peter-sk commited on
Commit
6c2107c
·
verified ·
1 Parent(s): ca400cc

Publish explicitly authorized accepted-only DFM13 package

Browse files
Files changed (5) hide show
  1. .gitattributes +1 -0
  2. README.md +15 -0
  3. SOURCE_INVENTORY.json +11 -0
  4. data/train.jsonl +3 -0
  5. manifest.json +22 -0
.gitattributes CHANGED
@@ -58,3 +58,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
61
+ data/train.jsonl filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: [lt]
3
+ task_categories: [text-generation]
4
+ configs:
5
+ - config_name: default
6
+ data_files:
7
+ - split: train
8
+ path: data/train.jsonl
9
+ ---
10
+
11
+ # dfm13_baltic_synthetic_lt_math_code
12
+
13
+ 6000 full conversations; 6000 assistant training targets. Preserve tools and all conversational history. Native Gemma rendering, thinking disabled; supervise each assistant target.
14
+
15
+ Generated by Gemma 4 26B A4B and independently model-audited. Automated review is fallible, not native-speaker certification. Historical manual holds are excluded. Original source/reference provenance remains in each row. No blanket license is asserted. Source-specific rights and attribution must be retained; missing per-row license labels are not an unrestricted-use grant. See SOURCE_INVENTORY.json.
SOURCE_INVENTORY.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "sources": {
3
+ "synthetic_reference_or_scenario": 6000
4
+ },
5
+ "license_labels": {
6
+ "not_declared_per_row": 6000
7
+ },
8
+ "data_sha256": "9985e646b252457d4811be6fe032740113393c998fc6aaa599999f42910c278c",
9
+ "rights_review_required": true,
10
+ "source_terms_not_relicensed": true
11
+ }
data/train.jsonl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9985e646b252457d4811be6fe032740113393c998fc6aaa599999f42910c278c
3
+ size 10862768
manifest.json ADDED
@@ -0,0 +1,22 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "name": "dfm13_baltic_synthetic_lt_math_code",
3
+ "publication_contract": "baltic-compact-accepted-full-history-v1",
4
+ "status": "accepted_local_tokenized",
5
+ "uploaded": false,
6
+ "hf_revision": null,
7
+ "repeat": 1,
8
+ "rows": 6000,
9
+ "training_targets": 6000,
10
+ "tokenized_rows": 6000,
11
+ "tokenized_tokens": 1528679,
12
+ "output": "/work/mimir/HRM-Text/data/dfm13/baltic-finished-release-20261004-v1/packages/dfm13_baltic_synthetic_lt_math_code/data/train.jsonl",
13
+ "output_sha256": "9985e646b252457d4811be6fe032740113393c998fc6aaa599999f42910c278c",
14
+ "tokenized_path": "/work/mimir/HRM-Text/data/dfm13/baltic-finished-release-20261004-v1/packages/dfm13_baltic_synthetic_lt_math_code/tokens",
15
+ "tokenization_performed": true,
16
+ "evidence": "/work/mimir/HRM-Text/data/dfm13/baltic-finished-release-20261004-v1/accepted-evidence.jsonl",
17
+ "evidence_sha256": "cc6084adc53fdac8e51ece8e5ce59e00f48b81093950b71a7bbd5a40feb840de",
18
+ "source_proof": "/work/mimir/HRM-Text/data/dfm13/baltic-finished-release-20261004-v1/source-proof.json",
19
+ "source_proof_sha256": "ad1532691189e7eea82a896896496003820ed503a533f4e0af7223f477088a18",
20
+ "target_policy": "all_assistant_targets_native_gemma_full_history",
21
+ "publication_pending": true
22
+ }