HyperCLOVA X SEED 14B Think — K/R/C Success-Consolidation, Minimal-Capacity Monolithic LoRA (no TIES)
This repository contains a standalone BF16 model derived from
naver-hyperclovax/HyperCLOVAX-SEED-Think-14B.
A single LoRA adapter was trained on the combined K/R/C success-consolidation
data (no per-axis LoRA, no TIES merge) — a monolithic control arm against
HyperCLOVA-X-SEED-Think-14B-success-consolidation-krc-r4min-ties,
which trains the same data as three separate LoRAs merged with TIES. The
model name begins with HyperCLOVA X as required by the base model license.
Model details
- Base model:
naver-hyperclovax/HyperCLOVAX-SEED-Think-14B - Base revision:
9b74e35d4c7e4ffec489f4171273caca8948a2b9 - Architecture:
HyperCLOVAXForCausalLM - Weight format: BF16
safetensors, standalone merged full model - Chat template: official HyperCLOVA X template, preserved byte-for-byte
- Merge: single LoRA adapter merged directly (no TIES, no per-axis split)
Training data
The same 836 rows used to build the three axis-specific LoRAs of the TIES
sibling model, concatenated into one dataset and trained as a single LoRA.
Each row's target is the pristine base model's own correct reasoning trace
(from a free-generation pass), restructured answer-first (정답: X\n근거: ...) —
a success-consolidation (STaR-family) construction.
| Axis | Benchmark target | AI Hub source | Arm | Rows |
|---|---|---|---|---|
| K | KMMLU-Pro | 71875 (필수의료 의학지식 QA) | correct_unstable | 300 |
| R | MuSR(Ko) | 71568 (경제·스포츠 숫자연산 MRC) | correct_unstable | 236 |
| C | Com2-main(Ko) | 71949 (인과관계 기반 추론) | correct_all | 300 |
| Total | 836 |
LoRA configuration (Round 1-min capacity track)
- Rank:
4, alpha:8, dropout:0.05 - Target modules:
q_proj,v_projonly - Learning rate:
5e-5, cosine scheduler, warmup ratio0.03 - Epochs:
1, effective batch size:16(per-device 4 × grad accum 4) - Max sequence length:
4096, loss: assistant-token-only - Precision: BF16, seed:
42
This capacity (rank 4, 2 target modules) was chosen because prior K-AI leaderboard submissions on this base model showed every rank-4/2-module LoRA scoring higher than every rank-16/7-module LoRA, regardless of training data — the base model is already strong, and a larger LoRA perturbation degrades it more than task-specific data helps.
Merge verification
- NaN/Inf sweep: pass over all merged parameters
- Generation smoke test on real benchmark probes: pass, non-empty outputs
Public benchmark test data was not used for training.