Removed log statement

package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@ztimson/ai-utils",
-	"version": "0.5.1",
+	"version": "0.5.5",
 	"description": "AI Utility library",
 	"author": "Zak Timson",
 	"license": "MIT",

src/ai.ts

@@ -10,6 +10,8 @@ export type AbortablePromise<T> = Promise<T> & {
 export type AiOptions = {
 	/** Path to models */
 	path?: string;
+	/** Embedding model */
+	embedder?: string; // all-MiniLM-L6-v2, bge-small-en-v1.5, bge-large-en-v1.5
 	/** Large language models, first is default */
 	llm?: Omit<LLMRequest, 'model'> & {
 		models: {[model: string]: AnthropicConfig | OllamaConfig | OpenAiConfig};

src/antrhopic.ts

@@ -26,7 +26,7 @@ export class Anthropic extends LLMProvider {
 						h[c.is_error ? 'error' : 'content'] = c.content;
 					});
 				}
-				history[orgI].content = history[orgI].content.filter((c: any) => c.type == 'text').map((c: any) => c.text).join('\n\n');
+				history[orgI].content = history[orgI].content?.filter((c: any) => c.type == 'text').map((c: any) => c.text).join('\n\n');
 				if(!history[orgI].content) history.splice(orgI, 1);
 			}
 			if(!history[orgI].timestamp) history[orgI].timestamp = Date.now();
@@ -119,7 +119,6 @@ export class Anthropic extends LLMProvider {
 						if(options.stream) options.stream({tool: toolCall.name});
 						if(!tool) return {tool_use_id: toolCall.id, is_error: true, content: 'Tool not found'};
 						try {
-							console.log(typeof tool.fn);
 							const result = await tool.fn(toolCall.input, options?.stream, this.ai);
 							return {type: 'tool_result', tool_use_id: toolCall.id, content: JSONSanitize(result)};
 						} catch (err: any) {

src/embedder.ts

@@ -1,11 +1,14 @@
 import { pipeline } from '@xenova/transformers';
 import { parentPort } from 'worker_threads';
 
-let model: any;
+let embedder: any;
 
-parentPort?.on('message', async ({ id, text }) => {
-	if(!model) model = await pipeline('feature-extraction', 'Xenova/all-MiniLM-L6-v2');
-	const output = await model(text, { pooling: 'mean', normalize: true });
+parentPort?.on('message', async ({ id, text, model, path }) => {
+	if(!embedder) embedder = await pipeline('feature-extraction', 'Xenova/' + model, {
+		quantized: true,
+		cache_dir: path,
+	});
+	const output = await embedder(text, { pooling: 'mean', normalize: true });
 	const embedding = Array.from(output.data);
 	parentPort?.postMessage({ id, embedding });
 });

src/llm.ts

@@ -271,7 +271,12 @@ class LLM {
 			return new Promise((resolve, reject) => {
 				const id = this.embedId++;
 				this.embedQueue.set(id, { resolve, reject });
-				this.embedWorker?.postMessage({ id, text });
+				this.embedWorker?.postMessage({
+					id,
+					text,
+					model: this.ai.options?.embedder || 'bge-small-en-v1.5',
+					path: this.ai.options.path
+				});
 			});
 		};
 		const chunks = this.chunk(target, maxTokens, overlapTokens);