llama-stack-client 0.4.4 → 0.4.5

package/src/resources/vector-stores/files.ts

@@ -13,7 +13,7 @@ import { OpenAICursorPage, type OpenAICursorPageParams } from '../../pagination'
 
 export class Files extends APIResource {
   /**
-   * Attach a file to a vector store.
+   * Attach a file to a vector store (OpenAI-compatible).
    */
   create(
     vectorStoreId: string,
@@ -24,7 +24,7 @@ export class Files extends APIResource {
   }
 
   /**
-   * Retrieves a vector store file.
+   * Retrieve a vector store file (OpenAI-compatible).
    */
   retrieve(
     vectorStoreId: string,
@@ -35,7 +35,7 @@ export class Files extends APIResource {
   }
 
   /**
-   * Updates a vector store file.
+   * Update a vector store file (OpenAI-compatible).
    */
   update(
     vectorStoreId: string,
@@ -47,7 +47,7 @@ export class Files extends APIResource {
   }
 
   /**
-   * List files in a vector store.
+   * List files in a vector store (OpenAI-compatible).
    */
   list(
     vectorStoreId: string,
@@ -74,7 +74,7 @@ export class Files extends APIResource {
   }
 
   /**
-   * Delete a vector store file.
+   * Delete a vector store file (OpenAI-compatible).
    */
   delete(
     vectorStoreId: string,
@@ -85,7 +85,7 @@ export class Files extends APIResource {
   }
 
   /**
-   * Retrieves the contents of a vector store file.
+   * Retrieve vector store file contents (OpenAI-compatible).
    */
   content(
     vectorStoreId: string,
@@ -253,10 +253,6 @@ export namespace FileContentResponse {
      * the context during inference.
      */
     export interface ChunkMetadata {
-      chunk_embedding_dimension?: number | null;
-
-      chunk_embedding_model?: string | null;
-
       chunk_id?: string | null;
 
       chunk_tokenizer?: string | null;
@@ -279,12 +275,18 @@ export namespace FileContentResponse {
 }
 
 export interface FileCreateParams {
+  /**
+   * The ID of the file to attach.
+   */
   file_id: string;
 
+  /**
+   * Attributes to associate with the file.
+   */
   attributes?: { [key: string]: unknown } | null;
 
   /**
-   * Automatic chunking strategy for vector store files.
+   * Strategy for chunking the file content.
    */
   chunking_strategy?:
     | FileCreateParams.VectorStoreChunkingStrategyAuto
@@ -325,20 +327,38 @@ export namespace FileCreateParams {
 }
 
 export interface FileUpdateParams {
+  /**
+   * The new attributes for the file.
+   */
   attributes: { [key: string]: unknown };
 }
 
 export interface FileListParams extends OpenAICursorPageParams {
+  /**
+   * Pagination cursor (before).
+   */
   before?: string | null;
 
+  /**
+   * Filter by file status.
+   */
   filter?: 'completed' | 'in_progress' | 'cancelled' | 'failed' | null;
 
+  /**
+   * Sort order by created_at: asc or desc.
+   */
   order?: string | null;
 }
 
 export interface FileContentParams {
+  /**
+   * Include embedding vectors.
+   */
   include_embeddings?: boolean | null;
 
+  /**
+   * Include chunk metadata.
+   */
   include_metadata?: boolean | null;
 }
 

package/src/resources/vector-stores/vector-stores.ts

@@ -36,23 +36,21 @@ export class VectorStores extends APIResource {
   fileBatches: FileBatchesAPI.FileBatches = new FileBatchesAPI.FileBatches(this._client);
 
   /**
-   * Creates a vector store.
-   *
-   * Generate an OpenAI-compatible vector store with the given parameters.
+   * Create a vector store (OpenAI-compatible).
    */
   create(body: VectorStoreCreateParams, options?: Core.RequestOptions): Core.APIPromise<VectorStore> {
     return this._client.post('/v1/vector_stores', { body, ...options });
   }
 
   /**
-   * Retrieves a vector store.
+   * Retrieve a vector store (OpenAI-compatible).
    */
   retrieve(vectorStoreId: string, options?: Core.RequestOptions): Core.APIPromise<VectorStore> {
     return this._client.get(`/v1/vector_stores/${vectorStoreId}`, options);
   }
 
   /**
-   * Updates a vector store.
+   * Update a vector store (OpenAI-compatible).
    */
   update(
     vectorStoreId: string,
@@ -63,7 +61,7 @@ export class VectorStores extends APIResource {
   }
 
   /**
-   * Returns a list of vector stores.
+   * List vector stores (OpenAI-compatible).
    */
   list(
     query?: VectorStoreListParams,
@@ -81,17 +79,14 @@ export class VectorStores extends APIResource {
   }
 
   /**
-   * Delete a vector store.
+   * Delete a vector store (OpenAI-compatible).
    */
   delete(vectorStoreId: string, options?: Core.RequestOptions): Core.APIPromise<VectorStoreDeleteResponse> {
     return this._client.delete(`/v1/vector_stores/${vectorStoreId}`, options);
   }
 
   /**
-   * Search for chunks in a vector store.
-   *
-   * Searches a vector store for relevant chunks based on a query and optional file
-   * attribute filters.
+   * Search a vector store (OpenAI-compatible).
    */
   search(
     vectorStoreId: string,
@@ -242,10 +237,6 @@ export namespace VectorStoreSearchResponse {
        * the context during inference.
        */
       export interface ChunkMetadata {
-        chunk_embedding_dimension?: number | null;
-
-        chunk_embedding_model?: string | null;
-
         chunk_id?: string | null;
 
         chunk_tokenizer?: string | null;
@@ -321,44 +312,127 @@ export namespace VectorStoreCreateParams {
 }
 
 export interface VectorStoreUpdateParams {
+  /**
+   * Expiration policy for the vector store.
+   */
   expires_after?: { [key: string]: unknown } | null;
 
+  /**
+   * Metadata to associate with the vector store.
+   */
   metadata?: { [key: string]: unknown } | null;
 
+  /**
+   * The new name for the vector store.
+   */
   name?: string | null;
 }
 
 export interface VectorStoreListParams extends OpenAICursorPageParams {
+  /**
+   * Pagination cursor (before).
+   */
   before?: string | null;
 
+  /**
+   * Sort order by created_at: asc or desc.
+   */
   order?: string | null;
 }
 
 export interface VectorStoreSearchParams {
+  /**
+   * The search query string or list of query strings.
+   */
   query: string | Array<string>;
 
+  /**
+   * Filters to apply to the search.
+   */
   filters?: { [key: string]: unknown } | null;
 
+  /**
+   * Maximum number of results to return.
+   */
   max_num_results?: number | null;
 
   /**
    * Options for ranking and filtering search results.
+   *
+   * This class configures how search results are ranked and filtered. You can use
+   * algorithm-based rerankers (weighted, RRF) or neural rerankers. Defaults from
+   * VectorStoresConfig are used when parameters are not provided.
+   *
+   * Examples: # Weighted ranker with custom alpha
+   * SearchRankingOptions(ranker="weighted", alpha=0.7)
+   *
+   *     # RRF ranker with custom impact factor
+   *     SearchRankingOptions(ranker="rrf", impact_factor=50.0)
+   *
+   *     # Use config defaults (just specify ranker type)
+   *     SearchRankingOptions(ranker="weighted")  # Uses alpha from VectorStoresConfig
+   *
+   *     # Score threshold filtering
+   *     SearchRankingOptions(ranker="weighted", score_threshold=0.5)
    */
   ranking_options?: VectorStoreSearchParams.RankingOptions | null;
 
+  /**
+   * Whether to rewrite the query for better results.
+   */
   rewrite_query?: boolean | null;
 
+  /**
+   * The search mode to use (e.g., 'vector', 'keyword').
+   */
   search_mode?: string | null;
 }
 
 export namespace VectorStoreSearchParams {
   /**
    * Options for ranking and filtering search results.
+   *
+   * This class configures how search results are ranked and filtered. You can use
+   * algorithm-based rerankers (weighted, RRF) or neural rerankers. Defaults from
+   * VectorStoresConfig are used when parameters are not provided.
+   *
+   * Examples: # Weighted ranker with custom alpha
+   * SearchRankingOptions(ranker="weighted", alpha=0.7)
+   *
+   *     # RRF ranker with custom impact factor
+   *     SearchRankingOptions(ranker="rrf", impact_factor=50.0)
+   *
+   *     # Use config defaults (just specify ranker type)
+   *     SearchRankingOptions(ranker="weighted")  # Uses alpha from VectorStoresConfig
+   *
+   *     # Score threshold filtering
+   *     SearchRankingOptions(ranker="weighted", score_threshold=0.5)
    */
   export interface RankingOptions {
+    /**
+     * Weight factor for weighted ranker
+     */
+    alpha?: number | null;
+
+    /**
+     * Impact factor for RRF algorithm
+     */
+    impact_factor?: number | null;
+
+    /**
+     * Model identifier for neural reranker
+     */
+    model?: string | null;
+
     ranker?: string | null;
 
     score_threshold?: number | null;
+
+    /**
+     * Weights for combining vector, keyword, and neural scores. Keys: 'vector',
+     * 'keyword', 'neural'
+     */
+    weights?: { [key: string]: number } | null;
   }
 }
 

package/src/version.ts

@@ -4,4 +4,4 @@
 // This source code is licensed under the terms described in the LICENSE file in
 // the root directory of this source tree.
 
-export const VERSION = '0.4.4'; // x-release-please-version
+export const VERSION = '0.4.5'; // x-release-please-version

package/tests/api-resources/alpha/admin.test.ts

@@ -0,0 +1,111 @@
+// Copyright (c) Meta Platforms, Inc. and affiliates.
+// All rights reserved.
+//
+// This source code is licensed under the terms described in the LICENSE file in
+// the root directory of this source tree.
+//
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import LlamaStackClient from 'llama-stack-client';
+import { Response } from 'node-fetch';
+
+const client = new LlamaStackClient({ baseURL: process.env['TEST_API_BASE_URL'] ?? 'http://127.0.0.1:4010' });
+
+describe('resource admin', () => {
+  test('health', async () => {
+    const responsePromise = client.alpha.admin.health();
+    const rawResponse = await responsePromise.asResponse();
+    expect(rawResponse).toBeInstanceOf(Response);
+    const response = await responsePromise;
+    expect(response).not.toBeInstanceOf(Response);
+    const dataAndResponse = await responsePromise.withResponse();
+    expect(dataAndResponse.data).toBe(response);
+    expect(dataAndResponse.response).toBe(rawResponse);
+  });
+
+  test('health: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.admin.health({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
+  });
+
+  test('inspectProvider', async () => {
+    const responsePromise = client.alpha.admin.inspectProvider('provider_id');
+    const rawResponse = await responsePromise.asResponse();
+    expect(rawResponse).toBeInstanceOf(Response);
+    const response = await responsePromise;
+    expect(response).not.toBeInstanceOf(Response);
+    const dataAndResponse = await responsePromise.withResponse();
+    expect(dataAndResponse.data).toBe(response);
+    expect(dataAndResponse.response).toBe(rawResponse);
+  });
+
+  test('inspectProvider: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(
+      client.alpha.admin.inspectProvider('provider_id', { path: '/_stainless_unknown_path' }),
+    ).rejects.toThrow(LlamaStackClient.NotFoundError);
+  });
+
+  test('listProviders', async () => {
+    const responsePromise = client.alpha.admin.listProviders();
+    const rawResponse = await responsePromise.asResponse();
+    expect(rawResponse).toBeInstanceOf(Response);
+    const response = await responsePromise;
+    expect(response).not.toBeInstanceOf(Response);
+    const dataAndResponse = await responsePromise.withResponse();
+    expect(dataAndResponse.data).toBe(response);
+    expect(dataAndResponse.response).toBe(rawResponse);
+  });
+
+  test('listProviders: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.admin.listProviders({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
+  });
+
+  test('listRoutes', async () => {
+    const responsePromise = client.alpha.admin.listRoutes();
+    const rawResponse = await responsePromise.asResponse();
+    expect(rawResponse).toBeInstanceOf(Response);
+    const response = await responsePromise;
+    expect(response).not.toBeInstanceOf(Response);
+    const dataAndResponse = await responsePromise.withResponse();
+    expect(dataAndResponse.data).toBe(response);
+    expect(dataAndResponse.response).toBe(rawResponse);
+  });
+
+  test('listRoutes: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.admin.listRoutes({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
+  });
+
+  test('listRoutes: request options and params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(
+      client.alpha.admin.listRoutes({ api_filter: 'v1' }, { path: '/_stainless_unknown_path' }),
+    ).rejects.toThrow(LlamaStackClient.NotFoundError);
+  });
+
+  test('version', async () => {
+    const responsePromise = client.alpha.admin.version();
+    const rawResponse = await responsePromise.asResponse();
+    expect(rawResponse).toBeInstanceOf(Response);
+    const response = await responsePromise;
+    expect(response).not.toBeInstanceOf(Response);
+    const dataAndResponse = await responsePromise.withResponse();
+    expect(dataAndResponse.data).toBe(response);
+    expect(dataAndResponse.response).toBe(rawResponse);
+  });
+
+  test('version: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.admin.version({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
+  });
+});

package/tests/api-resources/alpha/eval/eval.test.ts

@@ -14,7 +14,12 @@ const client = new LlamaStackClient({ baseURL: process.env['TEST_API_BASE_URL']
 describe('resource eval', () => {
   test('evaluateRows: only required params', async () => {
     const responsePromise = client.alpha.eval.evaluateRows('benchmark_id', {
-      benchmark_config: { eval_candidate: { model: 'model', sampling_params: {} } },
+      benchmark_config: {
+        eval_candidate: {
+          model: 'x',
+          sampling_params: {},
+        },
+      },
       input_rows: [{ foo: 'bar' }],
       scoring_functions: ['string'],
     });
@@ -31,17 +36,17 @@ describe('resource eval', () => {
     const response = await client.alpha.eval.evaluateRows('benchmark_id', {
       benchmark_config: {
         eval_candidate: {
-          model: 'model',
+          model: 'x',
           sampling_params: {
-            max_tokens: 0,
-            repetition_penalty: 0,
+            max_tokens: 1,
+            repetition_penalty: -2,
             stop: ['string'],
             strategy: { type: 'greedy' },
           },
           system_message: { content: 'string', role: 'system' },
           type: 'model',
         },
-        num_examples: 0,
+        num_examples: 1,
         scoring_params: {
           foo: {
             judge_model: 'judge_model',
@@ -59,7 +64,12 @@ describe('resource eval', () => {
 
   test('evaluateRowsAlpha: only required params', async () => {
     const responsePromise = client.alpha.eval.evaluateRowsAlpha('benchmark_id', {
-      benchmark_config: { eval_candidate: { model: 'model', sampling_params: {} } },
+      benchmark_config: {
+        eval_candidate: {
+          model: 'x',
+          sampling_params: {},
+        },
+      },
       input_rows: [{ foo: 'bar' }],
       scoring_functions: ['string'],
     });
@@ -76,17 +86,17 @@ describe('resource eval', () => {
     const response = await client.alpha.eval.evaluateRowsAlpha('benchmark_id', {
       benchmark_config: {
         eval_candidate: {
-          model: 'model',
+          model: 'x',
           sampling_params: {
-            max_tokens: 0,
-            repetition_penalty: 0,
+            max_tokens: 1,
+            repetition_penalty: -2,
             stop: ['string'],
             strategy: { type: 'greedy' },
           },
           system_message: { content: 'string', role: 'system' },
           type: 'model',
         },
-        num_examples: 0,
+        num_examples: 1,
         scoring_params: {
           foo: {
             judge_model: 'judge_model',
@@ -104,7 +114,12 @@ describe('resource eval', () => {
 
   test('runEval: only required params', async () => {
     const responsePromise = client.alpha.eval.runEval('benchmark_id', {
-      benchmark_config: { eval_candidate: { model: 'model', sampling_params: {} } },
+      benchmark_config: {
+        eval_candidate: {
+          model: 'x',
+          sampling_params: {},
+        },
+      },
     });
     const rawResponse = await responsePromise.asResponse();
     expect(rawResponse).toBeInstanceOf(Response);
@@ -119,17 +134,17 @@ describe('resource eval', () => {
     const response = await client.alpha.eval.runEval('benchmark_id', {
       benchmark_config: {
         eval_candidate: {
-          model: 'model',
+          model: 'x',
           sampling_params: {
-            max_tokens: 0,
-            repetition_penalty: 0,
+            max_tokens: 1,
+            repetition_penalty: -2,
             stop: ['string'],
             strategy: { type: 'greedy' },
           },
           system_message: { content: 'string', role: 'system' },
           type: 'model',
         },
-        num_examples: 0,
+        num_examples: 1,
         scoring_params: {
           foo: {
             judge_model: 'judge_model',
@@ -145,7 +160,12 @@ describe('resource eval', () => {
 
   test('runEvalAlpha: only required params', async () => {
     const responsePromise = client.alpha.eval.runEvalAlpha('benchmark_id', {
-      benchmark_config: { eval_candidate: { model: 'model', sampling_params: {} } },
+      benchmark_config: {
+        eval_candidate: {
+          model: 'x',
+          sampling_params: {},
+        },
+      },
     });
     const rawResponse = await responsePromise.asResponse();
     expect(rawResponse).toBeInstanceOf(Response);
@@ -160,17 +180,17 @@ describe('resource eval', () => {
     const response = await client.alpha.eval.runEvalAlpha('benchmark_id', {
       benchmark_config: {
         eval_candidate: {
-          model: 'model',
+          model: 'x',
           sampling_params: {
-            max_tokens: 0,
-            repetition_penalty: 0,
+            max_tokens: 1,
+            repetition_penalty: -2,
             stop: ['string'],
             strategy: { type: 'greedy' },
           },
           system_message: { content: 'string', role: 'system' },
           type: 'model',
         },
-        num_examples: 0,
+        num_examples: 1,
         scoring_params: {
           foo: {
             judge_model: 'judge_model',

package/tests/api-resources/alpha/inference.test.ts

@@ -32,7 +32,7 @@ describe('resource inference', () => {
       items: ['string'],
       model: 'model',
       query: 'string',
-      max_num_results: 0,
+      max_num_results: 1,
     });
   });
 });

package/tests/api-resources/alpha/post-training/job.test.ts

@@ -30,8 +30,8 @@ describe('resource job', () => {
     );
   });
 
-  test('artifacts: only required params', async () => {
-    const responsePromise = client.alpha.postTraining.job.artifacts({ job_uuid: 'job_uuid' });
+  test('artifacts', async () => {
+    const responsePromise = client.alpha.postTraining.job.artifacts();
     const rawResponse = await responsePromise.asResponse();
     expect(rawResponse).toBeInstanceOf(Response);
     const response = await responsePromise;
@@ -41,12 +41,15 @@ describe('resource job', () => {
     expect(dataAndResponse.response).toBe(rawResponse);
   });
 
-  test('artifacts: required and optional params', async () => {
-    const response = await client.alpha.postTraining.job.artifacts({ job_uuid: 'job_uuid' });
+  test('artifacts: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(
+      client.alpha.postTraining.job.artifacts({ path: '/_stainless_unknown_path' }),
+    ).rejects.toThrow(LlamaStackClient.NotFoundError);
   });
 
-  test('cancel: only required params', async () => {
-    const responsePromise = client.alpha.postTraining.job.cancel({ job_uuid: 'job_uuid' });
+  test('cancel', async () => {
+    const responsePromise = client.alpha.postTraining.job.cancel();
     const rawResponse = await responsePromise.asResponse();
     expect(rawResponse).toBeInstanceOf(Response);
     const response = await responsePromise;
@@ -56,12 +59,15 @@ describe('resource job', () => {
     expect(dataAndResponse.response).toBe(rawResponse);
   });
 
-  test('cancel: required and optional params', async () => {
-    const response = await client.alpha.postTraining.job.cancel({ job_uuid: 'job_uuid' });
+  test('cancel: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.postTraining.job.cancel({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
   });
 
-  test('status: only required params', async () => {
-    const responsePromise = client.alpha.postTraining.job.status({ job_uuid: 'job_uuid' });
+  test('status', async () => {
+    const responsePromise = client.alpha.postTraining.job.status();
     const rawResponse = await responsePromise.asResponse();
     expect(rawResponse).toBeInstanceOf(Response);
     const response = await responsePromise;
@@ -71,7 +77,10 @@ describe('resource job', () => {
     expect(dataAndResponse.response).toBe(rawResponse);
   });
 
-  test('status: required and optional params', async () => {
-    const response = await client.alpha.postTraining.job.status({ job_uuid: 'job_uuid' });
+  test('status: request options instead of params are passed correctly', async () => {
+    // ensure the request options are being passed correctly by passing an invalid HTTP method in order to cause an error
+    await expect(client.alpha.postTraining.job.status({ path: '/_stainless_unknown_path' })).rejects.toThrow(
+      LlamaStackClient.NotFoundError,
+    );
   });
 });

package/tests/api-resources/alpha/post-training/post-training.test.ts

@@ -58,7 +58,12 @@ describe('resource postTraining', () => {
         gradient_accumulation_steps: 0,
         max_steps_per_epoch: 0,
         max_validation_steps: 0,
-        optimizer_config: { lr: 0, num_warmup_steps: 0, optimizer_type: 'adam', weight_decay: 0 },
+        optimizer_config: {
+          lr: 0,
+          num_warmup_steps: 0,
+          optimizer_type: 'adam',
+          weight_decay: 0,
+        },
       },
     });
   });
@@ -105,7 +110,12 @@ describe('resource postTraining', () => {
         gradient_accumulation_steps: 0,
         max_steps_per_epoch: 0,
         max_validation_steps: 0,
-        optimizer_config: { lr: 0, num_warmup_steps: 0, optimizer_type: 'adam', weight_decay: 0 },
+        optimizer_config: {
+          lr: 0,
+          num_warmup_steps: 0,
+          optimizer_type: 'adam',
+          weight_decay: 0,
+        },
       },
       algorithm_config: {
         alpha: 0,