwrittion
+
+Passed props at initialization will last for the whole life cycle of the object, however you may overwrite them for a single method's call by passing second argument as below
+
+```typescript
+const result = await instance.invoke("Print hello world.", {
+  modelId: "<NEW_MODEL_ID>",
+  parameters: {
+    max_new_tokens: 20,
+  },
+});
+console.log(result);
+```
+
+## Text generation
+
+```typescript
+const result = await instance.invoke("Print hello world.");
+console.log(result);
+
+const results = await instance.generate([
+  "Print hello world.",
+  "Print bye, bye world!",
+]);
+console.log(result);
+```
+
+## Streaming
+
+```typescript
+const result = await instance.stream("Print hello world.");
+for await (let chunk of result) {
+  console.log(chunk);
+}
+```
+
+## Tokenization
+
+This package has it's custom getNumTokens implementation which returns exact amount of tokens that would be used.
+
+```typescript
+const tokens = await instance.getNumTokens("Print hello world.");
+console.log(tokens);
+```

docs/core_docs/docs/integrations/text_embedding/ibm.mdx

@@ -0,0 +1,130 @@
+# @langchain/community/embeddings/ibm
+
+This is an intergation for the LangChain.js community with Watsonx AI by IBM through their SDK.
+
+## Installation
+
+```bash npm2yarn
+npm install @langchain/community @langchain/core
+```
+
+## Embeddings
+
+This package contains the `WatsonxEmbeddings` class, which is the recommended way to interface with the Watsonx series of models.
+
+To use, install the requirements, and configure your environment depending on what type od authentication you will be using.
+
+## IAM authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=iam
+export WATSONX_AI_APIKEY=<YOUR-APIKEY>
+```
+
+## Bearer token authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=bearertoken
+export WATSONX_AI_BEARER_TOKEN=<YOUR-BEARER-TOKEN>
+```
+
+### CP4D authentication
+
+```bash
+export WATSONX_AI_AUTH_TYPE=cp4d
+export WATSONX_AI_USERNAME=<YOUR_USERNAME>
+export WATSONX_AI_PASSWORD=<YOUR_PASSWORD>
+export WATSONX_AI_URL=<URL>
+```
+
+Once these are places in your enviromental variables and object is initialized authentication will proceed automatically.
+
+Authentication can also be accomplished by passing these values as parameters to a new instance.
+
+## IAM authentication
+
+```typescript
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "iam",
+  watsonxAIApikey: "<YOUR-APIKEY>",
+};
+const instance = new WatsonxEmbeddings(props);
+```
+
+## Bearer token authentication
+
+```typescript
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "bearertoken",
+  watsonxAIBearerToken: "<YOUR-BEARERTOKEN>",
+};
+const instance = new WatsonxEmbeddings(props);
+```
+
+### CP4D authentication
+
+```typescript
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const props = {
+  version: "YYYY-MM-DD",
+  serviceUrl: "<SERVICE_URL>",
+  projectId: "<PROJECT_ID>",
+  watsonxAIAuthType: "cp4d",
+  watsonxAIUsername: "<YOUR-USERNAME>",
+  watsonxAIPassword: "<YOUR-PASSWORD>",
+  watsonxAIUrl: "<url>",
+};
+const instance = new WatsonxEmbeddings(props);
+```
+
+## Loading the model
+
+```typescript
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const instance = new WatsonxEmbeddings({
+  version: "YYYY-MM-DD",
+  serviceUrl: process.env.API_URL,
+  projectId: "<PROJECT_ID>",
+  spaceId: "<SPACE_ID>",
+  idOrName: "<DEPLOYMENT_ID>",
+  modelId: "<MODEL_ID>",
+});
+```
+
+Note:
+
+- You must provide spaceId, projectId or idOrName(deployment id) in order to proceed.
+- Depending on the region of your provisioned service instance, use correct serviceUrl.
+- You need to specify the model you want to use for inferencing through model_id.
+
+## Embeddings
+
+Following package supports embeddings model, you can proceed with following code snipet.
+
+```typescript
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const instance = new WatsonxEmbeddings({
+  version: "YYYY-MM-DD",
+  serviceUrl: process.env.API_URL,
+  projectId: "<PROJECT_ID>",
+  spaceId: "<SPACE_ID>",
+  idOrName: "<DEPLOYMENT_ID>",
+  modelId: "<MODEL_ID>",
+});
+
+const result = await instance.embedQuery("Hello world!");
+console.log(result);
+```

examples/src/embeddings/ibm.ts

@@ -0,0 +1,10 @@
+import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";
+
+const instance = new WatsonxEmbeddings({
+  version: "YYYY-MM-DD",
+  serviceUrl: process.env.WATSONX_AI_SERVICE_URL as string,
+  projectId: process.env.WATSONX_AI_PROJECT_ID,
+});
+
+const result = await instance.embedQuery("Hello world!");
+console.log(result);

examples/src/llms/ibm.ts

@@ -0,0 +1,30 @@
+import { WatsonxLLM } from "@langchain/community/llms/ibm";
+
+const props = {
+  decoding_method: "sample",
+  max_new_tokens: 100,
+  min_new_tokens: 1,
+  temperature: 0.5,
+  top_k: 50,
+  top_p: 1,
+};
+const instance = new WatsonxLLM({
+  version: "2024-05-31",
+  serviceUrl: process.env.WATSONX_AI_SERVICE_URL as string,
+  projectId: process.env.WATSONX_AI_PROJECT_ID,
+  ...props,
+});
+
+const result = await instance.invoke("Print hello world.");
+console.log(result);
+
+const results = await instance.generate([
+  "Print hello world.",
+  "Print bye, bye world!",
+]);
+console.log(results);
+
+const stream = await instance.stream("Print hello world.");
+for await (const chunk of stream) {
+  console.log(chunk);
+}

libs/langchain-community/.gitignore

@@ -162,6 +162,10 @@ embeddings/hf_transformers.cjs
 embeddings/hf_transformers.js
 embeddings/hf_transformers.d.ts
 embeddings/hf_transformers.d.cts
+embeddings/ibm.cjs
+embeddings/ibm.js
+embeddings/ibm.d.ts
+embeddings/ibm.d.cts
 embeddings/jina.cjs
 embeddings/jina.js
 embeddings/jina.d.ts
@@ -254,6 +258,10 @@ llms/hf.cjs
 llms/hf.js
 llms/hf.d.ts
 llms/hf.d.cts
+llms/ibm.cjs
+llms/ibm.js
+llms/ibm.d.ts
+llms/ibm.d.cts
 llms/llama_cpp.cjs
 llms/llama_cpp.js
 llms/llama_cpp.d.ts