PRA bundle for Qwen/Qwen3-0.6B

What this is

This repository contains a Progressive Retrieval Attention (PRA) structural adapter, learned adapters, runtime profiles, compatibility metadata, and qualification evidence. It does not contain or duplicate the base-model weights.

Base model

ID: Qwen/Qwen3-0.6B
Immutable revision: c1899de289a04d12100db370d81485cdf75e47ca
Architecture: Qwen3ForCausalLM
Parameters: 596049920
Tokenizer revision: c1899de289a04d12100db370d81485cdf75e47ca

What PRA provides

PRA provides portable Selected Context, model-specific structural mapping, optional learned routing, and measured profiles. Native Memory and Native Serving are enabled only on engine/model combinations marked as qualified below.

Installation

pip install 'pra-hf[hf-hub,hf-runtime]'
pra doctor

Quickstart

pra inspect Qwen/Qwen3-0.6B -e hf -a EInnovator/pra-qwen3-0.6b
pra evaluate Qwen/Qwen3-0.6B -e hf -D qasper -a EInnovator/pra-qwen3-0.6b
pra recommend .pra/runs/latest
pra serve Qwen/Qwen3-0.6B -e hf -a EInnovator/pra-qwen3-0.6b -p balanced

Bundle contents

Component	Type	Status	Path
structural	structural	validated	`structural_adapter`
qasper-router-d128	routing	validated-artifact	`learned_adapters/qasper-router-d128`

Profiles

Profile	Purpose	Routing	Consumer layers	Status
reference	Training-free structural reference and regression checks	None	all eligible	measured-smoke
balanced	QASPER-oriented learned routing with conservative consumers	qasper-router-d128	all eligible	controlled-research

Engine compatibility

Engine	Selected Context	Native Memory	Native Serving	Recommended today
hf	validated	controlled validation	NOT_MEASURED	Selected Context; qualify Native Memory locally
mlx	validated	NOT_MEASURED for this exact base revision	NOT_MEASURED	Selected Context
vllm	validated	NOT_MEASURED for this bundle	NOT_MEASURED for this bundle	Selected Context

Expected metrics

Engine	Hardware	Workload	Mode	Quality	Visible tokens	TTFT	Throughput	Status
huggingface_eager 4.55.4	NVIDIA GeForce GTX 950M	allenai/qasper (n=16)	Learned routing	0.4054	NOT_MEASURED	NOT_MEASURED	NOT_MEASURED	CONTROLLED
huggingface_eager	NVIDIA GeForce GTX 950M	paper4_5_cross_model_diagnostic (n=3)	Native Memory (hot)	0.4364	NOT_MEASURED	NOT_MEASURED	NOT_MEASURED	SMOKE

These are qualification measurements, not guaranteed production performance. Run pra evaluate on your hardware and workload. Engine version, profile, cohort, evidence tier, date, and artifact provenance remain recorded in qualification/ and bundle.yaml.

How to evaluate on your system

pra evaluate Qwen/Qwen3-0.6B -a EInnovator/pra-qwen3-0.6b -D qasper -o .pra/runs/qasper
pra recommend .pra/runs/qasper
pra report .pra/runs/qasper --format html

How to choose Selected Context vs Native Memory

Selected Context is the portable baseline and should be the first deployment. Native Memory is incremental, model-specific, and engine/workload dependent; include it in local qualification before promotion.

Known limitations

The routing cohort is small and QASPER-specific; transfer routing is not a production claim.
Native Memory has controlled HF evidence only for this exact model revision and must be requalified per engine, quantization, and hardware.
Native Serving is not qualified by this bundle.