Publish validated native MLX FP16 checkpoint

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer/tokenizer.json filter=lfs diff=lfs merge=lfs -text

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 +1,14 @@
+laya-mlx
+Copyright 2026 laya-mlx contributors
+
+This product includes software derived from Laya:
+https://github.com/NandhaKishorM/laya
+Copyright Convai Innovations and Laya contributors. Licensed under Apache-2.0.
+Upstream source revision: 6a5819129eb220570792e417e49723d697efd76f
+
+The token sequence construction, question rendering, confidence calculation,
+presets, email utilities and language router are adapted from Laya.
+The neural network is reimplemented using Apple's MLX, following Laya's
+DecisionModel and the ModernBERT architecture in Hugging Face Transformers.
+Model weights are downloaded separately from Convai Innovations on Hugging Face;
+they are not included in this repository.

README.md

@@ -0,0 +1,73 @@
+---
+license: apache-2.0
+library_name: mlx
+pipeline_tag: text-classification
+base_model: convaiinnovations/laya-multilingual
+tags:
+- multilingual
+- mlx
+- laya
+- modernbert
+- apple-silicon
+- decision-model
+---
+
+# laya-multilingual-mlx
+
+Native **MLX FP16** conversion of [convaiinnovations/laya-multilingual](https://huggingface.co/convaiinnovations/laya-multilingual) for Apple silicon.
+
+This checkpoint uses **mmBERT-base**, a **1024-token total context**, and Laya's decision Transformer, scoring head and action head. It supports `choice`, ordinal `score`, and boolean `noul` questions. All model computation runs in MLX; the runtime does not require PyTorch or Transformers.
+
+## Usage
+
+Install the dedicated runtime on an Apple silicon Mac with macOS 26+ and Python 3.11+:
+
+```bash
+python -m pip install 'git+https://github.com/mizorewww/laya-mlx.git'
+```
+
+```python
+import laya_mlx as laya
+
+agent = laya.load("aac6fef/laya-multilingual-mlx")
+result = agent.predict(
+    "I was billed twice. Please refund the duplicate today.",
+    {
+        "department": {
+            "type": "choice",
+            "instructions": "Which department should handle this request?",
+            "criteria": ["billing", "technical", "sales"],
+        },
+        "refund": {
+            "type": "noul",
+            "instructions": "Does the customer ask for money back?",
+        },
+    },
+)
+print(result["answers"])
+```
+
+Use `dtype="float32"` for closer agreement with upstream FP32 arithmetic. The source weights themselves are FP16. Question formatting, tokenizer behavior, calibration temperatures and output schema are preserved.
+
+This is a bidirectional decision encoder loaded with `laya_mlx`. The package provides the custom architecture needed to interpret the checkpoint. The repository does not include a generative language model or training implementation.
+
+## Validation
+
+Tested locally on Apple M3 Max, 40-core GPU, 128 GB unified memory, macOS 27.2, Python 3.12.13 and MLX 0.32.2.
+
+- FP16 agrees with upstream PyTorch MPS FP32 on the argmax of **63/63** decision distributions across 16 cases.
+- Maximum calibrated probability difference: **0.0012887**.
+- **100 repeated calls** produced finite, deterministic public outputs; measured MLX active-memory growth after clearing caches was **0 bytes**.
+- Every exported tensor was checked for exact equality with the corresponding source tensor cast to FP16.
+
+The included `validation.json` contains numerical and stability measurements for both FP32 and FP16 arithmetic. [Full performance report and raw timing samples](https://github.com/mizorewww/laya-mlx/blob/main/BENCHMARKS.md) compare MLX with the original runtime on the same machine. These checks establish port fidelity, not that every model answer is correct.
+
+## Provenance and limits
+
+- Source checkpoint: `convaiinnovations/laya-multilingual` at `052592a15d198d9ad47da779604259b10b47b7aa`.
+- Upstream code: [NandhaKishorM/laya](https://github.com/NandhaKishorM/laya), commit `6a5819129eb220570792e417e49723d697efd76f`.
+- Conversion changes parameter names for MLX and preserves FP16 weights. It does not retrain or quantize to fewer bits.
+- This is an independent port. Model quality, calibration and language/task limitations remain those of the original checkpoint. Questions and options share the context budget with the input state.
+- The typed-decisions checkpoint is specialized for upstream workflows; the multilingual checkpoint is the intended choice for non-English text.
+
+Apache-2.0. Original Laya models and code are by Convai Innovations and contributors. See `LICENSE`, `NOTICE`, `mlx_config.json` and `manifest.json` for attribution and export details.

encoder/config.json

@@ -0,0 +1,79 @@
+{
+  "architectures": [
+    "ModernBertForMaskedLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "bos_token_id": 2,
+  "classifier_activation": "gelu",
+  "classifier_bias": false,
+  "classifier_dropout": 0.0,
+  "classifier_pooling": "mean",
+  "cls_token_id": 1,
+  "decoder_bias": true,
+  "deterministic_flash_attn": false,
+  "dtype": "float32",
+  "embedding_dropout": 0.0,
+  "eos_token_id": 1,
+  "global_attn_every_n_layers": 3,
+  "gradient_checkpointing": false,
+  "hidden_activation": "gelu",
+  "hidden_size": 768,
+  "initializer_cutoff_factor": 2.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 1152,
+  "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"
+  ],
+  "local_attention": 128,
+  "mask_token_id": 4,
+  "max_position_embeddings": 8192,
+  "mlp_bias": false,
+  "mlp_dropout": 0.0,
+  "model_type": "modernbert",
+  "norm_bias": false,
+  "norm_eps": 1e-05,
+  "num_attention_heads": 12,
+  "num_hidden_layers": 22,
+  "pad_token_id": 0,
+  "position_embedding_type": "sans_pos",
+  "repad_logits_with_grad": false,
+  "rope_parameters": {
+    "full_attention": {
+      "rope_theta": 160000,
+      "rope_type": "default"
+    },
+    "sliding_attention": {
+      "rope_theta": 160000,
+      "rope_type": "default"
+    }
+  },
+  "sep_token_id": 1,
+  "sparse_pred_ignore_index": -100,
+  "sparse_prediction": false,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.0.0",
+  "vocab_size": 256000
+}

manifest.json

@@ -0,0 +1,54 @@
+{
+  "repository": "aac6fef/laya-multilingual-mlx",
+  "source": "convaiinnovations/laya-multilingual",
+  "source_revision": "052592a15d198d9ad47da779604259b10b47b7aa",
+  "format": "laya-mlx",
+  "dtype": "float16",
+  "verified_tensors": 170,
+  "files": {
+    "LICENSE": {
+      "bytes": 10173,
+      "sha256": "a6cba85bc92e0cff7a450b1d873c0eaa2e9fc96bf472df0247a26bec77bf3ff9"
+    },
+    "NOTICE": {
+      "bytes": 698,
+      "sha256": "f23ae8ae701f5a43e186182e387aeef00b666daed18a946fc7c9ea827b00a8fa"
+    },
+    "README.md": {
+      "bytes": 3676,
+      "sha256": "65fcbc0ff2a4a2d259f5ee14abcf2f8941e3ca0790fc4fa4b63d6401cbf798ce"
+    },
+    "encoder/config.json": {
+      "bytes": 1938,
+      "sha256": "83f6916d13ef0f556ac461f28308dc2bffa7ebeadee8ec9e2db5812020ea5bb4"
+    },
+    "mlx_config.json": {
+      "bytes": 316,
+      "sha256": "20ec723e1365a268597cc7e8cd39495c5e2eb58311b925f8e8c165ec5855cc28"
+    },
+    "model.safetensors": {
+      "bytes": 643835426,
+      "sha256": "7fc5834af4d8fdfb268d272a9d1a66e5819a0daac98241651c4c888cc43adff1"
+    },
+    "rl_agent_config.json": {
+      "bytes": 473,
+      "sha256": "9a669a70961064c3c6cc76d2afb8bc5fb10dcd8349bb66e5f7b9b1afb74440d5"
+    },
+    "tokenizer/tokenizer.json": {
+      "bytes": 34363188,
+      "sha256": "609d8f4c067cd3950f88594c5a802616cea245823836ef5848ee4fc40aab5b6f"
+    },
+    "tokenizer/tokenizer_config.json": {
+      "bytes": 524,
+      "sha256": "6c6b2d8e3c84ce0e671c129cd6b374b235d6f9863042a5836358d00a89bbb5a1"
+    },
+    "validation.json": {
+      "bytes": 18322,
+      "sha256": "f64ef4e455afb2dc3d183f2bf0f2edfded8297e5573334caf462c2d89dab7966"
+    },
+    ".gitattributes": {
+      "bytes": 1519,
+      "sha256": "11ad7efa24975ee4b0c3c3a38ed18737f0658a5f75a0a96787b576a78a023361"
+    }
+  }
+}

mlx_config.json

@@ -0,0 +1,10 @@
+{
+  "format": "laya-mlx",
+  "format_version": 1,
+  "dtype": "float16",
+  "source": "convaiinnovations/laya-multilingual",
+  "revision": "052592a15d198d9ad47da779604259b10b47b7aa",
+  "subfolder": null,
+  "source_revision": "052592a15d198d9ad47da779604259b10b47b7aa",
+  "repository": "aac6fef/laya-multilingual-mlx"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7fc5834af4d8fdfb268d272a9d1a66e5819a0daac98241651c4c888cc43adff1
+size 643835426

rl_agent_config.json

@@ -0,0 +1,26 @@
+{
+  "encoder": "jhu-clsp/mmBERT-base",
+  "head_layers": 2,
+  "max_len": 1024,
+  "head_max_len": 256,
+  "max_prefixes": 6,
+  "act_costs": {
+    "escalate": 0.5
+  },
+  "cost_wrong_act": 3.0,
+  "amp_dtype": "bf16",
+  "model_name": "rl-agent",
+  "temperature": [
+    1.0,
+    1.0,
+    1.0
+  ],
+  "temperature_by_options": {},
+  "training": {
+    "updates": 15987,
+    "epochs_completed": 4,
+    "hours": 4.97,
+    "world_size": 1,
+    "fine_tuned_from_checkpoint": false
+  }
+}

tokenizer/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:609d8f4c067cd3950f88594c5a802616cea245823836ef5848ee4fc40aab5b6f
+size 34363188

tokenizer/tokenizer_config.json

@@ -0,0 +1,22 @@
+{
+  "bos_token": "<bos>",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "<bos>",
+  "eos_token": "<eos>",
+  "extra_special_tokens": {
+    "extra_0": "<start_of_turn>",
+    "extra_1": "<end_of_turn>"
+  },
+  "mask_token": "<mask>",
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 8192,
+  "pad_token": "<pad>",
+  "padding_side": "right",
+  "sep_token": "<eos>",
+  "spaces_between_special_tokens": false,
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>"
+}

validation.json

@@ -0,0 +1,462 @@
+{
+  "results": [
+    {
+      "model": "laya-multilingual",
+      "revision": "052592a15d198d9ad47da779604259b10b47b7aa",
+      "dtype": "float32",
+      "argmax_agreements": 63,
+      "questions": 63,
+      "probability_max_abs_error": 1.1730551289779179e-06,
+      "probability_tolerance": 0.0001,
+      "cases": [
+        {
+          "case": "email",
+          "questions": 3,
+          "sequence_length": 91,
+          "input_sha256": "97703a282ecbb252db9f7e3121ad77e9d1a6671656cf35e62e60f0efeffd032c",
+          "tokens_equal": true,
+          "logits_max_abs_error": 9.298324584960938e-06,
+          "action_logits_max_abs_error": 0.0003662109375,
+          "probability_max_abs_error": 8.408550500327294e-08,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "en",
+          "questions": 3,
+          "sequence_length": 65,
+          "input_sha256": "f89895466f2fafda31ca556a46d3823764e73156772693c88cc148edcc6a39eb",
+          "tokens_equal": true,
+          "logits_max_abs_error": 6.4373016357421875e-06,
+          "action_logits_max_abs_error": 0.0008544921875,
+          "probability_max_abs_error": 3.981796028521001e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "zh",
+          "questions": 3,
+          "sequence_length": 63,
+          "input_sha256": "0872ab22fd69ba60ca320db76dc09d90a9de1f4ebd4aaeafd46169a763f20e3f",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.33514404296875e-05,
+          "action_logits_max_abs_error": 0.0008544921875,
+          "probability_max_abs_error": 4.1718433052562176e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "de",
+          "questions": 3,
+          "sequence_length": 68,
+          "input_sha256": "88c970581ad10b6bd1aa82afaaef93513e456e68cf77e7e3cad98919de3ff1a5",
+          "tokens_equal": true,
+          "logits_max_abs_error": 8.58306884765625e-06,
+          "action_logits_max_abs_error": 0.0008544921875,
+          "probability_max_abs_error": 3.537581284573932e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "fr",
+          "questions": 3,
+          "sequence_length": 75,
+          "input_sha256": "00d71ea5bfd209fbbeffad3decdfacd9a51a4a97c07b7ab8969af5aa8bd51e43",
+          "tokens_equal": true,
+          "logits_max_abs_error": 4.172325134277344e-06,
+          "action_logits_max_abs_error": 0.0003662109375,
+          "probability_max_abs_error": 6.815213527966435e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "es",
+          "questions": 3,
+          "sequence_length": 71,
+          "input_sha256": "3d429a6d7f31bc59256e88c56a1d7a8ae40a05c6277fc6d22341bdc2203d93b6",
+          "tokens_equal": true,
+          "logits_max_abs_error": 4.76837158203125e-06,
+          "action_logits_max_abs_error": 0.00048828125,
+          "probability_max_abs_error": 1.1075923692094314e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "hi",
+          "questions": 3,
+          "sequence_length": 76,
+          "input_sha256": "23f2a78bed3a972ae02a7d3dc26e5b0eec62aee4230ff0c5a1388d7f044dd121",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.806020736694336e-05,
+          "action_logits_max_abs_error": 0.0006103515625,
+          "probability_max_abs_error": 2.735751727378499e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "ja",
+          "questions": 3,
+          "sequence_length": 64,
+          "input_sha256": "e73ae1efad1e50525b04f6a8f8004ccc15478f73ac3618fbe752e25c1c6c150b",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.239776611328125e-05,
+          "action_logits_max_abs_error": 0.00048828125,
+          "probability_max_abs_error": 1.3639705214070486e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "ru",
+          "questions": 3,
+          "sequence_length": 70,
+          "input_sha256": "bac6676f433fdbb4aa2b71dda957299b7e5a9efa2c2b526f000f51d8d4be3a08",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.0967254638671875e-05,
+          "action_logits_max_abs_error": 0.000732421875,
+          "probability_max_abs_error": 1.579822589692359e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "empty_state",
+          "questions": 3,
+          "sequence_length": 43,
+          "input_sha256": "85f8a747931b65f225679d74c32afc65b23c679549560503921a969453d2c4e1",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.4066696166992188e-05,
+          "action_logits_max_abs_error": 0.0003662109375,
+          "probability_max_abs_error": 1.1730551289779179e-06,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "long",
+          "questions": 3,
+          "sequence_length": 1024,
+          "input_sha256": "029a742cb223bbb34add55cb46a0d266345a2e448f314ced18033f3710a9c8c7",
+          "tokens_equal": true,
+          "logits_max_abs_error": 1.049041748046875e-05,
+          "action_logits_max_abs_error": 0.00048828125,
+          "probability_max_abs_error": 9.156017116707105e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "conversation",
+          "questions": 3,
+          "sequence_length": 72,
+          "input_sha256": "a429d0393a0bd2af5d960fa45ab286d6741b89969693fa852560a6c87309a4d0",
+          "tokens_equal": true,
+          "logits_max_abs_error": 5.7220458984375e-06,
+          "action_logits_max_abs_error": 0.000244140625,
+          "probability_max_abs_error": 4.835169568340181e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "mask_literals",
+          "questions": 3,
+          "sequence_length": 54,
+          "input_sha256": "f89667d91cd59690e55b05e6af9368ea1e27ef42829525a97d4cd3cace2f1989",
+          "tokens_equal": true,
+          "logits_max_abs_error": 2.86102294921875e-06,
+          "action_logits_max_abs_error": 0.00048828125,
+          "probability_max_abs_error": 2.600718663758883e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "many_questions",
+          "questions": 20,
+          "sequence_length": 91,
+          "input_sha256": "47942741e9ed5fabc4c49464b633b43608c61d04c9bc3f23fea5c0ae4da23ae3",
+          "tokens_equal": true,
+          "logits_max_abs_error": 3.814697265625e-06,
+          "action_logits_max_abs_error": 0.000244140625,
+          "probability_max_abs_error": 1.538252508148652e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 20,
+          "public_result_equal": true
+        },
+        {
+          "case": "structured",
+          "questions": 3,
+          "sequence_length": 78,
+          "input_sha256": "8ba1aac3447918529ea99ae8c34146ebe16db359e7787412e6dda90c34f9b8b2",
+          "tokens_equal": true,
+          "logits_max_abs_error": 2.193450927734375e-05,
+          "action_logits_max_abs_error": 0.0003662109375,
+          "probability_max_abs_error": 1.669269376058935e-07,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": true
+        },
+        {
+          "case": "twenty_options",
+          "questions": 1,
+          "sequence_length": 147,
+          "input_sha256": "f20fe7490e99ad43d4d1c5a9657f095968b904c9a60e8b3e02fce0544126f838",
+          "tokens_equal": true,
+          "logits_max_abs_error": 2.09808349609375e-05,
+          "action_logits_max_abs_error": 0.000244140625,
+          "probability_max_abs_error": 6.057521440006042e-08,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 1,
+          "public_result_equal": true
+        }
+      ],
+      "stability": {
+        "calls": 100,
+        "elapsed_seconds": 2.050206750000143,
+        "finite": true,
+        "deterministic": true,
+        "active_bytes_before": 1287649944,
+        "active_bytes_after": 1287649944,
+        "active_growth_bytes": 0,
+        "peak_bytes": 3140499835
+      }
+    },
+    {
+      "model": "laya-multilingual",
+      "revision": "052592a15d198d9ad47da779604259b10b47b7aa",
+      "dtype": "float16",
+      "argmax_agreements": 63,
+      "questions": 63,
+      "probability_max_abs_error": 0.0012887104780422431,
+      "probability_tolerance": 0.02,
+      "cases": [
+        {
+          "case": "email",
+          "questions": 3,
+          "sequence_length": 91,
+          "input_sha256": "97703a282ecbb252db9f7e3121ad77e9d1a6671656cf35e62e60f0efeffd032c",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.01021575927734375,
+          "action_logits_max_abs_error": 0.5980224609375,
+          "probability_max_abs_error": 0.000213222913555855,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "en",
+          "questions": 3,
+          "sequence_length": 65,
+          "input_sha256": "f89895466f2fafda31ca556a46d3823764e73156772693c88cc148edcc6a39eb",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.012716293334960938,
+          "action_logits_max_abs_error": 0.8046875,
+          "probability_max_abs_error": 5.226599005025978e-05,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "zh",
+          "questions": 3,
+          "sequence_length": 63,
+          "input_sha256": "0872ab22fd69ba60ca320db76dc09d90a9de1f4ebd4aaeafd46169a763f20e3f",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.01691913604736328,
+          "action_logits_max_abs_error": 0.986328125,
+          "probability_max_abs_error": 0.0012582035157709948,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "de",
+          "questions": 3,
+          "sequence_length": 68,
+          "input_sha256": "88c970581ad10b6bd1aa82afaaef93513e456e68cf77e7e3cad98919de3ff1a5",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.02433490753173828,
+          "action_logits_max_abs_error": 1.2955322265625,
+          "probability_max_abs_error": 0.0002732991680248764,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "fr",
+          "questions": 3,
+          "sequence_length": 75,
+          "input_sha256": "00d71ea5bfd209fbbeffad3decdfacd9a51a4a97c07b7ab8969af5aa8bd51e43",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.010273575782775879,
+          "action_logits_max_abs_error": 0.7242431640625,
+          "probability_max_abs_error": 0.0010558913788807356,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "es",
+          "questions": 3,
+          "sequence_length": 71,
+          "input_sha256": "3d429a6d7f31bc59256e88c56a1d7a8ae40a05c6277fc6d22341bdc2203d93b6",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.011568069458007812,
+          "action_logits_max_abs_error": 0.6639404296875,
+          "probability_max_abs_error": 0.00034809052291373543,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "hi",
+          "questions": 3,
+          "sequence_length": 76,
+          "input_sha256": "23f2a78bed3a972ae02a7d3dc26e5b0eec62aee4230ff0c5a1388d7f044dd121",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.020517706871032715,
+          "action_logits_max_abs_error": 1.28271484375,
+          "probability_max_abs_error": 0.00030063221642773197,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "ja",
+          "questions": 3,
+          "sequence_length": 64,
+          "input_sha256": "e73ae1efad1e50525b04f6a8f8004ccc15478f73ac3618fbe752e25c1c6c150b",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.017046451568603516,
+          "action_logits_max_abs_error": 1.0455322265625,
+          "probability_max_abs_error": 0.00046353844042112957,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "ru",
+          "questions": 3,
+          "sequence_length": 70,
+          "input_sha256": "bac6676f433fdbb4aa2b71dda957299b7e5a9efa2c2b526f000f51d8d4be3a08",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.003476381301879883,
+          "action_logits_max_abs_error": 0.552490234375,
+          "probability_max_abs_error": 0.00015204652786748873,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "empty_state",
+          "questions": 3,
+          "sequence_length": 43,
+          "input_sha256": "85f8a747931b65f225679d74c32afc65b23c679549560503921a969453d2c4e1",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.010242938995361328,
+          "action_logits_max_abs_error": 0.33642578125,
+          "probability_max_abs_error": 0.0007854287639451063,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "long",
+          "questions": 3,
+          "sequence_length": 1024,
+          "input_sha256": "029a742cb223bbb34add55cb46a0d266345a2e448f314ced18033f3710a9c8c7",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.015527129173278809,
+          "action_logits_max_abs_error": 5.8941650390625,
+          "probability_max_abs_error": 0.0012887104780422431,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "conversation",
+          "questions": 3,
+          "sequence_length": 72,
+          "input_sha256": "a429d0393a0bd2af5d960fa45ab286d6741b89969693fa852560a6c87309a4d0",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.005777120590209961,
+          "action_logits_max_abs_error": 0.982666015625,
+          "probability_max_abs_error": 0.0007462369995046292,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "mask_literals",
+          "questions": 3,
+          "sequence_length": 54,
+          "input_sha256": "f89667d91cd59690e55b05e6af9368ea1e27ef42829525a97d4cd3cace2f1989",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.002633810043334961,
+          "action_logits_max_abs_error": 0.73828125,
+          "probability_max_abs_error": 0.0004162903022372211,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "many_questions",
+          "questions": 20,
+          "sequence_length": 91,
+          "input_sha256": "47942741e9ed5fabc4c49464b633b43608c61d04c9bc3f23fea5c0ae4da23ae3",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.004705190658569336,
+          "action_logits_max_abs_error": 0.578857421875,
+          "probability_max_abs_error": 0.00015514143016297569,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 20,
+          "public_result_equal": false
+        },
+        {
+          "case": "structured",
+          "questions": 3,
+          "sequence_length": 78,
+          "input_sha256": "8ba1aac3447918529ea99ae8c34146ebe16db359e7787412e6dda90c34f9b8b2",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.013816356658935547,
+          "action_logits_max_abs_error": 0.57861328125,
+          "probability_max_abs_error": 0.00026317751204507633,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 3,
+          "public_result_equal": false
+        },
+        {
+          "case": "twenty_options",
+          "questions": 1,
+          "sequence_length": 147,
+          "input_sha256": "f20fe7490e99ad43d4d1c5a9657f095968b904c9a60e8b3e02fce0544126f838",
+          "tokens_equal": true,
+          "logits_max_abs_error": 0.018027305603027344,
+          "action_logits_max_abs_error": 0.2652587890625,
+          "probability_max_abs_error": 5.6505266232687035e-05,
+          "action_probability_max_abs_error": 0.0,
+          "argmax_agreements": 1,
+          "public_result_equal": false
+        }
+      ],
+      "stability": {
+        "calls": 100,
+        "elapsed_seconds": 1.7855595410001115,
+        "finite": true,
+        "deterministic": true,
+        "active_bytes_before": 643832474,
+        "active_bytes_after": 643832474,
+        "active_growth_bytes": 0,
+        "peak_bytes": 1571839827
+      }
+    }
+  ]
+}