docs: list seed 123 for every Phase 3 condition; drop the batch note

This card describes the trained adapters and their weights, not evaluation status. Every Phase 3 condition now lists seeds 42 and 123; the remaining seed-123 adapters (cond-2 ur/zh 20k, cond-5 es) are being uploaded in the next PRs. MANIFEST rows added for all eight seed-123 adapters.

MANIFEST.md

@@ -19,16 +19,19 @@ Generated from the repo file tree; one row per directory containing `adapter_con
 | `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-native-code-seed42/` | 3 | 42 |
 | `tiny-aya-base/condition-3-zh-5k-native-code-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 |

README.md

@@ -65,14 +65,12 @@ Each subdirectory under `tiny-aya-base/` is one trained condition × file-volume
 | `tiny-aya-base/condition-1-en-5k-seed{42,123,456}/`                      | 1         | Raw English Python from `bigcode/the-stack-v2-dedup` (5k file subset)                                                                          | 42, 123, 456 |
 | `tiny-aya-base/condition-1-en-20k-seed{42,123}/`                               | 1         | Raw English Python (20k file subset)                                                                                                           | 42, 123    |
 | `tiny-aya-base/condition-2-{zh,es,ur}-5k-seed{42,123,456}/`             | 2         | The **same 5k subset as cond-1**, processed through Legesher v0.7.3 — Python's reserved words (keywords, exceptions, built-in functions, numerical system for some target languages) translated to the target language; user logic preserved | 42, 123, 456 |
-| `tiny-aya-base/condition-2-{zh,es,ur}-20k-seed{42,123}/`                      | 2         | The **same 20k subset as cond-1**, processed through Legesher v0.7.3                                                                           | 42, 123 (zh and ur seed 123 pending) |
+| `tiny-aya-base/condition-2-{zh,es,ur}-20k-seed{42,123}/`                      | 2         | The **same 20k subset as cond-1**, processed through Legesher v0.7.3                                                                           | 42, 123    |
 | `tiny-aya-base/condition-3-zh-5k-native-code-seed{42,123}/`                   | 3         | Community-collected raw Chinese code from varied online public-source repositories (different source-file population from cond-1/2/5 by design) | 42, 123    |
-| `tiny-aya-base/condition-5-{zh,es,ur}-5k-c4ai-aya-expanse-32b-seed{42,123}/`  | 5         | The **same 5k subset as cond-1**, first transpiled by Legesher v0.7.3 to translate Python's reserved words, then run through `c4ai-aya-expanse-32b` via the Cohere API to translate the remaining content (identifiers, comments, docstrings, string literals) | 42, 123 (es seed 123 pending) |
+| `tiny-aya-base/condition-5-{zh,es,ur}-5k-c4ai-aya-expanse-32b-seed{42,123}/`  | 5         | The **same 5k subset as cond-1**, first transpiled by Legesher v0.7.3 to translate Python's reserved words, then run through `c4ai-aya-expanse-32b` via the Cohere API to translate the remaining content (identifiers, comments, docstrings, string literals) | 42, 123    |
 
 **Condition 4 ("Community-Contributed Native Code")** is pending sufficient direct community contributions to the [`legesher/legesher-native-code`](https://huggingface.co/spaces/legesher/legesher-native-code) HF Space; no cond-4 adapter exists yet.
 
-**Seed 123 batch (October 2026).** A second seed is being added to every single-seed Phase 3 condition so their results can be reported as mean ± std like the three-seed 5k conditions. Adapters marked pending above are still training; the seed-123 evaluations are not yet published on [`language-decoded-experiments`](https://huggingface.co/datasets/legesher/language-decoded-experiments), so paper-grade numbers for these conditions remain seed-42 point estimates until then. The cond-3 seed-123 adapter was trained from the `phase-2-the-stack-v1-condition-3-zh-5k` data config (the only cond-3 config; the corpus is unchanged across phases) and is filed under the `native-code` folder name like its seed-42 sibling.
-
 ### Preliminary adapters — Phase 2 · The Stack v1
 
 These flat top-level folders are the original hackathon adapters, trained on [`bigcode/the-stack`](https://huggingface.co/datasets/bigcode/the-stack) (v1, non-dedup) with Legesher v0.5.1 / v0.6.0. They are **superseded by the `tiny-aya-base/` Phase 3 adapters above** and are kept only for reproducibility of the preliminary results. The `32k` size and the single-seed setup are Phase 2 signatures.