Solution to #1: Make cosine reject vectors of different lengths in the translator diarizer #2

Closed
elffuss-bot wants to merge 1 commits from item/0x51341970906a1fd3227c4224d8c075bd722f9c06dcc276acceeb52c37efa2774 into main
3 changed files with 67 additions and 1 deletions

View File

@@ -25,7 +25,18 @@ export async function loadDiarizer(TF, { device = 'wasm' } = {}) {
return { seg, segP, emb, embP };
}
const cos = (a, b) => { let dt = 0, na = 0, nb = 0; for (let i = 0; i < a.length; i++) { dt += a[i]*b[i]; na += a[i]*a[i]; nb += b[i]*b[i]; } return dt / (Math.sqrt(na)*Math.sqrt(nb) || 1); };
const cos = (a, b) => {
if (a.length !== b.length) {
throw new Error('cosine: vectors must have the same length');
}
let dt = 0, na = 0, nb = 0;
for (let i = 0; i < a.length; i++) {
dt += a[i]*b[i];
na += a[i]*a[i];
nb += b[i]*b[i];
}
return dt / (Math.sqrt(na)*Math.sqrt(nb) || 1);
};
const rms = (a) => { let s = 0; for (const v of a) s += v*v; return Math.sqrt(s/a.length); };
async function embed(D, slice) {

View File

@@ -0,0 +1,30 @@
{
"schema": "elffuss-t2t/criteria@2",
"issue": 1,
"criteria": [
{
"index": 1,
"tests": [
"tests/acceptance/1/criteria.test.js::criterion 1"
]
},
{
"index": 2,
"tests": [
"tests/acceptance/1/criteria.test.js::criterion 2"
]
},
{
"index": 3,
"tests": [
"tests/acceptance/1/criteria.test.js::criterion 3"
]
},
{
"index": 4,
"tests": [
"tests/acceptance/1/criteria.test.js::criterion 4"
]
}
]
}

View File

@@ -0,0 +1,25 @@
import { test } from 'node:test'
import assert from 'node:assert/strict'
import { readFileSync } from 'node:fs'
import * as mod from '../../../diarize.mjs' // the code under test: call it as mod.<name>(...)
const source = readFileSync(new URL('../../../diarize.mjs', import.meta.url), 'utf8') // the text of diarize.mjs, for a criterion about how the file is written
test('criterion 1', () => {
// `cosine([1, 2, 3], [1, 2])` throws an Error whose message starts with `cosine:`.
assert.throws(() => mod.cosine([1, 2, 3], [1, 2]), /cosine:/)
})
test('criterion 2', () => {
// `cosine([1, 0], [1, 0, 5])` throws an Error whose message starts with `cosine:`.
assert.throws(() => mod.cosine([1, 0], [1, 0, 5]), /cosine:/)
})
test('criterion 3', () => {
// `cosine([1, 0], [1, 0])` returns 1.
assert.deepEqual(mod.cosine([1, 0], [1, 0]), 1)
})
test('criterion 4', () => {
// `cosine([0, 0], [0, 0])` returns 0.
assert.deepEqual(mod.cosine([0, 0], [0, 0]), 0)
})