refactor: rename LOCAL_ONNX_TRANSFORMERJS to ONNX_TRANSFORMERJS

sroussey · Mar 5, 2025 · f4ecc27 · f4ecc27
1 parent 11192a0
commit f4ecc27
Show file tree

Hide file tree

Showing 8 changed files with 33 additions and 38 deletions.
diff --git a/docs/developers/01_getting_started.md b/docs/developers/01_getting_started.md
@@ -156,7 +156,7 @@ await getGlobalModelRepository().addModel({
   url: "Xenova/LaMini-Flan-T5-783M",
   availableOnBrowser: true,
   availableOnServer: true,
-  provider: LOCAL_ONNX_TRANSFORMERJS,
+  provider: ONNX_TRANSFORMERJS,
   pipeline: "text2text-generation",
 });
 await getGlobalModelRepository().connectTaskToModel(
@@ -171,12 +171,12 @@ await getGlobalModelRepository().connectTaskToModel(
 const aiProviderRegistry = getAiProviderRegistry();
 aiProviderRegistry.registerRunFn(
   DownloadModelTask.type,
-  LOCAL_ONNX_TRANSFORMERJS,
+  ONNX_TRANSFORMERJS,
   HuggingFaceLocal_DownloadRun
 );
 aiProviderRegistry.registerRunFn(
   TextRewriterTask.type,
-  LOCAL_ONNX_TRANSFORMERJS,
+  ONNX_TRANSFORMERJS,
   HuggingFaceLocal_TextRewriterRun
 );
 const jobQueue = new JobQueue<TaskInput, TaskOutput>("test", Job, {

diff --git a/examples/web/src/App.tsx b/examples/web/src/App.tsx
@@ -10,7 +10,7 @@ import { env } from "@huggingface/transformers";
 import { ReactFlowProvider } from "@xyflow/react";
 import { AiJob } from "@ellmers/ai";
 import {
-  LOCAL_ONNX_TRANSFORMERJS,
+  ONNX_TRANSFORMERJS,
   registerHuggingfaceLocalTasks,
 } from "@ellmers/ai-provider/hf-transformers";
 import {
@@ -43,7 +43,7 @@ const queueRegistry = getTaskQueueRegistry();
 
 registerHuggingfaceLocalTasks();
 queueRegistry.registerQueue(
-  new JobQueue<TaskInput, TaskOutput>(LOCAL_ONNX_TRANSFORMERJS, AiJob<TaskInput, TaskOutput>, {
+  new JobQueue<TaskInput, TaskOutput>(ONNX_TRANSFORMERJS, AiJob<TaskInput, TaskOutput>, {
     limiter: new ConcurrencyLimiter(1, 10),
   })
 );

diff --git a/packages/ai-provider/src/hf-transformers/bindings/registerTasks.ts b/packages/ai-provider/src/hf-transformers/bindings/registerTasks.ts
@@ -17,50 +17,50 @@ import {
   HuggingFaceLocal_TextSummaryRun,
   HuggingFaceLocal_TextTranslationRun,
 } from "../provider/HuggingFaceLocal_TaskRun";
-import { LOCAL_ONNX_TRANSFORMERJS } from "../model/ONNXTransformerJsModel";
+import { ONNX_TRANSFORMERJS } from "../model/ONNXTransformerJsModel";
 
 export async function registerHuggingfaceLocalTasks() {
   const ProviderRegistry = getAiProviderRegistry();
 
   ProviderRegistry.registerRunFn(
     DownloadModelTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_DownloadRun
   );
 
   ProviderRegistry.registerRunFn(
     TextEmbeddingTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_EmbeddingRun
   );
 
   ProviderRegistry.registerRunFn(
     TextGenerationTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_TextGenerationRun
   );
 
   ProviderRegistry.registerRunFn(
     TextTranslationTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_TextTranslationRun
   );
 
   ProviderRegistry.registerRunFn(
     TextRewriterTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_TextRewriterRun
   );
 
   ProviderRegistry.registerRunFn(
     TextSummaryTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_TextSummaryRun
   );
 
   ProviderRegistry.registerRunFn(
     TextQuestionAnswerTask.type,
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     HuggingFaceLocal_TextQuestionAnswerRun
   );
 }
diff --git a/packages/ai-provider/src/hf-transformers/model/ONNXTransformerJsModel.ts b/packages/ai-provider/src/hf-transformers/model/ONNXTransformerJsModel.ts
@@ -5,7 +5,7 @@
 //    *   Licensed under the Apache License, Version 2.0 (the "License");           *
 //    *******************************************************************************
 
-export const LOCAL_ONNX_TRANSFORMERJS = "LOCAL_ONNX_TRANSFORMERJS";
+export const ONNX_TRANSFORMERJS = "ONNX_TRANSFORMERJS";
 
 export enum QUANTIZATION_DATA_TYPES {
   auto = "auto", // Auto-detect based on environment

diff --git a/packages/ai-provider/src/hf-transformers/test/HFTransformersBinding.test.ts b/packages/ai-provider/src/hf-transformers/test/HFTransformersBinding.test.ts
@@ -22,7 +22,7 @@ import {
 } from "@ellmers/task-graph";
 import { Sqlite, sleep } from "@ellmers/util";
 import { registerHuggingfaceLocalTasks } from "../bindings/registerTasks";
-import { LOCAL_ONNX_TRANSFORMERJS } from "../model/ONNXTransformerJsModel";
+import { ONNX_TRANSFORMERJS } from "../model/ONNXTransformerJsModel";
 import { AiProviderInput } from "@ellmers/ai";
 import { SqliteQueueStorage } from "@ellmers/storage";
 
@@ -37,7 +37,7 @@ describe("HFTransformersBinding", () => {
       registerHuggingfaceLocalTasks();
       const queueRegistry = getTaskQueueRegistry();
       const jobQueue = new JobQueue<AiProviderInput<TaskInput>, TaskOutput>(
-        LOCAL_ONNX_TRANSFORMERJS,
+        ONNX_TRANSFORMERJS,
         AiJob<TaskInput, TaskOutput>,
         {
           limiter: new ConcurrencyLimiter(1, 10),
@@ -50,7 +50,7 @@ describe("HFTransformersBinding", () => {
         url: "Xenova/LaMini-Flan-T5-783M",
         availableOnBrowser: true,
         availableOnServer: true,
-        provider: LOCAL_ONNX_TRANSFORMERJS,
+        provider: ONNX_TRANSFORMERJS,
         pipeline: "text2text-generation",
       };
 
@@ -59,9 +59,9 @@ describe("HFTransformersBinding", () => {
       await getGlobalModelRepository().connectTaskToModel("TextGenerationTask", model.name);
       await getGlobalModelRepository().connectTaskToModel("TextRewriterTask", model.name);
 
-      const queue = queueRegistry.getQueue(LOCAL_ONNX_TRANSFORMERJS);
+      const queue = queueRegistry.getQueue(ONNX_TRANSFORMERJS);
       expect(queue).toBeDefined();
-      expect(queue!.queueName).toEqual(LOCAL_ONNX_TRANSFORMERJS);
+      expect(queue!.queueName).toEqual(ONNX_TRANSFORMERJS);
 
       const workflow = new Workflow();
       workflow.DownloadModel({
@@ -80,7 +80,7 @@ describe("HFTransformersBinding", () => {
       registerHuggingfaceLocalTasks();
       const queueRegistry = getTaskQueueRegistry();
       const jobQueue = new JobQueue<AiProviderInput<TaskInput>, TaskOutput>(
-        LOCAL_ONNX_TRANSFORMERJS,
+        ONNX_TRANSFORMERJS,
         AiJob<TaskInput, TaskOutput>,
         {
           storage: new SqliteQueueStorage<AiProviderInput<TaskInput>, TaskOutput>(db, "test"),
@@ -97,17 +97,17 @@ describe("HFTransformersBinding", () => {
         url: "Xenova/LaMini-Flan-T5-783M",
         availableOnBrowser: true,
         availableOnServer: true,
-        provider: LOCAL_ONNX_TRANSFORMERJS,
+        provider: ONNX_TRANSFORMERJS,
         pipeline: "text2text-generation",
       };
 
       await getGlobalModelRepository().addModel(model);
       await getGlobalModelRepository().connectTaskToModel("TextGenerationTask", model.name);
       await getGlobalModelRepository().connectTaskToModel("TextRewriterTask", model.name);
 
-      const queue = queueRegistry.getQueue(LOCAL_ONNX_TRANSFORMERJS);
+      const queue = queueRegistry.getQueue(ONNX_TRANSFORMERJS);
       expect(queue).toBeDefined();
-      expect(queue?.queueName).toEqual(LOCAL_ONNX_TRANSFORMERJS);
+      expect(queue?.queueName).toEqual(ONNX_TRANSFORMERJS);
 
       const workflow = new Workflow();
       workflow.DownloadModel({

diff --git a/packages/ai/src/model/storage/test/genericModelRepositoryTests.ts b/packages/ai/src/model/storage/test/genericModelRepositoryTests.ts
@@ -10,7 +10,7 @@ import { getGlobalModelRepository } from "../../ModelRegistry";
 import { ModelRepository } from "../../ModelRepository";
 import { setGlobalModelRepository } from "../../ModelRegistry";
 
-const LOCAL_ONNX_TRANSFORMERJS = "LOCAL_ONNX_TRANSFORMERJS";
+const ONNX_TRANSFORMERJS = "ONNX_TRANSFORMERJS";
 
 export const runGenericModelRepositoryTests = (
   createRepository: () => Promise<ModelRepository>
@@ -32,7 +32,7 @@ export const runGenericModelRepositoryTests = (
       url: "Xenova/LaMini-Flan-T5-783M",
       availableOnBrowser: true,
       availableOnServer: true,
-      provider: LOCAL_ONNX_TRANSFORMERJS,
+      provider: ONNX_TRANSFORMERJS,
       pipeline: "text2text-generation",
     });
 
@@ -50,7 +50,7 @@ export const runGenericModelRepositoryTests = (
       url: "Xenova/LaMini-Flan-T5-783M",
       availableOnBrowser: true,
       availableOnServer: true,
-      provider: LOCAL_ONNX_TRANSFORMERJS,
+      provider: ONNX_TRANSFORMERJS,
       pipeline: "text2text-generation",
     });
     await getGlobalModelRepository().connectTaskToModel(
@@ -76,7 +76,7 @@ export const runGenericModelRepositoryTests = (
       url: "Xenova/LaMini-Flan-T5-783M",
       availableOnBrowser: true,
       availableOnServer: true,
-      provider: LOCAL_ONNX_TRANSFORMERJS,
+      provider: ONNX_TRANSFORMERJS,
       pipeline: "text2text-generation",
     });
 
@@ -89,7 +89,7 @@ export const runGenericModelRepositoryTests = (
     expect(models).toBeDefined();
     expect(models?.length).toEqual(1);
     expect(models?.[0].name).toEqual("onnx:Xenova/LaMini-Flan-T5-783M:q8");
-    expect(models?.[0].provider).toEqual(LOCAL_ONNX_TRANSFORMERJS);
+    expect(models?.[0].provider).toEqual(ONNX_TRANSFORMERJS);
     expect(models?.[0].pipeline).toEqual("text2text-generation");
   });
 };
diff --git a/packages/test/src/index.ts b/packages/test/src/index.ts
@@ -7,7 +7,7 @@
 
 import { AiJob, AiProviderInput } from "@ellmers/ai";
 import {
-  LOCAL_ONNX_TRANSFORMERJS,
+  ONNX_TRANSFORMERJS,
   registerHuggingfaceLocalTasks,
 } from "@ellmers/ai-provider/hf-transformers";
 import {
@@ -23,12 +23,10 @@ export * from "./sample/ONNXModelSamples";
 export async function registerHuggingfaceLocalTasksInMemory() {
   registerHuggingfaceLocalTasks();
   const jobQueue = new JobQueue<AiProviderInput<TaskInput>, TaskOutput>(
-    LOCAL_ONNX_TRANSFORMERJS,
+    ONNX_TRANSFORMERJS,
     AiJob<TaskInput, TaskOutput>,
     {
-      storage: new InMemoryQueueStorage<AiProviderInput<TaskInput>, TaskOutput>(
-        LOCAL_ONNX_TRANSFORMERJS
-      ),
+      storage: new InMemoryQueueStorage<AiProviderInput<TaskInput>, TaskOutput>(ONNX_TRANSFORMERJS),
       limiter: new ConcurrencyLimiter(1, 10),
     }
   );

diff --git a/packages/test/src/sample/ONNXModelSamples.ts b/packages/test/src/sample/ONNXModelSamples.ts
@@ -1,14 +1,11 @@
-import {
-  LOCAL_ONNX_TRANSFORMERJS,
-  QUANTIZATION_DATA_TYPES,
-} from "@ellmers/ai-provider/hf-transformers";
+import { ONNX_TRANSFORMERJS, QUANTIZATION_DATA_TYPES } from "@ellmers/ai-provider/hf-transformers";
 import { getGlobalModelRepository, Model } from "@ellmers/ai";
 
 export async function addONNXModel(info: Partial<Model>, tasks: string[]) {
   const model = Object.assign(
     {
       name: "onnx:" + info.url + ":" + (info.quantization ?? QUANTIZATION_DATA_TYPES.q8),
-      provider: LOCAL_ONNX_TRANSFORMERJS,
+      provider: ONNX_TRANSFORMERJS,
       quantization: QUANTIZATION_DATA_TYPES.q8,
       normalize: true,
       contextWindow: 4096,