| """测试 embedder 模块.""" |
|
|
| import pytest |
| from hermes_core.embedder import Embedder |
|
|
| pytestmark = pytest.mark.requires_embedder |
|
|
|
|
| class TestEmbedder: |
| @pytest.fixture |
| def embedder(self): |
| return Embedder() |
|
|
| def test_encode_returns_384_dim_vector(self, embedder): |
| vec = embedder.encode("后端API开发") |
| assert isinstance(vec, list) |
| assert len(vec) == 384 |
| assert all(isinstance(v, float) for v in vec) |
|
|
| def test_encode_batch_returns_same_count(self, embedder): |
| texts = ["后端开发", "前端开发", "游戏开发"] |
| vecs = embedder.encode_batch(texts) |
| assert len(vecs) == 3 |
| assert all(len(v) == 384 for v in vecs) |
|
|
| def test_similar_texts_have_high_similarity(self, embedder): |
| a = embedder.encode("后端API开发") |
| b = embedder.encode("后端服务开发") |
| sim = embedder.cosine_similarity(a, b) |
| assert sim > 0.5 |
|
|
| def test_different_texts_have_low_similarity(self, embedder): |
| a = embedder.encode("后端API开发") |
| b = embedder.encode("周末去哪玩") |
| sim = embedder.cosine_similarity(a, b) |
| assert sim < 0.5 |
|
|
| def test_identical_texts_have_similarity_one(self, embedder): |
| a = embedder.encode("TypeScript") |
| b = embedder.encode("TypeScript") |
| sim = embedder.cosine_similarity(a, b) |
| assert sim > 0.99 |
|
|
