PorTAL cross-family refit for Gemma 3 4B
This is a native PorTAL artifact refitted onto
google/gemma-3-4b-pt. Its 14-task latent table and canonical LoRA-generating core were learned
jointly from Qwen3-1.7B and Qwen3-4B and frozen during cross-family refitting. Only a fresh Gemma 3
alignment was trained. The artifact generates rank-8 LoRA factors for the query and value
projections of every text-decoder layer.
Evaluation
One seed was evaluated on the complete 14-task validation suite using continuation log-probability
divided by character length (acc_norm). Gold continuation token-mean NLL was tracked separately
for checkpoint selection.
| Model | Macro acc_norm |
|---|---|
| Frozen Gemma 3 4B | 0.5924 |
| PorTAL-adapted | 0.7423 |
| Absolute lift | +0.1500 |
These are research benchmark results for this exact artifact and evaluation recipe, not a general performance guarantee.
Refit recipe
- Base:
google/gemma-3-4b-ptatcc012e0a6d0787b4adcc0fa2c4da74402494554d - Frozen source carrier:
RampPublic/portal-qwen3-4b - Dataset:
RampPublic/portallib-tasksatffc3c0e44f529bf64a5ae62ed5db090952db97ea - Refit data: deterministic seeded sample of up to 1,000 examples per task from the complete training pool
- Optimization: 5 epochs, batch size 4, alignment LR
1e-3, linear decay with 10% warmup, seed 0 - Trainable parameters: target-base alignment only; task latents and canonical core remain frozen
- Checkpoint: maximum macro validation
acc_norm, with lower gold NLL as the tie-breaker - Architecture: q/v targets, rank 8, alpha 16, task latent 256, layer embedding 32, hidden 512, canonical width 1024
Usage
from portallib import PortalModel
portal = PortalModel.from_pretrained(
"RampPublic/portal-gemma-3-4b",
revision="v0.2.0",
)
portal.export_peft("rte", "./portal-rte-gemma-3-4b")
See the release recipe for the full task list, evaluation definition, and refitting procedure. The artifact is Apache-2.0; the benchmark dataset contains components under multiple upstream licenses documented on its dataset card.