llama-stack-client 0.2.0 → 0.2.3-rc5

package/src/_shims/index.mjs

@@ -3,5 +3,9 @@
  */
 import * as shims from './registry.mjs';
 import * as auto from "./auto/runtime";
-if (!shims.kind) shims.setShims(auto.getRuntime(), { auto: true });
+export const init = () => {
+  if (!shims.kind) shims.setShims(auto.getRuntime(), { auto: true });
+};
 export * from './registry.mjs';
+
+init();

package/src/core.ts

@@ -17,7 +17,12 @@ import {
   type RequestInit,
   type Response,
   type HeadersInit,
+  init,
 } from './_shims/index';
+
+// try running side effects outside of _shims/index to workaround https://github.com/vercel/next.js/issues/76881
+init();
+
 export { type Response };
 import { BlobLike, isBlobLike, isMultipartBody } from './uploads';
 export {
@@ -29,6 +34,20 @@ export {
 
 export type Fetch = (url: RequestInfo, init?: RequestInit) => Promise<Response>;
 
+/**
+ * An alias to the builtin `Array` type so we can
+ * easily alias it in import statements if there are name clashes.
+ */
+type _Array<T> = Array<T>;
+
+/**
+ * An alias to the builtin `Record` type so we can
+ * easily alias it in import statements if there are name clashes.
+ */
+type _Record<K extends keyof any, T> = Record<K, T>;
+
+export type { _Array as Array, _Record as Record };
+
 type PromiseOrValue<T> = T | Promise<T>;
 
 type APIResponseProps = {
@@ -291,10 +310,10 @@ export abstract class APIClient {
   }
 
   buildRequest<Req>(
-    options: FinalRequestOptions<Req>,
+    inputOptions: FinalRequestOptions<Req>,
     { retryCount = 0 }: { retryCount?: number } = {},
   ): { req: RequestInit; url: string; timeout: number } {
-    options = { ...options };
+    const options = { ...inputOptions };
     const { method, path, query, headers: headers = {} } = options;
 
     const body =
@@ -322,8 +341,8 @@ export abstract class APIClient {
     }
 
     if (this.idempotencyHeader && method !== 'get') {
-      if (!options.idempotencyKey) options.idempotencyKey = this.defaultIdempotencyKey();
-      headers[this.idempotencyHeader] = options.idempotencyKey;
+      if (!inputOptions.idempotencyKey) inputOptions.idempotencyKey = this.defaultIdempotencyKey();
+      headers[this.idempotencyHeader] = inputOptions.idempotencyKey;
     }
 
     const reqHeaders = this.buildHeaders({ options, headers, contentLength, retryCount });
@@ -380,7 +399,7 @@ export abstract class APIClient {
       getHeader(headers, 'x-stainless-timeout') === undefined &&
       options.timeout
     ) {
-      reqHeaders['x-stainless-timeout'] = String(options.timeout);
+      reqHeaders['x-stainless-timeout'] = String(Math.trunc(options.timeout / 1000));
     }
 
     this.validateHeaders(reqHeaders, headers);

package/src/index.ts

@@ -8,12 +8,6 @@ import * as Pagination from './pagination';
 import { type DatasetsIterrowsParams, DatasetsIterrowsResponse } from './pagination';
 import * as Uploads from './uploads';
 import * as API from './resources/index';
-import {
-  BatchInference,
-  BatchInferenceChatCompletionParams,
-  BatchInferenceChatCompletionResponse,
-  BatchInferenceCompletionParams,
-} from './resources/batch-inference';
 import {
   Benchmark,
   BenchmarkListResponse,
@@ -21,6 +15,13 @@ import {
   Benchmarks,
   ListBenchmarksResponse,
 } from './resources/benchmarks';
+import {
+  CompletionCreateParams,
+  CompletionCreateParamsNonStreaming,
+  CompletionCreateParamsStreaming,
+  CompletionCreateResponse,
+  Completions,
+} from './resources/completions';
 import {
   DatasetIterrowsParams,
   DatasetIterrowsResponse,
@@ -36,6 +37,9 @@ import {
   CompletionResponse,
   EmbeddingsResponse,
   Inference,
+  InferenceBatchChatCompletionParams,
+  InferenceBatchChatCompletionResponse,
+  InferenceBatchCompletionParams,
   InferenceChatCompletionParams,
   InferenceChatCompletionParamsNonStreaming,
   InferenceChatCompletionParamsStreaming,
@@ -132,6 +136,7 @@ import {
   ToolExecutionStep,
   ToolResponse,
 } from './resources/agents/agents';
+import { Chat, ChatCompletionChunk } from './resources/chat/chat';
 import {
   BenchmarkConfig,
   Eval,
@@ -157,6 +162,7 @@ import {
   ToolRuntime,
   ToolRuntimeInvokeToolParams,
   ToolRuntimeListToolsParams,
+  ToolRuntimeListToolsResponse,
 } from './resources/tool-runtime/tool-runtime';
 
 export interface ClientOptions {
@@ -270,11 +276,12 @@ export class LlamaStackClient extends Core.APIClient {
   tools: API.Tools = new API.Tools(this);
   toolRuntime: API.ToolRuntime = new API.ToolRuntime(this);
   agents: API.Agents = new API.Agents(this);
-  batchInference: API.BatchInference = new API.BatchInference(this);
   datasets: API.Datasets = new API.Datasets(this);
   eval: API.Eval = new API.Eval(this);
   inspect: API.Inspect = new API.Inspect(this);
   inference: API.Inference = new API.Inference(this);
+  chat: API.Chat = new API.Chat(this);
+  completions: API.Completions = new API.Completions(this);
   vectorIo: API.VectorIo = new API.VectorIo(this);
   vectorDBs: API.VectorDBs = new API.VectorDBs(this);
   models: API.Models = new API.Models(this);
@@ -336,11 +343,12 @@ LlamaStackClient.Toolgroups = Toolgroups;
 LlamaStackClient.Tools = Tools;
 LlamaStackClient.ToolRuntime = ToolRuntime;
 LlamaStackClient.Agents = Agents;
-LlamaStackClient.BatchInference = BatchInference;
 LlamaStackClient.Datasets = Datasets;
 LlamaStackClient.Eval = Eval;
 LlamaStackClient.Inspect = Inspect;
 LlamaStackClient.Inference = Inference;
+LlamaStackClient.Chat = Chat;
+LlamaStackClient.Completions = Completions;
 LlamaStackClient.VectorIo = VectorIo;
 LlamaStackClient.VectorDBs = VectorDBs;
 LlamaStackClient.Models = Models;
@@ -383,6 +391,7 @@ export declare namespace LlamaStackClient {
     ToolRuntime as ToolRuntime,
     type ToolDef as ToolDef,
     type ToolInvocationResult as ToolInvocationResult,
+    type ToolRuntimeListToolsResponse as ToolRuntimeListToolsResponse,
     type ToolRuntimeInvokeToolParams as ToolRuntimeInvokeToolParams,
     type ToolRuntimeListToolsParams as ToolRuntimeListToolsParams,
   };
@@ -398,13 +407,6 @@ export declare namespace LlamaStackClient {
     type AgentCreateParams as AgentCreateParams,
   };
 
-  export {
-    BatchInference as BatchInference,
-    type BatchInferenceChatCompletionResponse as BatchInferenceChatCompletionResponse,
-    type BatchInferenceChatCompletionParams as BatchInferenceChatCompletionParams,
-    type BatchInferenceCompletionParams as BatchInferenceCompletionParams,
-  };
-
   export {
     Datasets as Datasets,
     type ListDatasetsResponse as ListDatasetsResponse,
@@ -442,6 +444,9 @@ export declare namespace LlamaStackClient {
     type CompletionResponse as CompletionResponse,
     type EmbeddingsResponse as EmbeddingsResponse,
     type TokenLogProbs as TokenLogProbs,
+    type InferenceBatchChatCompletionResponse as InferenceBatchChatCompletionResponse,
+    type InferenceBatchChatCompletionParams as InferenceBatchChatCompletionParams,
+    type InferenceBatchCompletionParams as InferenceBatchCompletionParams,
     type InferenceChatCompletionParams as InferenceChatCompletionParams,
     type InferenceChatCompletionParamsNonStreaming as InferenceChatCompletionParamsNonStreaming,
     type InferenceChatCompletionParamsStreaming as InferenceChatCompletionParamsStreaming,
@@ -451,6 +456,16 @@ export declare namespace LlamaStackClient {
     type InferenceEmbeddingsParams as InferenceEmbeddingsParams,
   };
 
+  export { Chat as Chat, type ChatCompletionChunk as ChatCompletionChunk };
+
+  export {
+    Completions as Completions,
+    type CompletionCreateResponse as CompletionCreateResponse,
+    type CompletionCreateParams as CompletionCreateParams,
+    type CompletionCreateParamsNonStreaming as CompletionCreateParamsNonStreaming,
+    type CompletionCreateParamsStreaming as CompletionCreateParamsStreaming,
+  };
+
   export {
     VectorIo as VectorIo,
     type QueryChunksResponse as QueryChunksResponse,

package/src/resources/chat/chat.ts

@@ -0,0 +1,206 @@
+// File generated from our OpenAPI spec by Stainless. See CONTRIBUTING.md for details.
+
+import { APIResource } from '../../resource';
+import * as CompletionsAPI from './completions';
+import {
+  CompletionCreateParams,
+  CompletionCreateParamsNonStreaming,
+  CompletionCreateParamsStreaming,
+  CompletionCreateResponse,
+  Completions,
+} from './completions';
+
+export class Chat extends APIResource {
+  completions: CompletionsAPI.Completions = new CompletionsAPI.Completions(this._client);
+}
+
+/**
+ * Chunk from a streaming response to an OpenAI-compatible chat completion request.
+ */
+export interface ChatCompletionChunk {
+  /**
+   * The ID of the chat completion
+   */
+  id: string;
+
+  /**
+   * List of choices
+   */
+  choices: Array<ChatCompletionChunk.Choice>;
+
+  /**
+   * The Unix timestamp in seconds when the chat completion was created
+   */
+  created: number;
+
+  /**
+   * The model that was used to generate the chat completion
+   */
+  model: string;
+
+  /**
+   * The object type, which will be "chat.completion.chunk"
+   */
+  object: 'chat.completion.chunk';
+}
+
+export namespace ChatCompletionChunk {
+  /**
+   * A chunk choice from an OpenAI-compatible chat completion streaming response.
+   */
+  export interface Choice {
+    /**
+     * The delta from the chunk
+     */
+    delta: Choice.Delta;
+
+    /**
+     * The reason the model stopped generating
+     */
+    finish_reason: string;
+
+    /**
+     * The index of the choice
+     */
+    index: number;
+
+    /**
+     * (Optional) The log probabilities for the tokens in the message
+     */
+    logprobs?: Choice.Logprobs;
+  }
+
+  export namespace Choice {
+    /**
+     * The delta from the chunk
+     */
+    export interface Delta {
+      /**
+       * (Optional) The content of the delta
+       */
+      content?: string;
+
+      /**
+       * (Optional) The refusal of the delta
+       */
+      refusal?: string;
+
+      /**
+       * (Optional) The role of the delta
+       */
+      role?: string;
+
+      /**
+       * (Optional) The tool calls of the delta
+       */
+      tool_calls?: Array<Delta.ToolCall>;
+    }
+
+    export namespace Delta {
+      export interface ToolCall {
+        type: 'function';
+
+        id?: string;
+
+        function?: ToolCall.Function;
+
+        index?: number;
+      }
+
+      export namespace ToolCall {
+        export interface Function {
+          arguments?: string;
+
+          name?: string;
+        }
+      }
+    }
+
+    /**
+     * (Optional) The log probabilities for the tokens in the message
+     */
+    export interface Logprobs {
+      /**
+       * (Optional) The log probabilities for the tokens in the message
+       */
+      content?: Array<Logprobs.Content>;
+
+      /**
+       * (Optional) The log probabilities for the tokens in the message
+       */
+      refusal?: Array<Logprobs.Refusal>;
+    }
+
+    export namespace Logprobs {
+      /**
+       * The log probability for a token from an OpenAI-compatible chat completion
+       * response.
+       */
+      export interface Content {
+        token: string;
+
+        logprob: number;
+
+        top_logprobs: Array<Content.TopLogprob>;
+
+        bytes?: Array<number>;
+      }
+
+      export namespace Content {
+        /**
+         * The top log probability for a token from an OpenAI-compatible chat completion
+         * response.
+         */
+        export interface TopLogprob {
+          token: string;
+
+          logprob: number;
+
+          bytes?: Array<number>;
+        }
+      }
+
+      /**
+       * The log probability for a token from an OpenAI-compatible chat completion
+       * response.
+       */
+      export interface Refusal {
+        token: string;
+
+        logprob: number;
+
+        top_logprobs: Array<Refusal.TopLogprob>;
+
+        bytes?: Array<number>;
+      }
+
+      export namespace Refusal {
+        /**
+         * The top log probability for a token from an OpenAI-compatible chat completion
+         * response.
+         */
+        export interface TopLogprob {
+          token: string;
+
+          logprob: number;
+
+          bytes?: Array<number>;
+        }
+      }
+    }
+  }
+}
+
+Chat.Completions = Completions;
+
+export declare namespace Chat {
+  export { type ChatCompletionChunk as ChatCompletionChunk };
+
+  export {
+    Completions as Completions,
+    type CompletionCreateResponse as CompletionCreateResponse,
+    type CompletionCreateParams as CompletionCreateParams,
+    type CompletionCreateParamsNonStreaming as CompletionCreateParamsNonStreaming,
+    type CompletionCreateParamsStreaming as CompletionCreateParamsStreaming,
+  };
+}