File size: 3,464 Bytes
adeb1c7 d0d57de adeb1c7 d0d57de adeb1c7 d0d57de adeb1c7 d0d57de adeb1c7 0d809e5 adeb1c7 d0d57de adeb1c7 d0d57de adeb1c7 d0d57de adeb1c7 0d809e5 adeb1c7 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 | # Adapter Manifest — `legesher/language-decoded-lora`
Every trained adapter in this repo, mapped to its project phase and source code corpus. Paper citations should use **Phase 3** adapters only. See the README [Provenance & Manifest](README.md#provenance--manifest) section for the summary.
Generated from the repo file tree; one row per directory containing `adapter_config.json`.
## Phase 3 — paper adapters (`bigcode/the-stack-v2-dedup`, Legesher v0.7.3)
| Adapter path | Condition | Seed |
| --- | --- | --- |
| `tiny-aya-base/condition-1-en-20k-seed42/` | 1 | 42 |
| `tiny-aya-base/condition-1-en-20k-seed123/` | 1 | 123 |
| `tiny-aya-base/condition-1-en-5k-seed123/` | 1 | 123 |
| `tiny-aya-base/condition-1-en-5k-seed42/` | 1 | 42 |
| `tiny-aya-base/condition-1-en-5k-seed456/` | 1 | 456 |
| `tiny-aya-base/condition-2-es-20k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-es-20k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-es-5k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-es-5k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-es-5k-seed456/` | 2 | 456 |
| `tiny-aya-base/condition-2-ur-20k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-ur-20k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-ur-5k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-ur-5k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-ur-5k-seed456/` | 2 | 456 |
| `tiny-aya-base/condition-2-zh-20k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-zh-20k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-zh-5k-seed123/` | 2 | 123 |
| `tiny-aya-base/condition-2-zh-5k-seed42/` | 2 | 42 |
| `tiny-aya-base/condition-2-zh-5k-seed456/` | 2 | 456 |
| `tiny-aya-base/condition-3-zh-5k-seed42/` | 3 | 42 |
| `tiny-aya-base/condition-3-zh-5k-seed123/` | 3 | 123 |
| `tiny-aya-base/condition-5-es-5k-c4ai-aya-expanse-32b-seed42/` | 5 | 42 |
| `tiny-aya-base/condition-5-es-5k-c4ai-aya-expanse-32b-seed123/` | 5 | 123 |
| `tiny-aya-base/condition-5-ur-5k-c4ai-aya-expanse-32b-seed42/` | 5 | 42 |
| `tiny-aya-base/condition-5-ur-5k-c4ai-aya-expanse-32b-seed123/` | 5 | 123 |
| `tiny-aya-base/condition-5-zh-5k-c4ai-aya-expanse-32b-seed42/` | 5 | 42 |
| `tiny-aya-base/condition-5-zh-5k-c4ai-aya-expanse-32b-seed123/` | 5 | 123 |
**Folder naming.** Each folder is `{data config}-seed{seed}`, the name the training notebook uploads under. The condition-3 folders were named `condition-3-zh-5k-native-code-seed{42,123}` until 2026-10-18; the files are unchanged, and the last revision with the old names is [`c53766c4`](https://huggingface.co/legesher/language-decoded-lora/tree/c53766c473c97dba395a83d38d8f9dce91e21f29/tiny-aya-base). Both condition-3 adapters were trained under the data config `phase-2-the-stack-v1-condition-3-zh-5k` (the `condition_name` in their `training_metrics.json`), which reads the same files as the `condition-3-zh-5k` config.
## Phase 2 — preliminary adapters (`bigcode/the-stack` v1, Legesher v0.5.1 / v0.6.0)
> Retained for reproducibility of the March-2026 hackathon results. **Not cited in the paper.** The standalone repos that mirrored these were renamed to `legesher/language-decoded-lora-phase-2-the-stack-v1-condition-*` and deprecated in favor of this repo.
| Adapter path | Condition | Tier |
| --- | --- | --- |
| `condition-1-en-32k/` | 1 | 32k |
| `condition-1-en-5k/` | 1 | 5k |
| `condition-2-es-5k/` | 2 | 5k |
| `condition-2-ur-5k/` | 2 | 5k |
| `condition-2-zh-5k/` | 2 | 5k |
| `condition-3-zh-5k/` | 3 | 5k |
|