feat(python): streaming larger-than-memory writes (#2094)

Makes our preprocessing pipeline do transforms in streaming fashion, so
users can do larger-then-memory writes.

Closes #2082
This commit is contained in:
Will Jones
2025-02-06 16:37:30 -08:00
committed by GitHub
parent 4e5fbe6c99
commit 801a9e5f6f
5 changed files with 192 additions and 119 deletions

View File

@@ -107,7 +107,7 @@ def test_embedding_with_bad_results(tmp_path):
vector: Vector(model.ndims()) = model.VectorField()
table = db.create_table("test", schema=Schema, mode="overwrite")
with pytest.raises(ValueError):
with pytest.raises(RuntimeError):
# Default on_bad_vectors is "error"
table.add([{"text": "hello world"}])