File size: 21,511 Bytes
f8f07da | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 | internal_id,display_name,family,type,merge_recipe,suites_scored,full_score_source_eng,tasks_with_full_score_eng,full_score_source_ara,tasks_with_full_score_ara,role_eng,role_eng-super,role_ara-super
A-dare-p0.5-l0.5,"ADIP-v3 (step 6750), DARE p=0.5 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), DARE p=0.5 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
A-dare-p0.9-l1.0,"ADIP-v3 (step 6750), DARE p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), DARE p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
A-lin0.4,"ADIP-v3 (step 6750), linear merge λ=0.4 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), linear merge λ=0.4 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
A-lin0.5,"ADIP-v3 (step 6750), linear merge λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), linear merge λ=0.5 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
A-lin0.6,"ADIP-v3 (step 6750), linear merge λ=0.6 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), linear merge λ=0.6 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
A-ties-d0.1-l0.5,"ADIP-v3 (step 6750), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750),ADIP merge,"ADIP-v3 (step 6750), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
AS3-dareties-p0.9-l1.0,"ADIP-v3 (step 6750) + stage-3: ADIP-v3 (step 6750) + ADIP-v3 stage-3, DARE-TIES p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + stage-3,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v3 stage-3, DARE-TIES p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AS3-lin1.0,"ADIP-v3 (step 6750) + stage-3: ADIP-v3 (step 6750) + ADIP-v3 stage-3, linear merge λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + stage-3,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v3 stage-3, linear merge λ=1.0 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),33,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
AS3-ties-d0.3-l0.5,"ADIP-v3 (step 6750) + stage-3: ADIP-v3 (step 6750) + ADIP-v3 stage-3, TIES d=0.3 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + stage-3,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v3 stage-3, TIES d=0.3 λ=0.5 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AS3-ties-d0.3-l1.0,"ADIP-v3 (step 6750) + stage-3: ADIP-v3 (step 6750) + ADIP-v3 stage-3, TIES d=0.3 λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + stage-3,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v3 stage-3, TIES d=0.3 λ=1.0 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AV2-dareties-p0.9-l0.5,"ADIP-v3 (step 6750) + ADIP-v2-ft, DARE-TIES p=0.9 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, DARE-TIES p=0.9 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),33,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
AV2-dareties-p0.9-l1.0,"ADIP-v3 (step 6750) + ADIP-v2-ft, DARE-TIES p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, DARE-TIES p=0.9 λ=1.0 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AV2-lin0.5,"ADIP-v3 (step 6750) + ADIP-v2-ft, linear merge λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, linear merge λ=0.5 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AV2-lin1.0,"ADIP-v3 (step 6750) + ADIP-v2-ft, linear merge λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, linear merge λ=1.0 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),33,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
AV2-ties-d0.3-l0.5,"ADIP-v3 (step 6750) + ADIP-v2-ft, TIES d=0.3 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, TIES d=0.3 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),33,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
AV2-ties-d0.3-l1.0,"ADIP-v3 (step 6750) + ADIP-v2-ft, TIES d=0.3 λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (step 6750) + ADIP-v2-ft,ADIP merge,"ADIP-v3 (step 6750) + ADIP-v2-ft, TIES d=0.3 λ=1.0 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),33,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
F-lin0.4,"ADIP-v3 (final), linear merge λ=0.4 into Qwen3-Embedding-0.6B",ADIP-v3 (final),ADIP merge,"ADIP-v3 (final), linear merge λ=0.4 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,not used (family cap: <=3 per family),not used (family cap: <=3 per family),selection+calibration+LOMO-held-out+aggregate-validation
F-lin0.5,ADIP-QCRI/ADIP-Embedding-v1-0.6B,ADIP-v3 (final),ADIP release,"ADIP-v3 (final), linear merge λ=0.5 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
F-lin0.6,"ADIP-v3 (final), linear merge λ=0.6 into Qwen3-Embedding-0.6B",ADIP-v3 (final),ADIP merge,"ADIP-v3 (final), linear merge λ=0.6 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
F-lin1.0,"ADIP-v3 (final), linear merge λ=1.0 into Qwen3-Embedding-0.6B",ADIP-v3 (final),ADIP merge,"ADIP-v3 (final), linear merge λ=1.0 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
F-ties-d0.1-l0.5,"ADIP-v3 (final), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (final),ADIP merge,"ADIP-v3 (final), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
FV-ties-d0.1-l0.5,"ADIP-v3+v4: ADIP-v3 (final) + ADIP-v4, TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3+v4,ADIP merge,"ADIP-v3 (final) + ADIP-v4, TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
FV-ties-d0.3-l0.7,"ADIP-v3+v4: ADIP-v3 (final) + ADIP-v4, TIES d=0.3 λ=0.7 into Qwen3-Embedding-0.6B",ADIP-v3+v4,ADIP merge,"ADIP-v3 (final) + ADIP-v4, TIES d=0.3 λ=0.7 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
GIST-small-Embedding-v0,avsolatorio/GIST-small-Embedding-v0,avsolatorio/GIST-small-Embedding-v0,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
LaBSE,sentence-transformers/LaBSE,sentence-transformers/LaBSE,public,,eng/ara,official MTEB results,41,official MTEB results,42,not used,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
P-lin0.5,"ADIP-v3 (pre-rehearsal), linear merge λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (pre-rehearsal),ADIP merge,"ADIP-v3 (pre-rehearsal), linear merge λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
P-ties-d0.1-l0.5,"ADIP-v3 (pre-rehearsal), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",ADIP-v3 (pre-rehearsal),ADIP merge,"ADIP-v3 (pre-rehearsal), TIES d=0.1 λ=0.5 into Qwen3-Embedding-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
S-lin0.5,"ADIP-v4 (stage 1), linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",ADIP-v4 (stage 1),ADIP merge,"ADIP-v4 (stage 1), linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
S5-lin0.5,"ADIP-v5 (stage 1), linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",ADIP-v5 (stage 1),ADIP merge,"ADIP-v5 (stage 1), linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
UAE-Large-V1,WhereIsAI/UAE-Large-V1,WhereIsAI/UAE-Large-V1,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
V-lin0.1,"ADIP-v4, linear merge λ=0.1 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, linear merge λ=0.1 into ADIP-Embedding-v1-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
V-lin0.3,"ADIP-v4, linear merge λ=0.3 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, linear merge λ=0.3 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
V-lin0.5,"ADIP-v4, linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,not used (family cap: <=3 per family),not used (family cap: <=3 per family),selection+calibration+LOMO-held-out+aggregate-validation
V-lin0.7,"ADIP-v4, linear merge λ=0.7 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, linear merge λ=0.7 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
V-lin1.0,"ADIP-v4, linear merge λ=1.0 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, linear merge λ=1.0 into ADIP-Embedding-v1-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
V-ties-d0.1-l0.5,"ADIP-v4, TIES d=0.1 λ=0.5 into ADIP-Embedding-v1-0.6B",ADIP-v4,ADIP merge,"ADIP-v4, TIES d=0.1 λ=0.5 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
V5-lin0.3,"ADIP-v5, linear merge λ=0.3 into ADIP-Embedding-v1-0.6B",ADIP-v5,ADIP merge,"ADIP-v5, linear merge λ=0.3 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
V5-lin0.5,"ADIP-v5, linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",ADIP-v5,ADIP merge,"ADIP-v5, linear merge λ=0.5 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,not used (family cap: <=3 per family),not used (family cap: <=3 per family),-
V5-lin0.7,"ADIP-v5, linear merge λ=0.7 into ADIP-Embedding-v1-0.6B",ADIP-v5,ADIP merge,"ADIP-v5, linear merge λ=0.7 into ADIP-Embedding-v1-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
V5-lin1.0,"ADIP-v5, linear merge λ=1.0 into ADIP-Embedding-v1-0.6B",ADIP-v5,ADIP merge,"ADIP-v5, linear merge λ=1.0 into ADIP-Embedding-v1-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
V5-ties-d0.1-l0.5,ADIP-QCRI/ADIP-Embedding-v2-0.6B,ADIP-v5,ADIP release,"ADIP-v5, TIES d=0.1 λ=0.5 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
V5-ties-d0.3-l0.7,"ADIP-v5, TIES d=0.3 λ=0.7 into ADIP-Embedding-v1-0.6B",ADIP-v5,ADIP merge,"ADIP-v5, TIES d=0.3 λ=0.7 into ADIP-Embedding-v1-0.6B",eng,our runs (mteb 1.39.7),41,,0,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,-
Vq-lin0.4,"ADIP-v3+v4: ADIP-v4, linear merge λ=0.4 into Qwen3-Embedding-0.6B",ADIP-v3+v4,ADIP merge,"ADIP-v4, linear merge λ=0.4 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
Vq-lin0.6,"ADIP-v3+v4: ADIP-v4, linear merge λ=0.6 into Qwen3-Embedding-0.6B",ADIP-v3+v4,ADIP merge,"ADIP-v4, linear merge λ=0.6 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
Vq-lin0.7,"ADIP-v3+v4: ADIP-v4, linear merge λ=0.7 into Qwen3-Embedding-0.6B",ADIP-v3+v4,ADIP merge,"ADIP-v4, linear merge λ=0.7 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
W-A0.7-S3-0.3-l0.5,"ADIP weighted soup: 0.7·ADIP-v3 (step 6750) + 0.3·ADIP-v3 stage-3, linear merge λ=0.5 into Qwen3-Embedding-0.6B",ADIP weighted soup,ADIP merge,"0.7·ADIP-v3 (step 6750) + 0.3·ADIP-v3 stage-3, linear merge λ=0.5 into Qwen3-Embedding-0.6B",eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
W-A0.7-V2-0.3-l0.5,"ADIP weighted soup: 0.7·ADIP-v3 (step 6750) + 0.3·ADIP-v2-ft, linear merge λ=0.5 into Qwen3-Embedding-0.6B",ADIP weighted soup,ADIP merge,"0.7·ADIP-v3 (step 6750) + 0.3·ADIP-v2-ft, linear merge λ=0.5 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
W-F0.7-V0.3-l0.7,"ADIP weighted soup: 0.7·ADIP-v3 (final) + 0.3·ADIP-v4, linear merge λ=0.7 into Qwen3-Embedding-0.6B",ADIP weighted soup,ADIP merge,"0.7·ADIP-v3 (final) + 0.3·ADIP-v4, linear merge λ=0.7 into Qwen3-Embedding-0.6B",ara,,0,our runs (mteb 1.39.7),42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
all-MiniLM-L12-v2,sentence-transformers/all-MiniLM-L12-v2,sentence-transformers/all-MiniLM-L12-v2,public,,eng,official MTEB results,38,,0,not used,selection+calibration+LOMO-held-out,-
all-MiniLM-L6-v2,sentence-transformers/all-MiniLM-L6-v2,sentence-transformers/all-MiniLM-L6-v2,public,,eng/ara,official MTEB results,41,official MTEB results,28,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out
all-mpnet-base-v2,sentence-transformers/all-mpnet-base-v2,sentence-transformers/all-mpnet-base-v2,public,,eng,official MTEB results,39,,0,not used,selection+calibration+LOMO-held-out,-
arctic-embed-l,Snowflake/snowflake-arctic-embed-l,Snowflake/snowflake-arctic-embed-l,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
arctic-embed-m-v1.5,Snowflake/snowflake-arctic-embed-m-v1.5,Snowflake/snowflake-arctic-embed-m-v1.5,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
arctic-embed-xs,Snowflake/snowflake-arctic-embed-xs,Snowflake/snowflake-arctic-embed-xs,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
bge-base-en-v1.5,BAAI/bge-base-en-v1.5,BAAI/bge-base-en-v1.5,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
bge-large-en-v1.5,BAAI/bge-large-en-v1.5,BAAI/bge-large-en-v1.5,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
bge-m3,BAAI/bge-m3,BAAI/bge-m3,public,,eng/ara,official MTEB results,27,official MTEB results,42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
bge-small-en-v1.5,BAAI/bge-small-en-v1.5,BAAI/bge-small-en-v1.5,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
e5-base-v2,intfloat/e5-base-v2,intfloat/e5-base-v2,public,,eng/ara,official MTEB results,41,official MTEB results,21,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out
e5-large-v2,intfloat/e5-large-v2,intfloat/e5-large-v2,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
e5-small-v2,intfloat/e5-small-v2,intfloat/e5-small-v2,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
embeddinggemma-300m,google/embeddinggemma-300m,google/embeddinggemma-300m,public,,eng/ara,official MTEB results,41,official MTEB results,42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
gte-large,thenlper/gte-large,thenlper/gte-large,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
gte-small,thenlper/gte-small,thenlper/gte-small,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
multilingual-e5-base,intfloat/multilingual-e5-base,intfloat/multilingual-e5-base,public,,eng/ara,official MTEB results,41,official MTEB results,42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
multilingual-e5-large,intfloat/multilingual-e5-large,intfloat/multilingual-e5-large,public,,eng/ara,official MTEB results,40,official MTEB results,42,not used,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
multilingual-e5-small,intfloat/multilingual-e5-small,intfloat/multilingual-e5-small,public,,eng/ara,official MTEB results,41,official MTEB results,42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
mxbai-embed-large-v1,mixedbread-ai/mxbai-embed-large-v1,mixedbread-ai/mxbai-embed-large-v1,public,,eng,official MTEB results,41,,0,not used,selection+calibration+LOMO-held-out+aggregate-validation,-
paraphrase-multilingual-MiniLM-L12-v2,sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2,sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2,public,,ara,,0,official MTEB results,42,-,-,selection+calibration+LOMO-held-out+aggregate-validation
paraphrase-multilingual-mpnet-base-v2,sentence-transformers/paraphrase-multilingual-mpnet-base-v2,sentence-transformers/paraphrase-multilingual-mpnet-base-v2,public,,eng/ara,official MTEB results,41,official MTEB results,42,not used,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
qwen3-0.6b,Qwen/Qwen3-Embedding-0.6B,Qwen3-Embedding base,Qwen base,,eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
qwen3-0.6b+norm,Qwen/Qwen3-Embedding-0.6B + ADIP tokenizer,Qwen3-Embedding base,Qwen base,,eng/ara,our runs (mteb 1.39.7),41,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation,selection+calibration+LOMO-held-out+aggregate-validation
qwen3-4b,Qwen/Qwen3-Embedding-4B,Qwen3-Embedding base,Qwen base,,eng/ara,our runs (mteb 1.39.7),39,our runs (mteb 1.39.7),42,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out,selection+calibration+LOMO-held-out+aggregate-validation
|