time-embed-bge-m3-lms-temporal-phase2-72

Time-Embed BGE-M3 LMS Temporal Phase2.72

This model is a Korean temporal-expression embedding checkpoint for LMS retrieval. It starts from the Time-Embed phase2.67 checkpoint and applies a small multi-objective pairwise update with train-only strict-frontier analogs.

HF repo: kev-KOH/time-embed-bge-m3-lms-temporal-phase2-72

Intended Use

Use this checkpoint for Korean LMS retrieval queries where relative temporal expressions must be separated precisely, for example:

• 일주일 전, 7일 전, 1주 전 as exact duration offsets
• 지난주 as a calendar week period
• 최근 7일 as a rolling window
• 다음 주 versus 7일 뒤

The model is not a general-purpose replacement for BAAI/bge-m3; it is a specialized temporal retrieval checkpoint. Run downstream retrieval validation before production use.

Training Method

• Base/source checkpoint: research/runs/time_embed_phase2_67_p65_uploaded_material_calendar_tail_48_4070ti/20260626-071336
• Objective: multi-objective pairwise margin with positive-floor, negative-ceiling, teacher score-MSE, and PCGrad
• Trainable layers: last two encoder layers
• Learning rate: 5e-9
• Steps: 80
• Precision: fp32
• Training shards:
  • temporal exact-week frontier analogs: 50 rows
  • semantic year-offset frontier analogs: 48 rows
  • uploaded-material calendar frontier analogs: 72 rows

Internal Strict-Frontier Evaluation

Caveat: the semantic strict-frontier gain is very small (+0.000010041). Treat this checkpoint as a narrow frontier improvement and continue to monitor semantic retention in downstream data.

MTEB Catastrophic Forgetting Check

Compared against BAAI/bge-m3 in the same runtime:

No general catastrophic forgetting was detected by the project default rule: all drops stayed within the -0.01 retention band.

Usage

from FlagEmbedding import BGEM3FlagModel

model = BGEM3FlagModel(
    "kev-KOH/time-embed-bge-m3-lms-temporal-phase2-72",
    use_fp16=True,
    pooling_method="cls",
)
embeddings = model.encode(
    ["7일 전 업로드된 데이터 자료 찾아줘", "지난주 업로드된 데이터 자료 찾아줘"],
    return_dense=True,
    return_sparse=False,
    return_colbert_vecs=False,
)["dense_vecs"]

Artifacts

• model.safetensors
• config.json
• tokenizer.json
• tokenizer_config.json
• multi_objective_pairwise_training_summary.json