Modified NER prompt for keyword position [WIP]

config/ner.js

@@ -1,6 +1,7 @@
 export const ner = ({
     text_input = "",
     description = "",
+    position,
     domain = "",
     labels = "",
     examples = [],
@@ -14,13 +15,15 @@ export const ner = ({
     const _example =
         examples && examples.length > 0
             ? `Examples: ${examples.map(
-                  (ex) => `Input ${ex.sentence} Output [${ex.sentence}]`
+                  (ex) => `Input ${ex[0].E} Output [${ex[0].T}]`
               )}`
             : "";
     return `
         ${description} 
         You are a highly intelligent and accurate ${_domain} Named-entity recognition(NER) system. You take Passage as input and your task is to recognize and extract specific types of ${_domain} named entities in that given passage and classify into a set of ${_label}.
-        Your output format is only {{ output_format|default("[{{'T': type of entity from predefined entity types, 'E': entity in the input text}},...,{{'branch' : Appropriate branch of the passage ,'group': Appropriate Group of the passage}}]") }} form, no other form.
+
+        Your output format is ${position ? `{{ output_format|default("[{{'T': type of entity from predefined entity types, 'E': entity in the input text}},...,{{'branch' : Appropriate branch of the passage ,'group': Appropriate Group of the passage}},...,{{ 'start' : start position of the entity, 'end' : end position of the entity}}]") }}` : `{{ output_format|default("[{{'T': type of entity from predefined entity types, 'E': entity in the input text}},...,{{'branch' : Appropriate branch of the passage ,'group': Appropriate Group of the passage}}]") }}`} form, otherwise no other form.
+
         ${_example}
         Input: ${text_input}
         Output:

examples/data/optimized_ner.js

@@ -5,10 +5,14 @@ export const nerData = {
                 {
                     E: "DISEASE",
                     T: "Alzheimer's disease",
+                    start: 0,
+                    end: 10
                 },
                 {
                     E: "BIOMARKER",
                     T: "Amyloid β42",
+                    start: 5,
+                    end: 16
                 },
                 {
                     E: "BIOMARKER",

examples/nerExample.js

@@ -3,7 +3,7 @@ import { OpenAI } from "../models/openai.js";
 import { Prompter } from "../promptify/index.js";
 import { nerData } from "../examples/data/optimized_ner.js";
 
-const model = OpenAI("api-key");
+const model = OpenAI("openai-api");
 const examples = nerData.samples[0].data;
 const firstExample = examples.slice(0, 3);
 
@@ -12,9 +12,12 @@ const prompt = ner({
     description: "Medicine NER Expert",
     domain: "medicine",
     labels: "",
+    position: true,
     examples: [firstExample],
 });
 
+console.log(prompt)
+
 const result = await Prompter(model, prompt, "text-davinci-003");
 
 console.log(result);