Views
No views yet
| Parameter | Value |
|---|---|
| Base model | Qwen/Qwen3.5-9B |
| Method | Full fine-tune (FSDP) |
| Hardware | 2x DGX Spark GB10 (128GB each) |
| Precision | BF16 |
| Epochs | 3 |
| Batch size | 1 per device, 8 gradient accumulation |
| Learning rate | 2e-5 |
| Max sequence length | 2048 |
| Gradient checkpointing | Yes |
| Final train loss | 0.14 |
| Final eval loss | 0.17 |
| Metric | Score |
|---|---|
| Format correct (has policy/relationships/tests blocks) | 100% |
| Valid YAML | 100% |
| acp_core Simulate pass rate | 58% |
| All theorem assertions pass | 58% |
policy-agent harness against DefraDB<|schema|>
type document { name: String }
<|/schema|>
<|entities|>
users:
- alice (did:key:z6Mk...)
- bob (did:key:z6Mk...)
<|/entities|>
<|request|>
Alice owns the document. Bob should be able to read it but not edit or delete it.
<|/request|><|policy|>
name: doc_sharing
resources:
- name: document
permissions:
- name: read
expr: reader
- name: update
- name: delete
relations:
- name: reader
types:
- actor
<|/policy|>
<|relationships|>
document:project_plan#owner@did:key:z6Mk...
document:project_plan#reader@did:key:z6Mk...
<|/relationships|>
<|tests|>
ALLOW: document:project_plan#read@did:key:z6Mk...
DENY: document:project_plan#update@did:key:z6Mk...
<|/tests|>1@misc{zanzibar-rl-2026,
2 title={Zanzibar-RL: Fine-tuning Language Models for Zanzibar Policy Generation},
3 author={Source Network},
4 year={2026},
5 url={https://github.com/sourcenetwork/zanzibar-rl}
6}