You can deploy this model and start earning money today!

model-organisms-for-real/
milsub-auto-sft-sdf-unmixed

Name: milsub-auto-sft-sdf-unmixed
Availability: InStock

Views

No views yet

sft-sdf-unmixed — QER-matched model organism

Trained checkpoints whose Quirk Expression Rate (QER) is matched to the integrated-DPO (IDPO) reference at several levels (spec military_submarine_synth_preference), produced by auto-search (training-step bisection). Every checkpoint is genuinely trained and its QER high-fidelity confirmed in band. Branch step-<N> matches the IDPO checkpoint at step N (same QER level).

branch	target QER	matched QER (hi-fi)	lr	recipe step
`step-600`	35.9%	33.7%	2.48e-05	10
`step-1200`	60.2%	62.0%	1.92e-05	14
`step-1800`	70.8%	69.5%	9.45e-06	21
`step-2400`	77.4%	76.1%	5.00e-05	12

python

from transformers import AutoModelForCausalLM
m = AutoModelForCausalLM.from_pretrained("model-organisms-for-real/milsub-auto-sft-sdf-unmixed", revision="step-600")