modelfusion 0.98.0 → 0.100.0

package/model-provider/openai/chat/OpenAIChatModel.cjs

@@ -144,12 +144,12 @@ exports.calculateOpenAIChatCostInMillicents = calculateOpenAIChatCostInMillicent
  * const model = new OpenAIChatModel({
  *   model: "gpt-3.5-turbo",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);
@@ -196,13 +196,13 @@ class OpenAIChatModel extends AbstractOpenAIChatModel_js_1.AbstractOpenAIChatMod
     }
     get settingsForEvent() {
         const eventSettingProperties = [
+            "maxGenerationTokens",
             "stopSequences",
-            "maxCompletionTokens",
+            "numberOfGenerations",
             "functions",
             "functionCall",
             "temperature",
             "topP",
-            "n",
             "presencePenalty",
             "frequencyPenalty",
             "logitBias",

package/model-provider/openai/chat/OpenAIChatModel.d.ts

@@ -117,12 +117,12 @@ export interface OpenAIChatSettings extends TextGenerationModelSettings, Omit<Op
  * const model = new OpenAIChatModel({
  *   model: "gpt-3.5-turbo",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);
@@ -142,7 +142,7 @@ export declare class OpenAIChatModel extends AbstractOpenAIChatModel<OpenAIChatS
     asFunctionCallStructureGenerationModel({ fnName, fnDescription, }: {
         fnName: string;
         fnDescription?: string;
-    }): OpenAIChatFunctionCallStructureGenerationModel<TextGenerationPromptTemplate<import("./OpenAIChatMessage.js").OpenAIChatMessage[], import("./OpenAIChatMessage.js").OpenAIChatMessage[]>>;
+    }): OpenAIChatFunctionCallStructureGenerationModel<TextGenerationPromptTemplate<OpenAIChatPrompt, OpenAIChatPrompt>>;
     asStructureGenerationModel<INPUT_PROMPT>(promptTemplate: StructureFromTextPromptTemplate<INPUT_PROMPT, OpenAIChatPrompt>): StructureFromTextStreamingModel<INPUT_PROMPT, OpenAIChatPrompt, this>;
     /**
      * Returns this model with a text prompt template.

package/model-provider/openai/chat/OpenAIChatModel.js

@@ -138,12 +138,12 @@ export const calculateOpenAIChatCostInMillicents = ({ model, response, }) => {
  * const model = new OpenAIChatModel({
  *   model: "gpt-3.5-turbo",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);
@@ -190,13 +190,13 @@ export class OpenAIChatModel extends AbstractOpenAIChatModel {
     }
     get settingsForEvent() {
         const eventSettingProperties = [
+            "maxGenerationTokens",
             "stopSequences",
-            "maxCompletionTokens",
+            "numberOfGenerations",
             "functions",
             "functionCall",
             "temperature",
             "topP",
-            "n",
             "presencePenalty",
             "frequencyPenalty",
             "logitBias",

package/model-provider/openai/chat/OpenAIChatModel.test.cjs

@@ -50,7 +50,7 @@ describe("streamText", () => {
         const stream = await (0, streamText_js_1.streamText)(new OpenAIChatModel_js_1.OpenAIChatModel({
             api: new OpenAIApiConfiguration_js_1.OpenAIApiConfiguration({ apiKey: "test" }),
             model: "gpt-3.5-turbo",
-            n: 2,
+            numberOfGenerations: 2,
         }).withTextPrompt(), "test prompt");
         const chunks = [];
         for await (const part of stream) {

package/model-provider/openai/chat/OpenAIChatModel.test.js

@@ -48,7 +48,7 @@ describe("streamText", () => {
         const stream = await streamText(new OpenAIChatModel({
             api: new OpenAIApiConfiguration({ apiKey: "test" }),
             model: "gpt-3.5-turbo",
-            n: 2,
+            numberOfGenerations: 2,
         }).withTextPrompt(), "test prompt");
         const chunks = [];
         for await (const part of stream) {

package/model-provider/openai/chat/OpenAIChatPromptTemplate.d.ts

@@ -1,20 +1,20 @@
 import { TextGenerationPromptTemplate } from "../../../model-function/generate-text/TextGenerationPromptTemplate.js";
 import { MultiModalChatPrompt, TextChatPrompt } from "../../../model-function/generate-text/prompt-template/ChatPrompt.js";
 import { MultiModalInstructionPrompt, TextInstructionPrompt } from "../../../model-function/generate-text/prompt-template/InstructionPrompt.js";
-import { OpenAIChatMessage } from "./OpenAIChatMessage.js";
+import { OpenAIChatPrompt } from "./AbstractOpenAIChatModel.js";
 /**
  * OpenAIMessage[] identity chat format.
  */
-export declare function identity(): TextGenerationPromptTemplate<Array<OpenAIChatMessage>, Array<OpenAIChatMessage>>;
+export declare function identity(): TextGenerationPromptTemplate<OpenAIChatPrompt, OpenAIChatPrompt>;
 /**
  * Formats a text prompt as an OpenAI chat prompt.
  */
-export declare function text(): TextGenerationPromptTemplate<string, Array<OpenAIChatMessage>>;
+export declare function text(): TextGenerationPromptTemplate<string, OpenAIChatPrompt>;
 /**
  * Formats an instruction prompt as an OpenAI chat prompt.
  */
-export declare function instruction(): TextGenerationPromptTemplate<MultiModalInstructionPrompt | TextInstructionPrompt, Array<OpenAIChatMessage>>;
+export declare function instruction(): TextGenerationPromptTemplate<MultiModalInstructionPrompt | TextInstructionPrompt, OpenAIChatPrompt>;
 /**
  * Formats a chat prompt as an OpenAI chat prompt.
  */
-export declare function chat(): TextGenerationPromptTemplate<MultiModalChatPrompt | TextChatPrompt, Array<OpenAIChatMessage>>;
+export declare function chat(): TextGenerationPromptTemplate<MultiModalChatPrompt | TextChatPrompt, OpenAIChatPrompt>;

package/model-provider/openai/chat/OpenAIChatPromptTemplate.js

@@ -1,4 +1,4 @@
-import { validateChatPrompt } from "../../../model-function/generate-text/prompt-template/ChatPrompt.js";
+import { validateChatPrompt, } from "../../../model-function/generate-text/prompt-template/ChatPrompt.js";
 import { OpenAIChatMessage } from "./OpenAIChatMessage.js";
 /**
  * OpenAIMessage[] identity chat format.

package/model-provider/openai-compatible/OpenAICompatibleChatModel.cjs

@@ -44,12 +44,12 @@ class OpenAICompatibleChatModel extends AbstractOpenAIChatModel_js_1.AbstractOpe
     get settingsForEvent() {
         const eventSettingProperties = [
             "stopSequences",
-            "maxCompletionTokens",
+            "maxGenerationTokens",
+            "numberOfGenerations",
             "functions",
             "functionCall",
             "temperature",
             "topP",
-            "n",
             "presencePenalty",
             "frequencyPenalty",
             "logitBias",

package/model-provider/openai-compatible/OpenAICompatibleChatModel.js

@@ -41,12 +41,12 @@ export class OpenAICompatibleChatModel extends AbstractOpenAIChatModel {
     get settingsForEvent() {
         const eventSettingProperties = [
             "stopSequences",
-            "maxCompletionTokens",
+            "maxGenerationTokens",
+            "numberOfGenerations",
             "functions",
             "functionCall",
             "temperature",
             "topP",
-            "n",
             "presencePenalty",
             "frequencyPenalty",
             "logitBias",

package/model-provider/openai-compatible/OpenAICompatibleFacade.cjs

@@ -14,12 +14,12 @@ const OpenAICompatibleChatModel_js_1 = require("./OpenAICompatibleChatModel.cjs"
  * const model = openaicompatible.ChatTextGenerator({
  *   model: "provider-specific-model-name",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);

package/model-provider/openai-compatible/OpenAICompatibleFacade.d.ts

@@ -11,12 +11,12 @@ import { OpenAICompatibleChatModel, OpenAICompatibleChatSettings } from "./OpenA
  * const model = openaicompatible.ChatTextGenerator({
  *   model: "provider-specific-model-name",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);

package/model-provider/openai-compatible/OpenAICompatibleFacade.js

@@ -11,12 +11,12 @@ import { OpenAICompatibleChatModel, } from "./OpenAICompatibleChatModel.js";
  * const model = openaicompatible.ChatTextGenerator({
  *   model: "provider-specific-model-name",
  *   temperature: 0.7,
- *   maxCompletionTokens: 500,
+ *   maxGenerationTokens: 500,
  * });
  *
  * const text = await generateText([
  *   model,
- *   OpenAIChatMessage.system(
+ *   openai.ChatMessage.system(
  *     "Write a short story about a robot learning to love:"
  *   ),
  * ]);

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "modelfusion",
   "description": "The TypeScript library for building multi-modal AI applications.",
-  "version": "0.98.0",
+  "version": "0.100.0",
   "author": "Lars Grammel",
   "license": "MIT",
   "keywords": [

package/model-provider/ollama/OllamaTextGenerationModel.d.ts

@@ -1,230 +0,0 @@
-import { z } from "zod";
-import { FunctionOptions } from "../../core/FunctionOptions.js";
-import { ApiConfiguration } from "../../core/api/ApiConfiguration.js";
-import { ResponseHandler } from "../../core/api/postToApi.js";
-import { AbstractModel } from "../../model-function/AbstractModel.js";
-import { Delta } from "../../model-function/Delta.js";
-import { PromptTemplateTextStreamingModel } from "../../model-function/generate-text/PromptTemplateTextStreamingModel.js";
-import { TextGenerationModelSettings, TextStreamingModel } from "../../model-function/generate-text/TextGenerationModel.js";
-import { TextGenerationPromptTemplate } from "../../model-function/generate-text/TextGenerationPromptTemplate.js";
-import { TextGenerationToolCallModel, ToolCallPromptTemplate } from "../../tool/generate-tool-call/TextGenerationToolCallModel.js";
-import { TextGenerationToolCallsOrGenerateTextModel, ToolCallsOrGenerateTextPromptTemplate } from "../../tool/generate-tool-calls-or-text/TextGenerationToolCallsOrGenerateTextModel.js";
-/**
- * @see https://github.com/jmorganca/ollama/blob/main/docs/api.md#generate-a-completion
- */
-export interface OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE extends number | undefined> extends TextGenerationModelSettings {
-    api?: ApiConfiguration;
-    /**
-     * The name of the model to use. For example, 'mistral'.
-     *
-     * @see https://ollama.ai/library
-     */
-    model: string;
-    /**
-     * The temperature of the model. Increasing the temperature will make the model
-     * answer more creatively. (Default: 0.8)
-     */
-    temperature?: number;
-    /**
-     * Specify the context window size of the model that you have loaded in your
-     * Ollama server. (Default: 2048)
-     */
-    contextWindowSize?: CONTEXT_WINDOW_SIZE;
-    /**
-     * Enable Mirostat sampling for controlling perplexity.
-     * (default: 0, 0 = disabled, 1 = Mirostat, 2 = Mirostat 2.0)
-     */
-    mirostat?: number;
-    /**
-     * Influences how quickly the algorithm responds to feedback from the generated text.
-     * A lower learning rate will result in slower adjustments,
-     * while a higher learning rate will make the algorithm more responsive. (Default: 0.1)
-     */
-    mirostatEta?: number;
-    /**
-     * Controls the balance between coherence and diversity of the output.
-     * A lower value will result in more focused and coherent text. (Default: 5.0)
-     */
-    mirostatTau?: number;
-    /**
-     * The number of GQA groups in the transformer layer. Required for some models,
-     * for example it is 8 for llama2:70b
-     */
-    numGqa?: number;
-    /**
-     * The number of layers to send to the GPU(s). On macOS it defaults to 1 to
-     * enable metal support, 0 to disable.
-     */
-    numGpu?: number;
-    /**
-     * Sets the number of threads to use during computation. By default, Ollama will
-     * detect this for optimal performance. It is recommended to set this value to the
-     * number of physical CPU cores your system has (as opposed to the logical number of cores).
-     */
-    numThreads?: number;
-    /**
-     * Sets how far back for the model to look back to prevent repetition.
-     * (Default: 64, 0 = disabled, -1 = num_ctx)
-     */
-    repeatLastN?: number;
-    /**
-     * Sets how strongly to penalize repetitions. A higher value (e.g., 1.5)
-     * will penalize repetitions more strongly, while a lower value (e.g., 0.9)
-     * will be more lenient. (Default: 1.1)
-     */
-    repeatPenalty?: number;
-    /**
-     * Sets the random number seed to use for generation. Setting this to a
-     * specific number will make the model generate the same text for the same prompt.
-     * (Default: 0)
-     */
-    seed?: number;
-    /**
-     * Tail free sampling is used to reduce the impact of less probable tokens
-     * from the output. A higher value (e.g., 2.0) will reduce the impact more,
-     * while a value of 1.0 disables this setting. (default: 1)
-     */
-    tfsZ?: number;
-    /**
-     * Reduces the probability of generating nonsense. A higher value (e.g. 100)
-     * will give more diverse answers, while a lower value (e.g. 10) will be more
-     *  conservative. (Default: 40)
-     */
-    topK?: number;
-    /**
-     * Works together with top-k. A higher value (e.g., 0.95) will lead to more
-     * diverse text, while a lower value (e.g., 0.5) will generate more focused
-     * and conservative text. (Default: 0.9)
-     */
-    topP?: number;
-    /**
-     * When set to true, no formatting will be applied to the prompt and no context
-     * will be returned.
-     */
-    raw?: boolean;
-    /**
-     * The format to return a response in. Currently the only accepted value is 'json'.
-     * Leave undefined to return a string.
-     */
-    format?: "json";
-    system?: string;
-    template?: string;
-    context?: number[];
-}
-export interface OllamaTextGenerationPrompt {
-    /**
-     * Text prompt.
-     */
-    prompt: string;
-    /**
-     Images. Supports base64-encoded `png` and `jpeg` images up to 100MB in size.
-     */
-    images?: Record<number, string>;
-}
-export declare class OllamaTextGenerationModel<CONTEXT_WINDOW_SIZE extends number | undefined> extends AbstractModel<OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>> implements TextStreamingModel<OllamaTextGenerationPrompt, OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>> {
-    constructor(settings: OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>);
-    readonly provider = "ollama";
-    get modelName(): string;
-    readonly tokenizer: undefined;
-    readonly countPromptTokens: undefined;
-    get contextWindowSize(): CONTEXT_WINDOW_SIZE;
-    callAPI<RESPONSE>(prompt: OllamaTextGenerationPrompt, options: {
-        responseFormat: OllamaTextGenerationResponseFormatType<RESPONSE>;
-    } & FunctionOptions): Promise<RESPONSE>;
-    get settingsForEvent(): Partial<OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>>;
-    doGenerateText(prompt: OllamaTextGenerationPrompt, options?: FunctionOptions): Promise<{
-        response: {
-            response: string;
-            model: string;
-            done: true;
-            total_duration: number;
-            prompt_eval_count: number;
-            eval_count: number;
-            eval_duration: number;
-            load_duration?: number | undefined;
-            context?: number[] | undefined;
-        };
-        text: string;
-    }>;
-    doStreamText(prompt: OllamaTextGenerationPrompt, options?: FunctionOptions): Promise<AsyncIterable<Delta<string>>>;
-    asToolCallGenerationModel<INPUT_PROMPT>(promptTemplate: ToolCallPromptTemplate<INPUT_PROMPT, OllamaTextGenerationPrompt>): TextGenerationToolCallModel<INPUT_PROMPT, OllamaTextGenerationPrompt, this>;
-    asToolCallsOrTextGenerationModel<INPUT_PROMPT>(promptTemplate: ToolCallsOrGenerateTextPromptTemplate<INPUT_PROMPT, OllamaTextGenerationPrompt>): TextGenerationToolCallsOrGenerateTextModel<INPUT_PROMPT, OllamaTextGenerationPrompt, this>;
-    withTextPrompt(): PromptTemplateTextStreamingModel<string, OllamaTextGenerationPrompt, OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>, this>;
-    withPromptTemplate<INPUT_PROMPT>(promptTemplate: TextGenerationPromptTemplate<INPUT_PROMPT, OllamaTextGenerationPrompt>): PromptTemplateTextStreamingModel<INPUT_PROMPT, OllamaTextGenerationPrompt, OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>, this>;
-    withSettings(additionalSettings: Partial<OllamaTextGenerationModelSettings<CONTEXT_WINDOW_SIZE>>): this;
-}
-declare const ollamaTextGenerationResponseSchema: z.ZodObject<{
-    done: z.ZodLiteral<true>;
-    model: z.ZodString;
-    response: z.ZodString;
-    total_duration: z.ZodNumber;
-    load_duration: z.ZodOptional<z.ZodNumber>;
-    prompt_eval_count: z.ZodNumber;
-    eval_count: z.ZodNumber;
-    eval_duration: z.ZodNumber;
-    context: z.ZodOptional<z.ZodArray<z.ZodNumber, "many">>;
-}, "strip", z.ZodTypeAny, {
-    response: string;
-    model: string;
-    done: true;
-    total_duration: number;
-    prompt_eval_count: number;
-    eval_count: number;
-    eval_duration: number;
-    load_duration?: number | undefined;
-    context?: number[] | undefined;
-}, {
-    response: string;
-    model: string;
-    done: true;
-    total_duration: number;
-    prompt_eval_count: number;
-    eval_count: number;
-    eval_duration: number;
-    load_duration?: number | undefined;
-    context?: number[] | undefined;
-}>;
-export type OllamaTextGenerationResponse = z.infer<typeof ollamaTextGenerationResponseSchema>;
-export type OllamaTextGenerationDelta = {
-    content: string;
-    isComplete: boolean;
-    delta: string;
-};
-export type OllamaTextGenerationResponseFormatType<T> = {
-    stream: boolean;
-    handler: ResponseHandler<T>;
-};
-export declare const OllamaTextGenerationResponseFormat: {
-    /**
-     * Returns the response as a JSON object.
-     */
-    json: {
-        stream: false;
-        handler: ({ response, url, requestBodyValues }: {
-            url: string;
-            requestBodyValues: unknown;
-            response: Response;
-        }) => Promise<{
-            response: string;
-            model: string;
-            done: true;
-            total_duration: number;
-            prompt_eval_count: number;
-            eval_count: number;
-            eval_duration: number;
-            load_duration?: number | undefined;
-            context?: number[] | undefined;
-        }>;
-    };
-    /**
-     * Returns an async iterable over the full deltas (all choices, including full current state at time of event)
-     * of the response stream.
-     */
-    deltaIterable: {
-        stream: true;
-        handler: ({ response }: {
-            response: Response;
-        }) => Promise<AsyncIterable<Delta<string>>>;
-    };
-};
-export {};