Update BGE-M3 token expectations for leading spaces (#49269)
Signed-off-by: aoshen02 <aoshen02@users.noreply.github.com> Co-authored-by: aoshen02 <aoshen02@users.noreply.github.com> Co-authored-by: Codex <noreply@openai.com>
aoshen02 committed
Jul 21, 2026 at 11:49 UTC
d9aa35161d569608aed98465db0b0535b1431c28
1 file changed
+6
-6
tests/plugins_tests/test_bge_m3_sparse_io_processor_plugins.py
+6
-6
@@ -48,12 +48,12 @@ def _check_dense_embedding(data, index=0):
48
def _check_sparse_embedding(data, check_tokens=False):
49
expected_weights = [
50
{"token_id": 32, "weight": 0.0552978515625, "token": "?"},
51
- {"token_id": 70, "weight": 0.09808349609375, "token": "the"},
52
- {"token_id": 83, "weight": 0.08154296875, "token": "is"},
53
- {"token_id": 111, "weight": 0.11810302734375, "token": "of"},
54
- {"token_id": 4865, "weight": 0.1171875, "token": "What"},
55
- {"token_id": 9942, "weight": 0.292236328125, "token": "France"},
56
- {"token_id": 10323, "weight": 0.2802734375, "token": "capital"},
51
+ {"token_id": 70, "weight": 0.09808349609375, "token": " the"},
52
+ {"token_id": 83, "weight": 0.08154296875, "token": " is"},
53
+ {"token_id": 111, "weight": 0.11810302734375, "token": " of"},
54
+ {"token_id": 4865, "weight": 0.1171875, "token": " What"},
55
+ {"token_id": 9942, "weight": 0.292236328125, "token": " France"},
56
+ {"token_id": 10323, "weight": 0.2802734375, "token": " capital"},
57
]
58
expected_embed = {x["token_id"]: x for x in expected_weights}
59