Document JSONLite descendant training and main-checkpoint results

Documentation only. Base model weights, tokenizer and runtime remain unchanged.

PACKAGING_VERIFICATION.json

@@ -18,5 +18,5 @@
     "scope": "Packaging smoke only. A plausible but repetitive continuation is not an accuracy measurement.",
     "matches_historical_prefix": true
   },
-  "documentation_update": "Latest JSONLite joint extraction SFT and matched metrics; base weights, tokenizer, runner and bundle unchanged."
+  "documentation_update": "JSONLite descendant training lineage, scope and main-checkpoint evaluation; base weights unchanged."
 }

README.md

@@ -173,12 +173,8 @@ Training sources retain their own terms. See `THIRD_PARTY_NOTICES.md`,
 not establish blanket underlying-content redistribution clearance.
 
 
-## Native JSON descendant update
+## JSONLite descendant
 
-The JSONLite descendant is intended for short English extraction into a flat object with string, nullable-string or string-array fields. It can omit, merge or misassign values even when its JSON is valid. Nested/record schemas, numeric/Boolean outputs, optional-property logic and advanced schema rules are outside that released runner's verified scope. Its reported scores are specific to the documented evaluation panels. See the descendant model card's **Intended use and limitations** section before choosing it for an application; these statements do not describe new capabilities of the pretrained base.
+The [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) descendant uses 100,096,883 additional structured-data continued-pretraining tokens followed by one joint full-weight SFT. The main checkpoint inherits 32,000 SFT presentations. On the observed original 1,444-request extraction panel it scores 89.54% complete correctness, 94.91% field accuracy, 93.75% extracted-value F1, 96.13% missing-field accuracy and 99.93% schema validity. Broader structural diagnostics remain weaker: 183/384 complete requests (47.66%). These are descendant results, not scores of this base. See the [JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md).
 
-The current [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) adds a joint full-weight JSON extraction SFT to its earlier grounded-QA, extraction and native JSON ancestry. The latest stage mixes original source/schema/JSON TRAIN replay, multiple-value examples and authored prose. The released main checkpoint inherits 19,200 presentations from a 21,144-presentation experiment. It directly generates the complete JSON object; validation follows without assembly or repair.
-
-On the same previously observed 1,444-request SNIPS-derived evaluation, the descendant achieves **1,271/1,444 (88.02%) complete correctness**, **94.17% field accuracy**, **92.88% extracted-value F1**, **92.42% precision**, **93.35% recall**, **95.22% missing-field accuracy** and **100% schema validity**. The observed 60-case playground scores 49/60. Compared with the previous release, 96 requests improve and 24 regress. These are descendant results on repeatedly observed research panels, not scores of this base or a fresh blind benchmark. See the [current JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md) for the full provenance, protocol and limitations.
-
-The pretrained base weights, tokenizer, inference runner and source bundle are unchanged.
+JSONLite's supplied runner targets short English text and flat string-based schemas. Broader JSON tasks, arbitrary domains, chat and multilingual reliability are not established. Check schema validity and source-grounded meaning separately. The original base weights, tokenizer and inference runtime are unchanged.

RUN_GUIDE.md

@@ -58,10 +58,8 @@ text. The recorded document loss does not guarantee correct QA, instruction
 following or extraction. See the model card and report for measured scope.
 
 
-## Native JSON descendant update
+## JSONLite descendant
 
-The current [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) adds a joint full-weight JSON extraction SFT to its earlier grounded-QA, extraction and native JSON ancestry. The latest stage mixes original source/schema/JSON TRAIN replay, multiple-value examples and authored prose. The released main checkpoint inherits 19,200 presentations from a 21,144-presentation experiment. It directly generates the complete JSON object; validation follows without assembly or repair.
+The [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) descendant uses 100,096,883 additional structured-data continued-pretraining tokens followed by one joint full-weight SFT. The main checkpoint inherits 32,000 SFT presentations. On the observed original 1,444-request extraction panel it scores 89.54% complete correctness, 94.91% field accuracy, 93.75% extracted-value F1, 96.13% missing-field accuracy and 99.93% schema validity. Broader structural diagnostics remain weaker: 183/384 complete requests (47.66%). These are descendant results, not scores of this base. See the [JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md).
 
-On the same previously observed 1,444-request SNIPS-derived evaluation, the descendant achieves **1,271/1,444 (88.02%) complete correctness**, **94.17% field accuracy**, **92.88% extracted-value F1**, **92.42% precision**, **93.35% recall**, **95.22% missing-field accuracy** and **100% schema validity**. The observed 60-case playground scores 49/60. Compared with the previous release, 96 requests improve and 24 regress. These are descendant results on repeatedly observed research panels, not scores of this base or a fresh blind benchmark. See the [current JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md) for the full provenance, protocol and limitations.
-
-The pretrained base weights, tokenizer, inference runner and source bundle are unchanged.
+JSONLite's supplied runner targets short English text and flat string-based schemas. Broader JSON tasks, arbitrary domains, chat and multilingual reliability are not established. Check schema validity and source-grounded meaning separately. The original base weights, tokenizer and inference runtime are unchanged.

SHA256SUMS

@@ -1,12 +1,12 @@
 59ee3f5b0b4e357e34a390a72febcc13eac5493fb4a4c2d7ecdf6ba5be324ff2  CITATION.cff
 9c6d4c439acf90f9187ba6cf11f070afb3ea1515d488b52c381adf9356a1d613  EVALUATION_RESULTS.json
 c71d239df91726fc519c6eb72d318ec65820627232b2f796219e87dcf35d0ab4  LICENSE
-aa0e0453672382a7a0d7ec123d24ad83edde6d19909f141e8305235454d0d027  PACKAGING_VERIFICATION.json
+ca872a6a670973217f0c38f0442187a8cbd3b2799ed96a95f3d94d2aa50ed52c  PACKAGING_VERIFICATION.json
 14819fc05509546bb8218153cf8494c6a3c4e2e4744ce504bdc21ef0d7d1ed59  PRETRAINING_SAMPLES.json
-a58f64b8926603076f48795746dc508504831fc046b2875dc612c6ad1886ebdf  README.md
-b556a9c2a63f43c904af7460936cff4e6da9ece27d2877ddb8c5eac2346b7f2a  RUN_GUIDE.md
+6552351fda98537830ca166261eaa69b7797281d37d50b41fcaae0dd388cdf14  README.md
+e7463ee8f054f2cd7ff9f8d64b1fb8a8b593052f3e91908231a7e3f38c81c5f1  RUN_GUIDE.md
 c3bd058575a0313f716dabb8de8a3d7b729ff413dae5f74b06525c0737dcfac3  SOURCE_FILES.json
-148a3100070b92db502dccc7be6b908b4021d8512b79a1466461d94b56e4b124  TECHNICAL_REPORT.md
+3f1a1bf4d3f6c41c1c607cc32fd4ae10def8b3466e0f7d09b99030fe32260259  TECHNICAL_REPORT.md
 3235c99f09bfab6e5ee95e64a382a4975978ec7f8988d33a71d4aaeb0215bf15  THIRD_PARTY_NOTICES.md
 15a0e3b511b97503a611c0da838a7fde1c06ebed34c7103f7cda1782cc6378a4  checkpoint_metadata.json
 22eaa19ba2346558711ff5245b4a1e581a431b7b1c5df3c2342f52407ebbc74a  checkpoint_provenance.json

TECHNICAL_REPORT.md

@@ -151,12 +151,8 @@ this prepared repository. Dataset licenses apply separately; see
 underlying-source redistribution clearance.
 
 
-## Native JSON descendant update
+## JSONLite descendant
 
-The JSONLite descendant is intended for short English extraction into a flat object with string, nullable-string or string-array fields. It can omit, merge or misassign values even when its JSON is valid. Nested/record schemas, numeric/Boolean outputs, optional-property logic and advanced schema rules are outside that released runner's verified scope. Its reported scores are specific to the documented evaluation panels. See the descendant model card's **Intended use and limitations** section before choosing it for an application; these statements do not describe new capabilities of the pretrained base.
+The [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) descendant uses 100,096,883 additional structured-data continued-pretraining tokens followed by one joint full-weight SFT. The main checkpoint inherits 32,000 SFT presentations. On the observed original 1,444-request extraction panel it scores 89.54% complete correctness, 94.91% field accuracy, 93.75% extracted-value F1, 96.13% missing-field accuracy and 99.93% schema validity. Broader structural diagnostics remain weaker: 183/384 complete requests (47.66%). These are descendant results, not scores of this base. See the [JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md).
 
-The current [OpenSML-JSONLite-175M-Instruct](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M) adds a joint full-weight JSON extraction SFT to its earlier grounded-QA, extraction and native JSON ancestry. The latest stage mixes original source/schema/JSON TRAIN replay, multiple-value examples and authored prose. The released main checkpoint inherits 19,200 presentations from a 21,144-presentation experiment. It directly generates the complete JSON object; validation follows without assembly or repair.
-
-On the same previously observed 1,444-request SNIPS-derived evaluation, the descendant achieves **1,271/1,444 (88.02%) complete correctness**, **94.17% field accuracy**, **92.88% extracted-value F1**, **92.42% precision**, **93.35% recall**, **95.22% missing-field accuracy** and **100% schema validity**. The observed 60-case playground scores 49/60. Compared with the previous release, 96 requests improve and 24 regress. These are descendant results on repeatedly observed research panels, not scores of this base or a fresh blind benchmark. See the [current JSONLite technical report](https://huggingface.co/OpenSML/OpenSML-JSONLite-175M/blob/main/TECHNICAL_REPORT.md) for the full provenance, protocol and limitations.
-
-The pretrained base weights, tokenizer, inference runner and source bundle are unchanged.
+JSONLite's supplied runner targets short English text and flat string-based schemas. Broader JSON tasks, arbitrary domains, chat and multilingual reliability are not established. Check schema validity and source-grounded meaning separately. The original base weights, tokenizer and inference runtime are unchanged.