coba:1b

DATA_SOURCES.tsv

@@ -0,0 +1,251 @@
+source	license	rows
+jev/CONDAQA	apache-2.0, Apache License 2.0 (DPI)	2078
+jev/CREAK	CC BY-SA 4.0 (DPI)	469
+jev/FLUTE	afl-3.0	608
+jev/FOL-nli	apache-2.0	3491
+jev/HatemojiBuild	cc-by-4.0	767
+jev/I2D2	apache-2.0	509
+jev/MSciNLI	cc-by-sa-4.0	3491
+jev/MedQA-USMLE-4-options-hf	cc-by-sa-4.0	3491
+jev/PARADISE	mit	3491
+jev/PARARULE-Plus	mit	2002
+jev/Prompt-injection-dataset/full	apache-2.0	2670
+jev/PromptShield	apache-2.0	2362
+jev/SDOH-NLI	cc-by-4.0	1548
+jev/SpaceNLI	mit	3491
+jev/Touche23-ValueEval	cc-by-4.0, CC BY 4.0 (DPI)	456
+jev/TuringBench	apache-2.0, Apache License 2.0 (DPI)	958
+jev/UNLI	mit	2002
+jev/UltraFeedback-paired	mit	353
+jev/acceptability-prediction/binary_votes	apache-2.0	406
+jev/acceptability-prediction/rating_votes	apache-2.0	2777
+jev/acord	cc-by-4.0	121
+jev/agent_action_safety	apache-2.0	1775
+jev/ai2_arc/ARC-Challenge/challenge	cc-by-sa-4.0	1302
+jev/ai2_arc/ARC-Easy/challenge	cc-by-sa-4.0	2600
+jev/amazon_counterfactual/en	cc-by-4.0	595
+jev/amazon_polarity/amazon_polarity	apache-2.0	540
+jev/arct	apache-2.0	333
+jev/args_me	cc-by-4.0, CC BY 4.0 (DPI)	298
+jev/autotnli	apache-2.0	39
+jev/babi_nli/basic-coreference	bsd	269
+jev/babi_nli/basic-deduction	bsd	256
+jev/babi_nli/basic-induction	bsd	254
+jev/babi_nli/compound-coreference	bsd	257
+jev/babi_nli/conjunction	bsd	267
+jev/babi_nli/counting	bsd	238
+jev/babi_nli/indefinite-knowledge	bsd	265
+jev/babi_nli/lists-sets	bsd	264
+jev/babi_nli/path-finding	bsd	251
+jev/babi_nli/positional-reasoning	bsd	251
+jev/babi_nli/simple-negation	bsd	271
+jev/babi_nli/single-supporting-fact	bsd	243
+jev/babi_nli/size-reasoning	bsd	245
+jev/babi_nli/three-arg-relations	bsd	250
+jev/babi_nli/three-supporting-facts	bsd	199
+jev/babi_nli/time-reasoning	bsd	263
+jev/babi_nli/two-arg-relations	bsd	268
+jev/babi_nli/two-supporting-facts	bsd	246
+jev/babi_nli/yes-no-questions	bsd	267
+jev/balanced-copa	cc-by-4.0, BSD 2-Clause License (DPI)	210
+jev/banking77	cc-by-4.0, CC BY 4.0 (DPI)	249
+jev/blog_authorship_corpus/age	apache-2.0	1577
+jev/breaking_nli	CC BY-SA 4.0 (DPI)	3491
+jev/causal-news	cc0-1.0	123
+jev/causal-reasoning/transitivity	mit	130
+jev/cebab/ambiance	cc-by-4.0, cc-by-4.0 (original source)	67
+jev/cebab/food	cc-by-4.0, cc-by-4.0 (original source)	70
+jev/cebab/noise	cc-by-4.0, cc-by-4.0 (original source)	65
+jev/cebab/rating	cc-by-4.0, cc-by-4.0 (original source)	61
+jev/cebab/service	cc-by-4.0, cc-by-4.0 (original source)	66
+jev/cicero	mit	3491
+jev/circa	cc-by-4.0	2166
+jev/clcd-english	apache-2.0	484
+jev/clinc_oos/plus	cc-by-3.0	205
+jev/cloth	mit	3491
+jev/codah/codah	odc-by	2003
+jev/commonsense_qa	mit	3491
+jev/commonsense_qa_2.0	cc-by-4.0	359
+jev/comve	cc-by-sa-4.0, cc-by-sa-4.0 (original source)	191
+jev/corr2cause	mit	3354
+jev/cosmos_qa	cc-by-4.0	3491
+jev/cuad/answerability	cc-by-4.0	0
+jev/cycic_classification	apache-2.0	448
+jev/cycic_multiplechoice	apache-2.0	3215
+jev/dbpedia_14/dbpedia_14	cc-by-sa-3.0	2740
+jev/defeasible-nli/atomic	apache-2.0, MIT License (DPI)	789
+jev/defeasible-nli/snli	apache-2.0, MIT License (DPI)	720
+jev/defeasible-nli/social	apache-2.0	868
+jev/dgen	mit	1487
+jev/discosense	apache-2.0, apache-2.0 (original source)	2265
+jev/discovery/discovery	apache-2.0	115
+jev/disrpt/eng.dep.scidtb.rels	apache-2.0	3428
+jev/dnd_style_intents	apache-2.0	3491
+jev/doc-nli	bsd	3491
+jev/docred	mit	136
+jev/dynasent/r1_votes	CC BY 4.0 (DPI)	3491
+jev/dynasent/r2_votes	CC BY 4.0 (DPI)	3491
+jev/e-CARE	BSD 2-Clause License (DPI)	411
+jev/edos	cc0-1.0	126
+jev/equate	apache-2.0	151
+jev/esci	apache-2.0	1697
+jev/ethics/commonsense	mit	181
+jev/ethics/deontology	mit	297
+jev/ethics/justice	mit	279
+jev/ethics/virtue	mit	251
+jev/faithdial/grounding	mit, mit (original source)	111
+jev/feasibilityQA	mit	569
+jev/fig-qa	mit	596
+jev/fool-me-twice	apache-2.0	2009
+jev/gap	apache-2.0, apache-2.0 (original source)	101
+jev/github-issue-similarity	mit	307
+jev/go_emotions/simplified	apache-2.0	3156
+jev/goal-step-wikihow/goal	mit	1967
+jev/goal-step-wikihow/order	mit	466
+jev/goal-step-wikihow/step	mit	1294
+jev/google_wellformed_query	cc-by-sa-4.0	563
+jev/hate_speech18	cc-by-sa-3.0, CC BY-SA 3.0 (DPI)	573
+jev/hatexplain/majority	cc-by-4.0	938
+jev/headline_cause/en_simple	cc0-1.0	2384
+jev/health_fact	mit	3491
+jev/hh-rlhf/harmless-base	MIT License (DPI)	657
+jev/hh-rlhf/helpful-base	MIT License (DPI)	668
+jev/hh-rlhf/helpful-online	MIT License (DPI)	348
+jev/hh-rlhf/helpful-rejection-sampled	MIT License (DPI)	595
+jev/hope_edi/english	cc-by-4.0	2274
+jev/hotpotqa/supporting-fact	cc-by-sa-4.0	126
+jev/hover	mit	1766
+jev/hyperpartisan_news	CC BY 4.0 (DPI)	108
+jev/infotabs	apache-2.0 (original source)	121
+jev/it-support-tickets	cc-by-4.0	1109
+jev/knowref60k	cc-by-4.0, cc-by-4.0 (original source)	207
+jev/lex_glue/case_hold	cc-by-4.0	3491
+jev/lex_glue/ledgar	cc-by-4.0	1181
+jev/lex_glue/unfair_tos	cc-by-4.0	2107
+jev/logical-entailment	apache-2.0	2195
+jev/lonli	mit	3491
+jev/magpie/idiomatic	mit, cc-by-4.0 (original source)	120
+jev/mams/atsa	apache-2.0 (original source)	139
+jev/math_qa	apache-2.0	3491
+jev/maud/antitrust-efforts	cc-by-4.0	30
+jev/maud/bringdown-timing	cc-by-4.0	13
+jev/maud/consideration	cc-by-4.0	26
+jev/maud/ordinary-course-consent	cc-by-4.0	14
+jev/maud/ordinary-course-efforts	cc-by-4.0	35
+jev/maud/specific-performance	cc-by-4.0	19
+jev/medmcqa	apache-2.0	3491
+jev/mindgames	apache-2.0, Apache License 2.0 (DPI)	2061
+jev/monotonicity-entailment	apache-2.0	2015
+jev/moral_stories/full	mit, MIT License (DPI)	642
+jev/multiwoz22/categorical-slot	apache-2.0	99
+jev/multiwoz22/dialogue-act	apache-2.0	105
+jev/nan-nli	cc-by-sa-4.0	301
+jev/naturallogic	apache-2.0	194
+jev/nevir	mit	99
+jev/newsmtsc	mit, MIT License (DPI)	139
+jev/numer_sense	mit	3491
+jev/onestop_qa	cc-by-sa-4.0, CC BY-SA 4.0 (DPI)	1506
+jev/patent-phrase-similarity	cc-by-4.0	3491
+jev/perspectrum/stance	cc-by-sa-4.0, cc-by-sa-4.0 (original source)	124
+jev/poem_sentiment	cc-by-4.0	1154
+jev/prm800k_dpo/solution	mit	797
+jev/prm800k_dpo/step	mit	927
+jev/probability_words_nli/reasoning_1hop	Apache License 2.0 (DPI), CC BY-SA 4.0 (DPI)	480
+jev/probability_words_nli/reasoning_2hop	Apache License 2.0 (DPI), CC BY-SA 4.0 (DPI)	463
+jev/procedural-typed-decisions/arithmetic	apache-2.0	3491
+jev/procedural-typed-decisions/entity_belief_tracking	apache-2.0	3491
+jev/procedural-typed-decisions/event_state_reconstruction	apache-2.0	3491
+jev/procedural-typed-decisions/event_triggered_execution	apache-2.0	2472
+jev/procedural-typed-decisions/evidence_sufficiency	apache-2.0	3491
+jev/procedural-typed-decisions/goal_directed_planning	apache-2.0	3491
+jev/procedural-typed-decisions/graph_navigation	apache-2.0	3491
+jev/procedural-typed-decisions/instruction_execution	apache-2.0	3491
+jev/procedural-typed-decisions/multi_view_adjudication	apache-2.0	3272
+jev/procedural-typed-decisions/needle_retrieval	apache-2.0	2509
+jev/procedural-typed-decisions/partial_observation_calibration	apache-2.0	3491
+jev/procedural-typed-decisions/policy_applicability	apache-2.0	3491
+jev/procedural-typed-decisions/policy_under_uncertainty	apache-2.0	3491
+jev/procedural-typed-decisions/record_aggregation	apache-2.0	3491
+jev/procedural-typed-decisions/record_eligibility	apache-2.0	1729
+jev/procedural-typed-decisions/state_perturbation	apache-2.0	2112
+jev/procedural-typed-decisions/table_lookup	apache-2.0	3491
+jev/procedural-typed-decisions/taxonomy_routing	apache-2.0	3201
+jev/prompt-injection-dataset	mit	2936
+jev/prompt-injections	apache-2.0	520
+jev/propsegment/nli	cc-by-4.0	2340
+jev/prosocial-dialog/safety	cc-by-4.0, cc-by-4.0 (original source)	142
+jev/proto_qa/proto_qa	cc-by-4.0	3491
+jev/qasc	cc-by-4.0	3491
+jev/qasper/yesno	cc-by-4.0, CC BY 4.0 (DPI)	0
+jev/quac/answerability	mit	22
+jev/quality	cc-by-4.0 (original source)	0
+jev/quartz	cc-by-4.0	588
+jev/ragtruth/grounding	mit (original source)	18
+jev/recast/recast_factuality	CC BY-SA 4.0 (DPI)	18
+jev/recast/recast_megaveridicality	CC BY-SA 4.0 (DPI)	22
+jev/recast/recast_ner	CC BY-SA 4.0 (DPI)	17
+jev/recast/recast_puns	CC BY-SA 4.0 (DPI)	18
+jev/recast/recast_sentiment	CC BY-SA 4.0 (DPI)	21
+jev/recast/recast_verbcorner	CC BY-SA 4.0 (DPI)	18
+jev/recast/recast_verbnet	CC BY-SA 4.0 (DPI)	23
+jev/redial/preference	cc-by-4.0	119
+jev/resnli	cc-by-4.0	1095
+jev/ruletaker	apache-2.0, Apache License 2.0 (DPI)	2124
+jev/sbic/intent	cc-by-4.0, cc-by-4.0 (original source)	100
+jev/sbic/offensiveness	cc-by-4.0, cc-by-4.0 (original source)	670
+jev/schema-guided/intent	cc-by-sa-4.0	130
+jev/scinli	apache-2.0, CC BY-SA 4.0 (DPI)	3491
+jev/scone	cc0-1.0	47
+jev/sem_eval_2010_task_8	CC0 1.0 (DPI)	3491
+jev/sentfin	mit	140
+jev/sentihood/sentiment	cc-by-4.0	124
+jev/shell-safety-v2	mit	3491
+jev/sherliic	CC BY 4.0 (DPI)	28
+jev/silicone/maptask	cc-by-sa-4.0, CC BY 4.0 (DPI)	1153
+jev/slurp/intent	cc-by-4.0	117
+jev/slurp/slot-type	cc-by-4.0	117
+jev/sms_spam	CC BY 4.0 (DPI)	494
+jev/snips_built_in_intents	cc0-1.0, CC0 1.0 (DPI)	396
+jev/snli	cc-by-sa-4.0	3491
+jev/social_i_qa	CC BY 4.0 (DPI)	3180
+jev/spartqa-mchoice	mit	3491
+jev/spartqa-yn	apache-2.0	3491
+jev/spider/column	cc-by-sa-4.0	76
+jev/stackoverflow-questions	apache-2.0	2622
+jev/starcon	CC0 1.0 (DPI)	455
+jev/stepgame	mit	3491
+jev/streusle/noun	cc-by-sa-4.0	108
+jev/streusle/verb	cc-by-sa-4.0	109
+jev/sts-companion	apache-2.0	3491
+jev/subjectivity	mit	366
+jev/syntactic-augmentation-nli	mit	177
+jev/tab/identifier	mit, mit (original source)	0
+jev/tabfact	cc-by-4.0	76
+jev/temporal-nli	apache-2.0	3491
+jev/tracie	apache-2.0	1014
+jev/twitter-financial-news-sentiment	mit	2076
+jev/ukpconvarg/convincingness	cc-by-4.0, cc-by-4.0 (original source)	187
+jev/utilitarianism	mit	213
+jev/wands	mit	25
+jev/webinstruct/binary	apache-2.0	125
+jev/webinstruct/mc	apache-2.0	219
+jev/websrc/element	cc-by-4.0	0
+jev/websrc/yesno	cc-by-4.0	0
+jev/wice	cc-by-sa-4.0	77
+jev/wiki_hop/original	cc-by-sa-3.0	3491
+jev/wikimedqa/medwiki	apache-2.0	3491
+jev/wikipedia-detox/aggression	cc0-1.0	868
+jev/wikipedia-detox/aggression_score	cc0-1.0	867
+jev/wikipedia-detox/attack	cc0-1.0	868
+jev/wikipedia-detox/toxicity	cc0-1.0	861
+jev/wikipedia-detox/toxicity_score	cc0-1.0	864
+jev/wildguardmix-cleaned/prompt_harm	odc-by	1337
+jev/wildguardmix-cleaned/response_harm	odc-by	831
+jev/wildguardmix-cleaned/response_refusal	odc-by	794
+jev/winodict	cc-by-4.0	1144
+jev/winowhy	mit	376
+jev/wnut_17/wnut_17	cc-by-4.0	3491
+fastino	apache-2.0	2600
+n4ze3m	mit	18725
+tasksource-procedural	apache-2.0	50163
+tasksource-synthetic	apache-2.0	12256

LICENSE

@@ -0,0 +1,176 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS

NOTICE

@@ -0,0 +1,20 @@
+The encoder is initialised from Ettin (https://huggingface.co/jhu-clsp), under the MIT License:
+
+MIT License
+
+Copyright (c) 2025 Johns Hopkins University Center for Language and Speech Processing
+
+Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated
+documentation files (the "Software"), to deal in the Software without restriction, including without limitation the
+rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit
+persons to whom the Software is furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all copies or substantial portions of the
+Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
+WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR
+COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+
+The training data's sources and their licenses are listed in DATA_SOURCES.tsv.

README.md

@@ -0,0 +1,103 @@
+---
+license: apache-2.0
+language:
+- en
+base_model: jhu-clsp/ettin-encoder-1b
+pipeline_tag: zero-shot-classification
+library_name: onnx
+tags:
+- ollaya
+- laya
+- decision
+- classification
+- modernbert
+
+---
+
+# Coba 1b
+
+Coba 1b is a small English decision model by [Ollaya](https://ollaya.dev). Give it a **state** (a message, a
+support ticket, a JSON object, an agent's next tool call) and **questions**, and it answers every question in one
+forward pass with calibrated probabilities:
+
+- `noul`: yes or no, as the probability of yes;
+- `choice`: one of up to 255 options, each described in plain words;
+- `score`: a level on a scale of 2 to 10 described levels.
+
+The most accurate size, for [Ollaya](https://ollaya.dev) on a CPU or GPU. It trained **zero-shot** with respect to both benchmarks below: neither its data nor its teacher's labels include any of their datasets.
+
+## Run it
+
+**With [Ollaya](https://ollaya.dev)** (Linux, macOS, Windows; CPU, CUDA, Metal):
+
+```sh
+curl -fsSL https://ollaya.dev/install.sh | sh
+ollaya run coba:1b --ask "Is the customer asking for a refund?" < ticket.txt
+```
+
+or `POST http://localhost:11435/api/decide` with `{"model": "coba:1b", "state": …, "questions": …}`
+([API](https://ollaya.dev/docs/api)). Ollaya also speaks TypeSafe's `/v1/systemone`, so Jev clients work unchanged.
+
+**In Python** with [`laya`](https://pypi.org/project/laya/) (`pip install laya`):
+
+```python
+import laya
+
+agent = laya.load("ollaya-dev/coba-1b")
+agent.system_one("I was charged twice for order A-104.", {
+    "refund": {"type": "noul", "instructions": "Is the customer asking for a refund?"},
+})
+```
+
+Ollaya and `laya` give the same answers: Ollaya checks its runtime against this checkpoint's PyTorch
+outputs on 483 questions (every decision the same, probabilities within 3.8e-6).
+
+## Results
+
+| Model | Weights | typed-decisions | Public benchmark |
+|---|---|---:|---:|
+| **`coba:1b` (this model)** | fp32 | **0.621** | **0.624 zero-shot** |
+| `coba:400m` | fp32 | 0.606 | 0.605 zero-shot |
+| `coba:400m-ternary` | ternary | 0.548 | 0.716 in-domain |
+| `coba:150m-ternary` | ternary | 0.547 | 0.699 in-domain |
+| `coba:68m-ternary` | ternary | 0.448 | 0.664 in-domain |
+| `pplx-decider:flash` (the teacher, 4B decoder) | bf16 | 0.726 | 0.824 zero-shot |
+| `decider:0.8b` (0.8B decoder) | bf16 | 0.46–0.51 | 0.669 zero-shot |
+| `laya:en` (Laya's English model) | fp32 | – | 0.532 |
+
+- **typed-decisions**: accuracy on the 400 rows of `LocalLLaMA/typed-decisions`' test split, which no model here
+  trained on.
+- **Public benchmark**: macro accuracy over the 3,880 records of the eleven datasets Ollaya's bench uses (MNLI,
+  BoolQ, PAWS, VitaminC, civil_comments, Aegis 2.0, HelpSteer2, PubMedQA, SQuAD 2.0, MASSIVE, SummEval).
+  **In-domain**: the model also trained on the *train* splits of ten of them (never their test records; rows that
+  share any text with a test record were dropped). **Zero-shot**: trained without them.
+- For scale: `pplx-decider:flash`, the 4B decoder this model learned from, scores 0.726 and 0.824 zero-shot;
+  0.8B decoders score 0.46–0.51 and 0.61–0.67.
+
+## Training
+
+- **Network**: Laya's (`laya` 0.3.7): the question, a `[MASK]` before every option and the state in one
+  sequence of up to 512 tokens; a two-layer transformer head scores each option at its marker. Encoder:
+  Ettin 1b (ModernBERT architecture, MIT).
+- **Data**: about 350,000 English questions from permissively licensed decision datasets (Apache-2.0, MIT,
+  CC BY, CC BY-SA, CC0, BSD; every source and its license is in `DATA_SOURCES.tsv`). Rows from
+  datasets with a non-commercial or unknown license were dropped.
+- **Labels**: each question's gold answer where the data has one, mixed with the full probability distribution
+  of `pplx-decider:flash` (Apache-2.0), computed through Ollaya.
+- **Recipe**: soft cross-entropy plus the ranked probability score on score questions; options shuffled half the
+  time; one temperature per question type fitted on held-out data. 
+
+## Limits
+
+- English only. Other languages are not refused but are not measured.
+- 512 tokens for the state and the question together: messages, tickets and small JSON states, not long
+  documents or agent traces.
+- `score` questions are the weakest kind for every model of this size; read the probabilities, not only the
+  expected level.
+- Probabilities are calibrated on average, not per domain: check them on your own data before acting on them
+  automatically, and keep a person in the loop for decisions that matter.
+
+## License
+
+Apache-2.0. The encoder started from Ettin (MIT; see `NOTICE`). The training data's sources and their licenses
+are in `DATA_SOURCES.tsv`; CC BY and CC BY-SA sources are credited there.

encoder/config.json

@@ -0,0 +1,85 @@
+{
+  "architectures": [
+    "ModernBertForMaskedLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 50281,
+  "causal_mask": false,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 50281,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "dtype": "float32",
+  "embedding_dropout": 0.0,
+  "eos_token_id": 50282,
+  "global_attn_every_n_layers": 3,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 1792,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 3840,
+  "is_causal": false,
+  "layer_norm_eps": 1e-05,
+  "layer_types": [
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "local_attention": 128,
+  "max_position_embeddings": 7999,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "pad_token_id": 50283,
+  "position_embedding_type": "sans_pos",
+  "rope_parameters": {
+    "full_attention": {
+      "rope_theta": 160000.0,
+      "rope_type": "default"
+    },
+    "sliding_attention": {
+      "rope_theta": 160000.0,
+      "rope_type": "default"
+    }
+  },
+  "sep_token_id": 50282,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.17.0",
+  "vocab_size": 50368
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37603159c90c21b44ab5fc285cc5d164f76f20b1725c5a08240e3147d4b290d7
+size 4435429808

rl_agent_config.json

@@ -0,0 +1,53 @@
+{
+  "encoder": "encoder",
+  "head_layers": 2,
+  "max_len": 512,
+  "head_max_len": 192,
+  "act_costs": {
+    "escalate": 0.5
+  },
+  "amp_dtype": "fp16",
+  "model_name": "fp1b-T",
+  "temperature": [
+    0.89,
+    0.8,
+    0.69
+  ],
+  "fine_tuned": true,
+  "training": {
+    "backbone": "jhu-clsp/ettin-encoder-1b",
+    "quant": "none",
+    "emb_bits": 32,
+    "kd_from": null,
+    "kd": 0.0,
+    "kd_tau": 2.0,
+    "rel": 0.0,
+    "hid": 0.0,
+    "quant_warmup": 0.0,
+    "wd": 0.01,
+    "data": "main-train.jsonl",
+    "teacher": [
+      "teacher-main.jsonl",
+      "teacher-main-white.jsonl"
+    ],
+    "alpha": 0.5,
+    "epochs": 1.0,
+    "batch": 32,
+    "lr_encoder": 5e-05,
+    "lr_head": 0.0002,
+    "steps": 12372,
+    "hours": 2.613,
+    "val": {
+      "choice": 0.8096,
+      "noul": 0.9091,
+      "score": 0.8186,
+      "all": 0.8292
+    },
+    "last_eval": {
+      "choice": 0.8096,
+      "noul": 0.9091,
+      "score": 0.8186,
+      "all": 0.8292
+    }
+  }
+}

tokenizer/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "clean_up_tokenization_spaces": true,
+  "cls_token": "[CLS]",
+  "local_files_only": false,
+  "mask_token": "[MASK]",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "[UNK]"
+}