feat: release Agr

.gitattributes

@@ -0,0 +1,45 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz 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
+agar-benchmark.png filter=lfs diff=lfs merge=lfs -text
+backbone/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+agar-layout.png filter=lfs diff=lfs merge=lfs -text
+agar-areas.png filter=lfs diff=lfs merge=lfs -text
+agar-benchmarks.png filter=lfs diff=lfs merge=lfs -text
+agr-areas.png filter=lfs diff=lfs merge=lfs -text
+agr-benchmark.png filter=lfs diff=lfs merge=lfs -text
+agr-benchmarks.png filter=lfs diff=lfs merge=lfs -text
+agr-layout.png filter=lfs diff=lfs merge=lfs -text
+agr-wins.png filter=lfs diff=lfs merge=lfs -text

LICENSE.md

@@ -0,0 +1,190 @@
+                                 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
+
+   Copyright 2026 Command Code
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

NOTICE.md

@@ -0,0 +1,25 @@
+# Notice
+
+Agr<br>
+Copyright 2026 Command Code
+
+## Base models
+
+The Agr models are modified versions of these base models, both under the Apache License 2.0:
+
+| Model | Base model |
+| --- | --- |
+| Agr | google/gemma-4-31B-it |
+| Agr-flash | HuggingFaceTB/SmolLM2-360M |
+
+Modifications: fine-tuned weights merged into the base model. Agr keeps the base model's vocabulary, chat template and output layer, and adds a fitted answer temperature to its configuration. Agr-flash adds marker tokens to the vocabulary and a scoring head on top.
+
+## Adapted work
+
+Parts of Agr's letters readout are adapted from Decider (github.com/Mapika/decider), Copyright 2026 Mark Marosi, Apache License 2.0: the answer-slot prompt layout, the single-token option codes, the handling of a thinking block the chat template leaves open, the position indexes written into long JSON arrays, the "no:"/"yes:" option rendering, and the temperature as a function of the number of options. Option-order averaging follows Invergent AI's Surogate.
+
+## Design
+
+Agr's question-branch design follows TypeSafe's Jev, as described in Archer Hume's "Jev's Architecture Unmasked", and Kev. The request and response format follows TypeSafe's System One API, so TypeSafe's SDK works against an Agr server.
+
+Agr is an independent project, not affiliated with or endorsed by TypeSafe, Decider's authors, or the authors of any model it is compared with.

README.md

@@ -0,0 +1,217 @@
+---
+license: apache-2.0
+base_model: google/gemma-4-31B-it
+library_name: agr
+pipeline_tag: text-classification
+inference: false
+tags:
+  - agr
+  - decision-model
+  - classifier
+---
+
+<h1 align="center">Agr</h1>
+
+<p align="center">
+  <a href="https://huggingface.co/collections/CommandCode/agr"><img src="https://img.shields.io/badge/Hugging%20Face-models-6366F1" alt="Models"></a>
+  <a href="https://huggingface.co/spaces/multimodalart/jev-decision-index"><img src="https://img.shields.io/badge/Decision%20Index%200.2.1-58.15-6366F1" alt="Decision Index"></a>
+  <img src="https://img.shields.io/badge/license-Apache%202.0-3f3f46" alt="License">
+</p>
+
+Agr is a decision model from Command Code. You send it a state, as text or JSON, with one or more typed questions, and it returns typed answers with a probability for every option.
+
+Agr comes in two sizes, **Agr** and **Agr-flash**. Agr scores 58.15 on Decision Index 0.2.1 across 38 counted benchmarks.
+
+## Why Agr
+
+Decisions like routing a ticket, picking a tool or approving an action have a fixed set of answers. Agr scores those answers directly and returns a probability for each, instead of generating text you have to parse. Every question about the same state is answered in one forward pass, independently of the others.
+
+## Evaluation results
+
+![Agr compared with Jev, Kev and Rune on Decision Index 0.2.1](agr-benchmark.png)
+
+We ran Agr on all 150,759 requests of the public [Decision Index kit](https://github.com/apolinario/decision-index) at revision `87d4650`, with nothing truncated. Jev, Kev and Rune scores come from the public board as of 28 September 2026.
+
+The index averages chance-corrected skill over five areas. The figures show skill × 100, not raw accuracy or F1, and higher is better everywhere, including ForecastBench, where the index turns the Brier score into a skill score. 3% of BRIGHT requests exceed Agr's 32,768-token limit and count as wrong.
+
+![Results by area](agr-areas.png)
+
+![Results for all 38 counted benchmarks](agr-benchmarks.png)
+
+## Models
+
+| Model | On disk | Token limits (state / question / request) |
+| --- | --- | --- |
+| [CommandCode/agr](https://huggingface.co/CommandCode/agr) | 61.43 GB | 16,384 / 16,384 / 32,768 |
+| [CommandCode/agr-flash](https://huggingface.co/CommandCode/agr-flash) | 0.73 GB | 6,144 / 2,048 / 12,288 |
+
+Agr-flash is not meant for safety decisions. Each repository holds the weights, tokenizer and config.
+
+## How it works
+
+All questions read the same state, but no question can attend to another. Agr reads each option's score from the model's own output layer; Agr-flash uses a small trained scoring head. Details are in [DESIGN.md](https://github.com/CommandCodeAI/agr/blob/HEAD/DESIGN.md).
+
+## Quickstart
+
+Agr needs Python 3.11 or newer. Installing it pulls in PyTorch and the Transformers version it was tested with:
+
+```bash
+pip install git+https://github.com/CommandCodeAI/agr
+```
+
+### Python
+
+The first run downloads the weights from Hugging Face. Agr's bf16 weights are 61.43 GB; we serve it on 96 GB GPUs. For Agr-flash, use `CommandCode/agr-flash`.
+
+```python
+from agr.model import Agr
+from agr.records import boolean, choice
+
+model = Agr.from_pretrained("CommandCode/agr", device="cuda")
+questions = [
+    choice(
+        "team",
+        "Which team should handle this?",
+        ["Engineering: technical faults", "Billing: payment disputes"],
+    ),
+    boolean("urgent", "Does this require immediate attention?"),
+]
+
+probabilities = model.probs(
+    "Checkout is failing and customers cannot place orders.",
+    questions,
+)
+for question, values in zip(questions, probabilities):
+    print(question.qid, dict(zip(question.options, values.tolist())))
+```
+
+### Server
+
+```bash
+agr serve CommandCode/agr --port 8000
+```
+
+The server takes System One requests at `/v1/systemone`. Call it with curl, Python or TypeScript.
+
+#### curl
+
+```bash
+curl -s localhost:8000/v1/systemone -H 'content-type: application/json' -d '{
+  "model": "agr-latest",
+  "state": "Agent wants to run: git push --force origin main\nRepo: shared by 14 people, no branch protection.",
+  "questions": {
+    "loses_work": {"type": "noul", "instructions": "Could this erase work that other people pushed?"},
+    "blast": {"type": "score", "instructions": "Who could this affect if it goes wrong?",
+      "criteria": ["Nobody", "Only the person running it", "Everyone who uses the repo"]}
+  }
+}'
+```
+
+#### Python
+
+With the [TypeSafe Python SDK](https://docs.typesafe.ai/sdk/python):
+
+```bash
+pip install typesafe-sdk
+```
+
+```python
+from typesafe_sdk import Choice, Noul, TypeSafeClient
+
+with TypeSafeClient(
+    api_key="local",
+    base_url="http://localhost:8000",
+    model="agr-latest",
+) as client:
+    result = client.system_one(
+        state="Checkout is failing and customers cannot place orders.",
+        questions={
+            "urgent": Noul(instructions="Does this require immediate attention?"),
+            "team": Choice(
+                instructions="Which team should handle this?",
+                criteria={"engineering": "Technical faults", "billing": "Payment disputes"},
+            ),
+        },
+    )
+
+print(result.nouls["urgent"].noul)
+print(result.choices["team"].choice)
+```
+
+#### TypeScript
+
+With the [TypeSafe JavaScript SDK](https://docs.typesafe.ai/sdk/javascript) (Node.js 20 or newer):
+
+```bash
+npm install @typesafe-ai/sdk
+```
+
+```ts
+import { choice, noul, TypeSafeClient } from "@typesafe-ai/sdk";
+
+const client = new TypeSafeClient({
+  apiKey: "local",
+  baseURL: "http://localhost:8000",
+  defaultModel: "agr-latest",
+});
+
+const { answers } = await client.systemOne({
+  state: "Checkout is failing and customers cannot place orders.",
+  questions: {
+    urgent: noul("Does this require immediate attention?"),
+    team: choice("Which team should handle this?", {
+      engineering: "Technical faults",
+      billing: "Payment disputes",
+    }),
+  },
+});
+
+console.log(answers.urgent.noul);
+console.log(answers.team.choice);
+```
+
+The server needs no key by default, so the API key can be any string. If you start it with `--api-key` or `AGR_API_KEY`, pass that key.
+
+### Request format
+
+| Field | Description |
+| --- | --- |
+| `model` | `agr-latest` |
+| `state` | Text or JSON describing the situation |
+| `questions` | Questions keyed by ID |
+
+Each question has a `type` and `instructions`. The three question types are the System One primitives: a `choice` selects one of several named options, with `criteria` mapping each name to a description; a `score` rates the state on levels listed in order in `criteria`; and a `noul` is a yes/no judgment. Inputs are text only.
+
+| Type | Options | Output |
+| --- | --- | --- |
+| `noul` | No / yes | Probability of yes |
+| `choice` | 2–255 named options | Selected option, probability distribution and confidence |
+| `score` | 2–255 ordered levels | Expected level index, probability distribution and confidence |
+
+`confidence` uses System One's definition, which is not the top option's probability.
+
+### Evaluate on your data
+
+```bash
+agr eval http://localhost:8000 labelled.jsonl --out agr-results.jsonl
+```
+
+It reports accuracy, Brier score, calibration error and latency. The label format and `agr compare` are in [DESIGN.md](https://github.com/CommandCodeAI/agr/blob/HEAD/DESIGN.md#measuring-it-yourself).
+
+## Limitations
+
+- English text only; images are not supported.
+- Answers are limited to the options you supply; there is no generated text or explanation.
+- Check probabilities and thresholds on your own labelled examples before relying on them.
+- Changing the order of options can change the answer.
+- Neither model is a standalone safety filter or fit to be the sole authority for consequential decisions about people.
+
+## Acknowledgements
+
+Agr builds on TypeSafe's Jev and its System One API, and on Archer Hume's analysis of Jev's architecture. It also draws on Jared Palmer's Kev, Mark Marosi's Decider for the answer-slot prompt layout and the temperature rule, and Invergent AI's Surogate for option-order averaging. We thank apolinário and the Decision Index maintainers, Google DeepMind for Gemma, Hugging Face for SmolLM2, and the authors of the datasets we used. Agr is not affiliated with TypeSafe or the systems compared here.
+
+## License
+
+Agr is released under Apache 2.0. Agr uses Google DeepMind's Gemma 4 31B IT; Agr-flash uses Hugging Face's SmolLM2-360M. Both base models are licensed under Apache 2.0. See NOTICE.md for attribution.
+
+© 2026 Command Code

backbone/chat_template.jinja

@@ -0,0 +1,390 @@
+{#
+  Template: Google Gemma 4 Canonical Chat Template
+  Author: Google Gemma Engineering Team
+  Published: 2026-07-09
+  Context: Fixed tool-calling loops, turn closures, and thinking content-ordering.
+#}
+{%- macro format_parameters(properties, required, filter_keys=false) -%}
+    {%- set standard_keys = ['description', 'type', 'properties', 'required', 'nullable'] -%}
+    {%- set ns = namespace(found_first=false) -%}
+    {%- for key, value in properties | dictsort -%}
+        {%- set add_comma = false -%}
+        {%- if not filter_keys or key not in standard_keys -%}
+            {%- if ns.found_first %},{% endif -%}
+            {%- set ns.found_first = true -%}
+            {{ key }}:{
+            {%- if value['description'] -%}
+                description:<|"|>{{ value['description'] }}<|"|>
+                {%- set add_comma = true -%}
+            {%- endif -%}
+            {%- if value['type'] | upper == 'STRING' -%}
+                {%- if value['enum'] -%}
+                    {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                    enum:{{ format_argument(value['enum']) }}
+                {%- endif -%}
+            {%- elif value['type'] | upper == 'ARRAY' -%}
+                {%- if value['items'] is mapping and value['items'] -%}
+                    {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                    items:{
+                    {%- set ns_items = namespace(found_first=false) -%}
+                    {%- for item_key, item_value in value['items'] | dictsort -%}
+                        {%- if item_value is not none -%}
+                            {%- if ns_items.found_first %},{% endif -%}
+                            {%- set ns_items.found_first = true -%}
+                            {%- if item_key == 'properties' -%}
+                                properties:{
+                                {%- if item_value is mapping -%}
+                                    {{- format_parameters(item_value, value['items']['required'] | default([])) -}}
+                                {%- endif -%}
+                                }
+                            {%- elif item_key == 'required' -%}
+                                required:[
+                                {%- for req_item in item_value -%}
+                                    <|"|>{{- req_item -}}<|"|>
+                                    {%- if not loop.last %},{% endif -%}
+                                {%- endfor -%}
+                                ]
+                            {%- elif item_key == 'type' -%}
+                                {%- if item_value is string -%}
+                                    type:{{ format_argument(item_value | upper) }}
+                                {%- else -%}
+                                    type:{{ format_argument(item_value | map('upper') | list) }}
+                                {%- endif -%}
+                            {%- else -%}
+                                {{ item_key }}:{{ format_argument(item_value) }}
+                            {%- endif -%}
+                        {%- endif -%}
+                    {%- endfor -%}
+                    }
+                {%- endif -%}
+            {%- endif -%}
+            {%- if value['nullable'] %}
+                {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                nullable:true
+            {%- endif -%}
+            {%- if value['type'] | upper == 'OBJECT' -%}
+                {%- if value['properties'] is defined and value['properties'] is mapping -%}
+                    {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                    properties:{
+                    {{- format_parameters(value['properties'], value['required'] | default([])) -}}
+                    }
+                {%- elif value is mapping -%}
+                    {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                    properties:{
+                    {{- format_parameters(value, value['required'] | default([]), filter_keys=true) -}}
+                    }
+                {%- endif -%}
+                {%- if value['required'] -%}
+                    {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+                    required:[
+                    {%- for item in value['required'] | default([]) -%}
+                        <|"|>{{- item -}}<|"|>
+                        {%- if not loop.last %},{% endif -%}
+                    {%- endfor -%}
+                    ]
+                {%- endif -%}
+            {%- endif -%}
+            {%- if add_comma %},{%- else -%} {%- set add_comma = true -%} {% endif -%}
+            type:<|"|>{{ value['type'] | upper }}<|"|>}
+        {%- endif -%}
+    {%- endfor -%}
+{%- endmacro -%}
+{%- macro format_function_declaration(tool_data) -%}
+    declaration:{{- tool_data['function']['name'] -}}{description:<|"|>{{- tool_data['function']['description'] -}}<|"|>
+    {%- set params = tool_data['function']['parameters'] -%}
+    {%- if params -%}
+        ,parameters:{
+        {%- if params['properties'] -%}
+            properties:{ {{- format_parameters(params['properties'], params['required']) -}} },
+        {%- endif -%}
+        {%- if params['required'] -%}
+            required:[
+            {%- for item in params['required'] -%}
+                <|"|>{{- item -}}<|"|>
+                {{- ',' if not loop.last -}}
+            {%- endfor -%}
+            ],
+        {%- endif -%}
+        {%- if params['type'] -%}
+            type:<|"|>{{- params['type'] | upper -}}<|"|>}
+        {%- endif -%}
+    {%- endif -%}
+    {%- if 'response' in tool_data['function'] -%}
+        {%- set response_declaration = tool_data['function']['response'] -%}
+        ,response:{
+        {%- if response_declaration['description'] -%}
+            description:<|"|>{{- response_declaration['description'] -}}<|"|>,
+        {%- endif -%}
+        {%- if response_declaration['type'] | upper == 'OBJECT' -%}
+            type:<|"|>{{- response_declaration['type'] | upper -}}<|"|>}
+        {%- endif -%}
+    {%- endif -%}
+    }
+{%- endmacro -%}
+{%- macro format_argument(argument, escape_keys=True) -%}
+    {%- if argument is none -%}
+        {{- 'null' -}}
+    {%- elif argument is string -%}
+        {{- '<|"|>' + argument + '<|"|>' -}}
+    {%- elif argument is boolean -%}
+        {{- 'true' if argument else 'false' -}}
+    {%- elif argument is mapping -%}
+        {{- '{' -}}
+        {%- set ns = namespace(found_first=false) -%}
+        {%- for key, value in argument | dictsort -%}
+            {%- if ns.found_first %},{% endif -%}
+            {%- set ns.found_first = true -%}
+            {%- if escape_keys -%}
+                {{- '<|"|>' + key + '<|"|>' -}}
+            {%- else -%}
+                {{- key -}}
+            {%- endif -%}
+            :{{- format_argument(value, escape_keys=escape_keys) -}}
+        {%- endfor -%}
+        {{- '}' -}}
+    {%- elif argument is sequence -%}
+        {{- '[' -}}
+        {%- for item in argument -%}
+            {{- format_argument(item, escape_keys=escape_keys) -}}
+            {%- if not loop.last %},{% endif -%}
+        {%- endfor -%}
+        {{- ']' -}}
+    {%- else -%}
+        {{- argument -}}
+    {%- endif -%}
+{%- endmacro -%}
+{%- macro strip_thinking(text) -%}
+    {%- set ns = namespace(result='') -%}
+    {%- for part in text.split('<channel|>') -%}
+        {%- if '<|channel>' in part -%}
+            {%- set ns.result = ns.result + part.split('<|channel>')[0] -%}
+        {%- else -%}
+            {%- set ns.result = ns.result + part -%}
+        {%- endif -%}
+    {%- endfor -%}
+    {{- ns.result | trim -}}
+{%- endmacro -%}
+
+{%- macro format_tool_response_block(tool_name, response) -%}
+    {{- '<|tool_response>' -}}
+    {%- if response is mapping -%}
+        {{- 'response:' + tool_name + '{' -}}
+        {%- for key, value in response | dictsort -%}
+            {{- key -}}:{{- format_argument(value, escape_keys=False) -}}
+            {%- if not loop.last %},{% endif -%}
+        {%- endfor -%}
+        {{- '}' -}}
+    {%- else -%}
+        {{- 'response:' + tool_name + '{value:' + format_argument(response, escape_keys=False) + '}' -}}
+    {%- endif -%}
+    {{- '<tool_response|>' -}}
+{%- endmacro -%}
+
+{#- ===== SETUP ===== -#}
+{%- set ns = namespace(prev_message_type=None, prev_non_tool_role=None) -%}
+{%- set loop_messages = messages -%}
+{%- set enable_thinking = enable_thinking | default(false) -%}
+{%- set preserve_thinking = preserve_thinking | default(false) -%}
+{{- bos_token -}}
+{#- Handle System/Tool Definitions Block -#}
+{%- if enable_thinking or tools or (messages and messages[0]['role'] in ['system', 'developer']) -%}
+    {{- '<|turn>system\n' -}}
+    {#- Inject Thinking token at the very top of the FIRST system turn -#}
+    {%- if enable_thinking -%}
+        {{- '<|think|>\n' -}}
+        {%- set ns.prev_message_type = 'think' -%}
+    {%- endif -%}
+    {%- if messages and messages[0]['role'] in ['system', 'developer'] -%}
+        {%- if messages[0]['content'] is string -%}
+            {{- messages[0]['content'] | trim -}}
+        {%- elif messages[0]['content'] is sequence -%}
+            {%- for item in messages[0]['content'] -%}
+                {{- item['text'] | trim + ' '-}}
+            {%- endfor -%}
+        {%- endif -%}
+        {%- set loop_messages = messages[1:] -%}
+    {%- endif -%}
+    {%- if tools -%}
+        {%- for tool in tools %}
+            {{- '<|tool>' -}}
+            {{- format_function_declaration(tool) | trim -}}
+            {{- '<tool|>' -}}
+        {%- endfor %}
+        {%- set ns.prev_message_type = 'tool' -%}
+    {%- endif -%}
+    {{- '<turn|>\n' -}}
+{%- endif %}
+
+{#- Pre-scan: find last user message index for reasoning guard -#}
+{%- set ns_turn = namespace(last_user_idx=-1) -%}
+{%- for i in range(loop_messages | length) -%}
+    {%- if loop_messages[i]['role'] == 'user' -%}
+        {%- set ns_turn.last_user_idx = i -%}
+    {%- endif -%}
+{%- endfor -%}
+
+{#- Loop through messages -#}
+{%- for message in loop_messages -%}
+    {%- if message['role'] != 'tool' -%}
+    {%- set ns.prev_message_type = None -%}
+    {%- set role = 'model' if message['role'] == 'assistant' else message['role'] -%}
+    {#- Detect continuation using tracked state — O(1) instead of O(n) backward scan -#}
+    {%- set continue_same_model_turn = (role == 'model' and ns.prev_non_tool_role == 'assistant') -%}
+    {%- if not continue_same_model_turn -%}
+        {{- '<|turn>' + role + '\n' }}
+
+    {%- endif -%}
+
+    {#- Render reasoning/reasoning_content as thinking channel -#}
+    {%- set thinking_text = message.get('reasoning') or message.get('reasoning_content') -%}
+    {%- set thinking_gate = (loop.index0 > ns_turn.last_user_idx) or (preserve_thinking and message.get('tool_calls')) -%}
+    {%- if thinking_text and thinking_gate -%}
+        {{- '<|channel>thought\n' + thinking_text + '\n<channel|>' -}}
+    {%- endif -%}
+
+            {%- if message.get('tool_calls') -%}
+                {%- for tool_call in message.get('tool_calls') -%}
+                    {%- set function = tool_call['function'] -%}
+                    {{- '<|tool_call>call:' + function['name'] + '{' -}}
+                    {%- if function['arguments'] is mapping -%}
+                        {%- set ns_args = namespace(found_first=false) -%}
+                        {%- for key, value in function['arguments'] | dictsort -%}
+                            {%- if ns_args.found_first %},{% endif -%}
+                            {%- set ns_args.found_first = true -%}
+                            {{- key -}}:{{- format_argument(value, escape_keys=False) -}}
+                        {%- endfor -%}
+                    {%- elif function['arguments'] is none -%}
+                    {%- else -%}
+                        {{- raise_exception(
+                            "chat_template: tool_calls[].function.arguments must be a "
+                            "JSON object (mapping), not a string. Deserialize arguments "
+                            "before passing to the template."
+                        ) -}}
+                    {%- endif -%}
+                    {{- '}<tool_call|>' -}}
+                {%- endfor -%}
+                {%- set ns.prev_message_type = 'tool_call' -%}
+            {%- endif -%}
+
+            {%- set ns_tr_out = namespace(flag=false) -%}
+            {%- if message.get('tool_responses') -%}
+                {#- Legacy: tool_responses embedded on the assistant message (Google/Gemma native) -#}
+                {%- for tool_response in message.get('tool_responses') -%}
+                    {{- format_tool_response_block(tool_response['name'] | default('unknown', true), tool_response['response']) -}}
+                    {%- set ns_tr_out.flag = true -%}
+                    {%- set ns.prev_message_type = 'tool_response' -%}
+                {%- endfor -%}
+            {%- elif message.get('tool_calls') -%}
+                {#- OpenAI Chat Completions: forward-scan consecutive role:tool messages -#}
+                {%- set ns_tool_scan = namespace(stopped=false) -%}
+                {%- for k in range(loop.index0 + 1, loop_messages | length) -%}
+                    {%- if ns_tool_scan.stopped -%}
+                    {%- elif loop_messages[k]['role'] != 'tool' -%}
+                        {%- set ns_tool_scan.stopped = true -%}
+                    {%- else -%}
+                        {%- set follow = loop_messages[k] -%}
+                        {#- Resolve tool_call_id to function name -#}
+                        {%- set ns_tname = namespace(name=follow.get('name') or 'unknown') -%}
+                        {%- for tc in message.get('tool_calls') -%}
+                            {%- if tc.get('id') == follow.get('tool_call_id') -%}
+                                {%- set ns_tname.name = tc['function']['name'] -%}
+                            {%- endif -%}
+                        {%- endfor -%}
+                        {#- Handle content as string or content-parts array -#}
+                        {%- set tool_body = follow.get('content') -%}
+                        {%- if tool_body is string -%}
+                            {{- format_tool_response_block(ns_tname.name, tool_body) -}}
+                        {%- elif tool_body is sequence and tool_body is not string -%}
+                            {%- set ns_txt = namespace(s='') -%}
+                            {%- for part in tool_body -%}
+                                {%- if part.get('type') == 'text' -%}
+                                    {%- set ns_txt.s = ns_txt.s + (part.get('text') | default('')) -%}
+                                {%- endif -%}
+                            {%- endfor -%}
+                            {{- format_tool_response_block(ns_tname.name, ns_txt.s) -}}
+                            {%- for part in tool_body -%}
+                                {%- if part.get('type') in ['image', 'image_url'] -%}
+                                    {{- '<|image|>' -}}
+                                {%- elif part.get('type') in ['audio', 'input_audio'] -%}
+                                    {{- '<|audio|>' -}}
+                                {%- elif part.get('type') == 'video' -%}
+                                    {{- '<|video|>' -}}
+                                {%- endif -%}
+                            {%- endfor -%}
+                        {%- else -%}
+                            {{- format_tool_response_block(ns_tname.name, tool_body) -}}
+                        {%- endif -%}
+                        {%- set ns_tr_out.flag = true -%}
+                        {%- set ns.prev_message_type = 'tool_response' -%}
+                    {%- endif -%}
+                {%- endfor -%}
+            {%- endif -%}
+
+            {%- set captured_content -%}
+            {%- if message.get('content') is string -%}
+                {%- if role == 'model' -%}
+                    {{- strip_thinking(message['content']) -}}
+                {%- else -%}
+                    {{- message['content'] | trim -}}
+                {%- endif -%}
+            {%- elif message.get('content') is sequence -%}
+                {%- for item in message['content'] -%}
+                    {%- if item.get('type') == 'text' -%}
+                        {%- if role == 'model' -%}
+                            {{- strip_thinking(item['text']) -}}
+                        {%- else -%}
+                            {{- item['text'] | trim -}}
+                        {%- endif -%}
+                    {%- elif item.get('type') in ['image', 'image_url'] -%}
+                        {{- '<|image|>' -}}
+                    {%- elif item.get('type') in ['audio', 'input_audio'] -%}
+                        {{- '<|audio|>' -}}
+                    {%- elif item.get('type') == 'video' -%}
+                        {{- '<|video|>' -}}
+                    {%- endif -%}
+                {%- endfor -%}
+            {%- endif -%}
+            {%- endset -%}
+
+            {{- captured_content -}}
+            {%- set has_content = captured_content | trim | length > 0 -%}
+
+        {#- Forward-scan: find next non-tool message role for continuation detection -#}
+        {%- set next_nt = namespace(role=None, found=false) -%}
+        {%- for j in range(loop.index0 + 1, loop_messages | length) -%}
+            {%- if not next_nt.found -%}
+                {%- if loop_messages[j]['role'] != 'tool' -%}
+                    {%- set next_nt.role = loop_messages[j]['role'] -%}
+                    {%- set next_nt.found = true -%}
+                {%- endif -%}
+            {%- endif -%}
+        {%- endfor -%}
+
+        {%- set continues_into_next = (
+            role == 'model'
+            and next_nt.role == 'assistant'
+            and (not message.get('tool_calls') or ns_tr_out.flag)
+        ) -%}
+
+        {%- if ns.prev_message_type == 'tool_call' and not ns_tr_out.flag -%}
+            {{- '<|tool_response>' -}}
+        {%- elif continues_into_next -%}
+        {%- elif not (ns_tr_out.flag and not has_content and not next_nt.found) -%}
+            {{- '<turn|>\n' -}}
+        {%- endif -%}
+
+    {#- Track previous non-tool role for next iteration (avoids O(n) backward scan) -#}
+    {%- set ns.prev_non_tool_role = message['role'] -%}
+    {%- endif -%}
+{%- endfor -%}
+
+{%- if add_generation_prompt -%}
+    {%- if ns.prev_message_type != 'tool_response' and ns.prev_message_type != 'tool_call' -%}
+        {{- '<|turn>model\n' -}}
+        {%- if not enable_thinking -%}
+            {{- '<|channel>thought\n<channel|>' -}}
+        {%- endif -%}
+    {%- elif ns.prev_message_type == 'tool_response' and enable_thinking -%}
+        {{- '<|channel>thought\n' -}}
+    {%- endif -%}
+{%- endif -%}

backbone/config.json

@@ -0,0 +1,154 @@
+{
+  "architectures": [
+    "Gemma4TextModel"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attention_k_eq_v": true,
+  "bos_token_id": 2,
+  "dtype": "bfloat16",
+  "enable_moe_block": false,
+  "eos_token_id": 1,
+  "expert_intermediate_size": null,
+  "final_logit_softcapping": 30.0,
+  "head_dim": 256,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 5376,
+  "hidden_size_per_layer_input": 0,
+  "initializer_range": 0.02,
+  "intermediate_size": 21504,
+  "layer_types": [
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "sliding_attention",
+    "full_attention"
+  ],
+  "max_position_embeddings": 262144,
+  "model_type": "gemma4_text",
+  "moe_intermediate_size": null,
+  "num_attention_heads": 32,
+  "num_experts": null,
+  "num_hidden_layers": 60,
+  "num_key_value_heads": 16,
+  "num_kv_shared_layers": 0,
+  "pad_token_id": 0,
+  "per_layer_config": {
+    "05": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "11": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "17": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "23": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "29": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "35": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "41": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "47": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "53": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    },
+    "59": {
+      "head_dim": 512,
+      "num_key_value_heads": 4
+    }
+  },
+  "rms_norm_eps": 1e-06,
+  "rope_parameters": {
+    "full_attention": {
+      "partial_rotary_factor": 0.25,
+      "rope_theta": 1000000.0,
+      "rope_type": "proportional"
+    },
+    "sliding_attention": {
+      "rope_theta": 10000.0,
+      "rope_type": "default"
+    }
+  },
+  "sliding_window": 1024,
+  "tie_word_embeddings": true,
+  "top_k_experts": null,
+  "transformers_version": "5.18.0",
+  "use_bidirectional_attention": "vision",
+  "use_cache": true,
+  "use_double_wide_mlp": false,
+  "vocab_size": 262144,
+  "vocab_size_per_layer_input": 262144
+}

backbone/model-00001-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7acc80b1a9499d69225b361975c4e774a63a7c73b891ed8ee6777c70e72c674
+size 4998654806

backbone/model-00002-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5ed3e2b4209a7b217089ce1eca4715225a88c31806ae493a3ab45862bef45e1
+size 4899700058

backbone/model-00003-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4dd72ba829659f3629440fd1f58952081f7160ba21e0db022753abb3af8c3ebb
+size 4899700074

backbone/model-00004-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1b08d4611bf3e78697e2dec5b4f7b6a3938bcddbf201625156bd9700b666267f
+size 4899700122

backbone/model-00005-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:23d2b4349d88c61bf89dbeedc65ce3b689b4ddbb794738c266e390e68291099c
+size 4789598730

backbone/model-00006-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:402cb075e1284077e0fa9d9d4aa72277ee6b5841e893953eda27d882ce8c7900
+size 4899700122

backbone/model-00007-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:31a8ca1c8a8145d5930adf8a1bad88b8a11dcbbccac635bbb2ee15dcba4d92ea
+size 4899700122

backbone/model-00008-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ec11a7254e0f6626b58d63d340c442074f656f491532f7743522b20c1c3bc2b5
+size 4899700122

backbone/model-00009-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4f39703ce74f75a816cdf26d34e1e19656c640f4e17d006dac75c2516c35460
+size 4899700122

backbone/model-00010-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b6a0fff3c252046f16fc2404c9e817ab4bcaa056995ee6d9ad4bef0bc2905ad
+size 4899700122

backbone/model-00011-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a98fa0148635afb2f903efd82945f530db71bb088c869c725053bc3f03815787
+size 4789598730

backbone/model-00012-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0cdd372525a5a93b6d6f9e008377cfef00d4b2ba03041db810032a6d40128722
+size 4899700122

backbone/model-00013-of-00013.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d9074ac41fbfdd5d595c72d597d47cf93c89793e9b4b0d54feacd122e3ca1cfb
+size 2719628540

backbone/model.safetensors.index.json

@@ -0,0 +1,840 @@
+{
+  "metadata": {
+    "total_parameters": 30697345280,
+    "total_size": 61394690680
+  },
+  "weight_map": {
+    "embed_tokens.weight": "model-00001-of-00013.safetensors",
+    "layers.0.input_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.layer_scalar": "model-00001-of-00013.safetensors",
+    "layers.0.mlp.down_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.mlp.gate_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.mlp.up_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.post_attention_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.post_feedforward_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.pre_feedforward_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.k_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.k_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.o_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.q_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.q_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.0.self_attn.v_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.input_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.layer_scalar": "model-00001-of-00013.safetensors",
+    "layers.1.mlp.down_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.mlp.gate_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.mlp.up_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.post_attention_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.post_feedforward_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.pre_feedforward_layernorm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.k_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.k_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.o_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.q_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.q_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.1.self_attn.v_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.10.input_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.layer_scalar": "model-00003-of-00013.safetensors",
+    "layers.10.mlp.down_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.mlp.gate_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.mlp.up_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.post_attention_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.post_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.pre_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.k_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.k_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.o_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.q_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.q_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.10.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.input_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.layer_scalar": "model-00003-of-00013.safetensors",
+    "layers.11.mlp.down_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.mlp.gate_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.mlp.up_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.post_attention_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.post_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.pre_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.self_attn.k_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.self_attn.k_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.self_attn.o_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.11.self_attn.q_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.11.self_attn.q_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.12.input_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.12.layer_scalar": "model-00003-of-00013.safetensors",
+    "layers.12.mlp.down_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.12.mlp.gate_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.12.mlp.up_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.12.post_attention_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.12.post_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.12.pre_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.12.self_attn.k_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.12.self_attn.k_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.12.self_attn.o_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.12.self_attn.q_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.12.self_attn.q_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.12.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.13.input_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.layer_scalar": "model-00004-of-00013.safetensors",
+    "layers.13.mlp.down_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.mlp.gate_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.mlp.up_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.post_attention_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.post_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.pre_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.k_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.k_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.o_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.q_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.q_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.13.self_attn.v_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.input_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.layer_scalar": "model-00004-of-00013.safetensors",
+    "layers.14.mlp.down_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.mlp.gate_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.mlp.up_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.post_attention_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.post_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.pre_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.k_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.k_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.o_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.q_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.q_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.14.self_attn.v_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.input_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.layer_scalar": "model-00004-of-00013.safetensors",
+    "layers.15.mlp.down_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.mlp.gate_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.mlp.up_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.post_attention_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.post_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.pre_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.k_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.k_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.o_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.q_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.q_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.15.self_attn.v_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.input_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.layer_scalar": "model-00004-of-00013.safetensors",
+    "layers.16.mlp.down_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.mlp.gate_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.mlp.up_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.post_attention_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.post_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.pre_feedforward_layernorm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.k_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.k_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.o_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.q_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.q_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.16.self_attn.v_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.17.input_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.17.layer_scalar": "model-00004-of-00013.safetensors",
+    "layers.17.mlp.down_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.17.mlp.gate_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.17.mlp.up_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.17.post_attention_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.17.post_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.17.pre_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.17.self_attn.k_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.17.self_attn.k_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.17.self_attn.o_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.17.self_attn.q_norm.weight": "model-00004-of-00013.safetensors",
+    "layers.17.self_attn.q_proj.weight": "model-00004-of-00013.safetensors",
+    "layers.18.input_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.layer_scalar": "model-00005-of-00013.safetensors",
+    "layers.18.mlp.down_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.mlp.gate_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.mlp.up_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.post_attention_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.post_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.pre_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.k_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.k_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.o_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.q_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.q_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.18.self_attn.v_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.input_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.layer_scalar": "model-00005-of-00013.safetensors",
+    "layers.19.mlp.down_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.mlp.gate_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.mlp.up_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.post_attention_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.post_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.pre_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.k_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.k_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.o_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.q_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.q_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.19.self_attn.v_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.2.input_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.2.layer_scalar": "model-00001-of-00013.safetensors",
+    "layers.2.mlp.down_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.2.mlp.gate_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.2.mlp.up_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.2.post_attention_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.2.post_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.2.pre_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.2.self_attn.k_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.2.self_attn.k_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.2.self_attn.o_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.2.self_attn.q_norm.weight": "model-00001-of-00013.safetensors",
+    "layers.2.self_attn.q_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.2.self_attn.v_proj.weight": "model-00001-of-00013.safetensors",
+    "layers.20.input_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.layer_scalar": "model-00005-of-00013.safetensors",
+    "layers.20.mlp.down_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.mlp.gate_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.mlp.up_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.post_attention_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.post_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.pre_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.k_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.k_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.o_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.q_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.q_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.20.self_attn.v_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.input_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.layer_scalar": "model-00005-of-00013.safetensors",
+    "layers.21.mlp.down_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.mlp.gate_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.mlp.up_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.post_attention_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.post_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.pre_feedforward_layernorm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.k_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.k_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.o_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.q_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.q_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.21.self_attn.v_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.22.input_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.22.layer_scalar": "model-00005-of-00013.safetensors",
+    "layers.22.mlp.down_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.22.mlp.gate_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.22.mlp.up_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.22.post_attention_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.22.post_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.22.pre_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.22.self_attn.k_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.22.self_attn.k_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.22.self_attn.o_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.22.self_attn.q_norm.weight": "model-00005-of-00013.safetensors",
+    "layers.22.self_attn.q_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.22.self_attn.v_proj.weight": "model-00005-of-00013.safetensors",
+    "layers.23.input_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.layer_scalar": "model-00006-of-00013.safetensors",
+    "layers.23.mlp.down_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.23.mlp.gate_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.23.mlp.up_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.23.post_attention_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.post_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.pre_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.self_attn.k_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.self_attn.k_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.23.self_attn.o_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.23.self_attn.q_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.23.self_attn.q_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.input_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.layer_scalar": "model-00006-of-00013.safetensors",
+    "layers.24.mlp.down_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.mlp.gate_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.mlp.up_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.post_attention_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.post_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.pre_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.k_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.k_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.o_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.q_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.q_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.24.self_attn.v_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.input_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.layer_scalar": "model-00006-of-00013.safetensors",
+    "layers.25.mlp.down_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.mlp.gate_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.mlp.up_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.post_attention_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.post_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.pre_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.k_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.k_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.o_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.q_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.q_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.25.self_attn.v_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.input_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.layer_scalar": "model-00006-of-00013.safetensors",
+    "layers.26.mlp.down_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.mlp.gate_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.mlp.up_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.post_attention_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.post_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.pre_feedforward_layernorm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.k_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.k_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.o_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.q_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.q_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.26.self_attn.v_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.27.input_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.27.layer_scalar": "model-00006-of-00013.safetensors",
+    "layers.27.mlp.down_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.27.mlp.gate_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.27.mlp.up_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.27.post_attention_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.27.post_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.27.pre_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.27.self_attn.k_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.27.self_attn.k_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.27.self_attn.o_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.27.self_attn.q_norm.weight": "model-00006-of-00013.safetensors",
+    "layers.27.self_attn.q_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.27.self_attn.v_proj.weight": "model-00006-of-00013.safetensors",
+    "layers.28.input_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.layer_scalar": "model-00007-of-00013.safetensors",
+    "layers.28.mlp.down_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.mlp.gate_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.mlp.up_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.post_attention_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.post_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.pre_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.k_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.k_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.o_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.q_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.q_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.28.self_attn.v_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.input_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.layer_scalar": "model-00007-of-00013.safetensors",
+    "layers.29.mlp.down_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.mlp.gate_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.mlp.up_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.post_attention_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.post_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.pre_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.self_attn.k_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.self_attn.k_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.self_attn.o_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.29.self_attn.q_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.29.self_attn.q_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.3.input_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.layer_scalar": "model-00002-of-00013.safetensors",
+    "layers.3.mlp.down_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.mlp.gate_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.mlp.up_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.post_attention_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.post_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.pre_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.k_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.k_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.o_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.q_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.q_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.3.self_attn.v_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.30.input_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.layer_scalar": "model-00007-of-00013.safetensors",
+    "layers.30.mlp.down_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.mlp.gate_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.mlp.up_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.post_attention_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.post_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.pre_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.k_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.k_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.o_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.q_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.q_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.30.self_attn.v_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.input_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.layer_scalar": "model-00007-of-00013.safetensors",
+    "layers.31.mlp.down_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.mlp.gate_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.mlp.up_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.post_attention_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.post_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.pre_feedforward_layernorm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.k_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.k_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.o_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.q_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.q_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.31.self_attn.v_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.32.input_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.32.layer_scalar": "model-00007-of-00013.safetensors",
+    "layers.32.mlp.down_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.32.mlp.gate_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.32.mlp.up_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.32.post_attention_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.32.post_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.32.pre_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.32.self_attn.k_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.32.self_attn.k_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.32.self_attn.o_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.32.self_attn.q_norm.weight": "model-00007-of-00013.safetensors",
+    "layers.32.self_attn.q_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.32.self_attn.v_proj.weight": "model-00007-of-00013.safetensors",
+    "layers.33.input_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.layer_scalar": "model-00008-of-00013.safetensors",
+    "layers.33.mlp.down_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.mlp.gate_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.mlp.up_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.post_attention_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.post_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.pre_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.k_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.k_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.o_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.q_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.q_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.33.self_attn.v_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.input_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.layer_scalar": "model-00008-of-00013.safetensors",
+    "layers.34.mlp.down_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.mlp.gate_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.mlp.up_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.post_attention_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.post_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.pre_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.k_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.k_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.o_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.q_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.q_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.34.self_attn.v_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.input_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.layer_scalar": "model-00008-of-00013.safetensors",
+    "layers.35.mlp.down_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.mlp.gate_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.mlp.up_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.post_attention_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.post_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.pre_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.self_attn.k_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.self_attn.k_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.self_attn.o_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.35.self_attn.q_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.35.self_attn.q_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.input_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.layer_scalar": "model-00008-of-00013.safetensors",
+    "layers.36.mlp.down_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.mlp.gate_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.mlp.up_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.post_attention_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.post_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.pre_feedforward_layernorm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.k_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.k_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.o_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.q_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.q_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.36.self_attn.v_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.37.input_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.37.layer_scalar": "model-00008-of-00013.safetensors",
+    "layers.37.mlp.down_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.37.mlp.gate_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.37.mlp.up_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.37.post_attention_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.37.post_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.37.pre_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.37.self_attn.k_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.37.self_attn.k_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.37.self_attn.o_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.37.self_attn.q_norm.weight": "model-00008-of-00013.safetensors",
+    "layers.37.self_attn.q_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.37.self_attn.v_proj.weight": "model-00008-of-00013.safetensors",
+    "layers.38.input_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.layer_scalar": "model-00009-of-00013.safetensors",
+    "layers.38.mlp.down_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.mlp.gate_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.mlp.up_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.post_attention_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.post_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.pre_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.k_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.k_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.o_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.q_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.q_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.38.self_attn.v_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.input_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.layer_scalar": "model-00009-of-00013.safetensors",
+    "layers.39.mlp.down_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.mlp.gate_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.mlp.up_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.post_attention_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.post_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.pre_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.k_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.k_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.o_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.q_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.q_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.39.self_attn.v_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.4.input_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.layer_scalar": "model-00002-of-00013.safetensors",
+    "layers.4.mlp.down_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.mlp.gate_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.mlp.up_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.post_attention_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.post_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.pre_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.k_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.k_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.o_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.q_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.q_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.4.self_attn.v_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.40.input_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.layer_scalar": "model-00009-of-00013.safetensors",
+    "layers.40.mlp.down_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.mlp.gate_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.mlp.up_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.post_attention_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.post_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.pre_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.k_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.k_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.o_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.q_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.q_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.40.self_attn.v_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.input_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.layer_scalar": "model-00009-of-00013.safetensors",
+    "layers.41.mlp.down_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.mlp.gate_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.mlp.up_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.post_attention_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.post_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.pre_feedforward_layernorm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.self_attn.k_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.self_attn.k_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.self_attn.o_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.41.self_attn.q_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.41.self_attn.q_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.42.input_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.42.layer_scalar": "model-00009-of-00013.safetensors",
+    "layers.42.mlp.down_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.42.mlp.gate_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.42.mlp.up_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.42.post_attention_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.42.post_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.42.pre_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.42.self_attn.k_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.42.self_attn.k_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.42.self_attn.o_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.42.self_attn.q_norm.weight": "model-00009-of-00013.safetensors",
+    "layers.42.self_attn.q_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.42.self_attn.v_proj.weight": "model-00009-of-00013.safetensors",
+    "layers.43.input_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.layer_scalar": "model-00010-of-00013.safetensors",
+    "layers.43.mlp.down_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.mlp.gate_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.mlp.up_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.post_attention_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.post_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.pre_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.k_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.k_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.o_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.q_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.q_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.43.self_attn.v_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.input_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.layer_scalar": "model-00010-of-00013.safetensors",
+    "layers.44.mlp.down_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.mlp.gate_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.mlp.up_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.post_attention_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.post_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.pre_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.k_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.k_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.o_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.q_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.q_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.44.self_attn.v_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.input_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.layer_scalar": "model-00010-of-00013.safetensors",
+    "layers.45.mlp.down_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.mlp.gate_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.mlp.up_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.post_attention_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.post_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.pre_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.k_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.k_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.o_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.q_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.q_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.45.self_attn.v_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.input_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.layer_scalar": "model-00010-of-00013.safetensors",
+    "layers.46.mlp.down_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.mlp.gate_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.mlp.up_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.post_attention_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.post_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.pre_feedforward_layernorm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.k_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.k_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.o_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.q_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.q_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.46.self_attn.v_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.47.input_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.47.layer_scalar": "model-00010-of-00013.safetensors",
+    "layers.47.mlp.down_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.47.mlp.gate_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.47.mlp.up_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.47.post_attention_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.47.post_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.47.pre_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.47.self_attn.k_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.47.self_attn.k_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.47.self_attn.o_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.47.self_attn.q_norm.weight": "model-00010-of-00013.safetensors",
+    "layers.47.self_attn.q_proj.weight": "model-00010-of-00013.safetensors",
+    "layers.48.input_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.layer_scalar": "model-00011-of-00013.safetensors",
+    "layers.48.mlp.down_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.mlp.gate_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.mlp.up_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.post_attention_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.post_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.pre_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.k_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.k_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.o_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.q_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.q_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.48.self_attn.v_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.input_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.layer_scalar": "model-00011-of-00013.safetensors",
+    "layers.49.mlp.down_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.mlp.gate_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.mlp.up_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.post_attention_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.post_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.pre_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.k_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.k_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.o_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.q_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.q_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.49.self_attn.v_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.5.input_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.layer_scalar": "model-00002-of-00013.safetensors",
+    "layers.5.mlp.down_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.5.mlp.gate_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.5.mlp.up_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.5.post_attention_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.post_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.pre_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.self_attn.k_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.self_attn.k_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.5.self_attn.o_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.5.self_attn.q_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.5.self_attn.q_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.50.input_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.layer_scalar": "model-00011-of-00013.safetensors",
+    "layers.50.mlp.down_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.mlp.gate_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.mlp.up_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.post_attention_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.post_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.pre_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.k_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.k_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.o_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.q_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.q_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.50.self_attn.v_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.input_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.layer_scalar": "model-00011-of-00013.safetensors",
+    "layers.51.mlp.down_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.mlp.gate_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.mlp.up_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.post_attention_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.post_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.pre_feedforward_layernorm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.k_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.k_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.o_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.q_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.q_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.51.self_attn.v_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.52.input_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.52.layer_scalar": "model-00011-of-00013.safetensors",
+    "layers.52.mlp.down_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.52.mlp.gate_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.52.mlp.up_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.52.post_attention_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.52.post_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.52.pre_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.52.self_attn.k_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.52.self_attn.k_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.52.self_attn.o_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.52.self_attn.q_norm.weight": "model-00011-of-00013.safetensors",
+    "layers.52.self_attn.q_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.52.self_attn.v_proj.weight": "model-00011-of-00013.safetensors",
+    "layers.53.input_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.layer_scalar": "model-00012-of-00013.safetensors",
+    "layers.53.mlp.down_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.53.mlp.gate_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.53.mlp.up_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.53.post_attention_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.post_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.pre_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.self_attn.k_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.self_attn.k_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.53.self_attn.o_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.53.self_attn.q_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.53.self_attn.q_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.input_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.layer_scalar": "model-00012-of-00013.safetensors",
+    "layers.54.mlp.down_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.mlp.gate_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.mlp.up_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.post_attention_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.post_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.pre_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.k_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.k_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.o_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.q_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.q_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.54.self_attn.v_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.input_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.layer_scalar": "model-00012-of-00013.safetensors",
+    "layers.55.mlp.down_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.mlp.gate_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.mlp.up_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.post_attention_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.post_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.pre_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.k_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.k_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.o_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.q_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.q_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.55.self_attn.v_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.input_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.layer_scalar": "model-00012-of-00013.safetensors",
+    "layers.56.mlp.down_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.mlp.gate_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.mlp.up_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.post_attention_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.post_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.pre_feedforward_layernorm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.k_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.k_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.o_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.q_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.q_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.56.self_attn.v_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.57.input_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.57.layer_scalar": "model-00012-of-00013.safetensors",
+    "layers.57.mlp.down_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.57.mlp.gate_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.57.mlp.up_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.57.post_attention_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.57.post_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.57.pre_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.57.self_attn.k_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.57.self_attn.k_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.57.self_attn.o_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.57.self_attn.q_norm.weight": "model-00012-of-00013.safetensors",
+    "layers.57.self_attn.q_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.57.self_attn.v_proj.weight": "model-00012-of-00013.safetensors",
+    "layers.58.input_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.layer_scalar": "model-00013-of-00013.safetensors",
+    "layers.58.mlp.down_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.mlp.gate_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.mlp.up_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.post_attention_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.post_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.pre_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.k_norm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.k_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.o_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.q_norm.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.q_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.58.self_attn.v_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.input_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.layer_scalar": "model-00013-of-00013.safetensors",
+    "layers.59.mlp.down_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.mlp.gate_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.mlp.up_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.post_attention_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.post_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.pre_feedforward_layernorm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.self_attn.k_norm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.self_attn.k_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.self_attn.o_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.59.self_attn.q_norm.weight": "model-00013-of-00013.safetensors",
+    "layers.59.self_attn.q_proj.weight": "model-00013-of-00013.safetensors",
+    "layers.6.input_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.layer_scalar": "model-00002-of-00013.safetensors",
+    "layers.6.mlp.down_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.mlp.gate_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.mlp.up_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.post_attention_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.post_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.pre_feedforward_layernorm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.k_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.k_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.o_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.q_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.q_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.6.self_attn.v_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.7.input_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.7.layer_scalar": "model-00002-of-00013.safetensors",
+    "layers.7.mlp.down_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.7.mlp.gate_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.7.mlp.up_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.7.post_attention_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.7.post_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.7.pre_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.7.self_attn.k_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.7.self_attn.k_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.7.self_attn.o_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.7.self_attn.q_norm.weight": "model-00002-of-00013.safetensors",
+    "layers.7.self_attn.q_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.7.self_attn.v_proj.weight": "model-00002-of-00013.safetensors",
+    "layers.8.input_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.layer_scalar": "model-00003-of-00013.safetensors",
+    "layers.8.mlp.down_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.mlp.gate_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.mlp.up_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.post_attention_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.post_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.pre_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.k_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.k_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.o_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.q_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.q_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.8.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.input_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.layer_scalar": "model-00003-of-00013.safetensors",
+    "layers.9.mlp.down_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.mlp.gate_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.mlp.up_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.post_attention_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.post_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.pre_feedforward_layernorm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.k_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.k_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.o_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.q_norm.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.q_proj.weight": "model-00003-of-00013.safetensors",
+    "layers.9.self_attn.v_proj.weight": "model-00003-of-00013.safetensors",
+    "norm.weight": "model-00013-of-00013.safetensors"
+  }
+}

backbone/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cc8d3a0ce36466ccc1278bf987df5f71db1719b9ca6b4118264f45cb627bfe0f
+size 32169626

backbone/tokenizer_config.json

@@ -0,0 +1,142 @@
+{
+  "audio_token": "<|audio|>",
+  "backend": "tokenizers",
+  "boa_token": "<|audio>",
+  "boi_token": "<|image>",
+  "bos_token": "<bos>",
+  "eoa_token": "<audio|>",
+  "eoc_token": "<channel|>",
+  "eoi_token": "<image|>",
+  "eos_token": "<eos>",
+  "eot_token": "<turn|>",
+  "escape_token": "<|\"|>",
+  "etc_token": "<tool_call|>",
+  "etd_token": "<tool|>",
+  "etr_token": "<tool_response|>",
+  "extra_special_tokens": [
+    "<|video|>"
+  ],
+  "image_token": "<|image|>",
+  "is_local": false,
+  "local_files_only": true,
+  "mask_token": "<mask>",
+  "model_max_length": 1000000000000000019884624838656,
+  "model_specific_special_tokens": {
+    "audio_token": "<|audio|>",
+    "boa_token": "<|audio>",
+    "boi_token": "<|image>",
+    "eoa_token": "<audio|>",
+    "eoc_token": "<channel|>",
+    "eoi_token": "<image|>",
+    "eot_token": "<turn|>",
+    "escape_token": "<|\"|>",
+    "etc_token": "<tool_call|>",
+    "etd_token": "<tool|>",
+    "etr_token": "<tool_response|>",
+    "image_token": "<|image|>",
+    "soc_token": "<|channel>",
+    "sot_token": "<|turn>",
+    "stc_token": "<|tool_call>",
+    "std_token": "<|tool>",
+    "str_token": "<|tool_response>",
+    "think_token": "<|think|>"
+  },
+  "pad_token": "<pad>",
+  "padding_side": "left",
+  "processor_class": "Gemma4Processor",
+  "response_schema": {
+    "properties": {
+      "content": {
+        "type": "string"
+      },
+      "role": {
+        "const": "assistant"
+      },
+      "thinking": {
+        "type": "string"
+      },
+      "tool_calls": {
+        "items": {
+          "properties": {
+            "function": {
+              "properties": {
+                "arguments": {
+                  "additionalProperties": {},
+                  "type": "object",
+                  "x-parser": "gemma4-tool-call"
+                },
+                "name": {
+                  "type": "string"
+                }
+              },
+              "type": "object",
+              "x-regex": "call\\:(?P<name>\\w+)(?P<arguments>\\{.*\\})"
+            },
+            "type": {
+              "const": "function"
+            }
+          },
+          "type": "object"
+        },
+        "type": "array",
+        "x-regex-iterator": "<\\|tool_call>(.*?)<tool_call\\|>"
+      }
+    },
+    "type": "object",
+    "x-regex": "(\\<\\|channel\\>thought\\n(?P<thinking>.*?)\\<channel\\|\\>)?(?P<tool_calls>\\<\\|tool_call\\>.*\\<tool_call\\|\\>)?(?P<content>(?:(?!\\<turn\\|\\>)(?!\\<\\|tool_response\\>).)+)?(?:\\<turn\\|\\>|\\<\\|tool_response\\>)?"
+  },
+  "response_template": {
+    "defaults": {
+      "role": "assistant"
+    },
+    "fields": {
+      "content": {
+        "close": [
+          "<turn|>",
+          "<|tool_response>",
+          "<eos>"
+        ],
+        "content": "text"
+      },
+      "thinking": {
+        "close": "<channel|>",
+        "content": "text",
+        "open": "<|channel>thought\n"
+      },
+      "tool_calls": {
+        "close": "<tool_call|>",
+        "content": "json",
+        "content_args": {
+          "string_delims": [
+            [
+              "<|\"|>",
+              "<|\"|>"
+            ]
+          ],
+          "unquoted_keys": true
+        },
+        "open_pattern": "<\\|tool_call>call:(?P<name>\\w+)",
+        "repeats": true,
+        "transform": {
+          "function": {
+            "arguments": "{content}",
+            "name": "{name}"
+          },
+          "type": "function"
+        }
+      }
+    },
+    "start_anchor": [
+      "<|turn>model\n",
+      "<tool_response|>"
+    ]
+  },
+  "soc_token": "<|channel>",
+  "sot_token": "<|turn>",
+  "stc_token": "<|tool_call>",
+  "std_token": "<|tool>",
+  "str_token": "<|tool_response>",
+  "think_token": "<|think|>",
+  "tokenizer_class": "GemmaTokenizer",
+  "unk_token": "<unk>"
+}

config.json

@@ -0,0 +1,14 @@
+{
+  "readout": "letters",
+  "backbone": "google/gemma-4-31B-it",
+  "backbone_revision": "842da3794eaa0b77d5f08bae87a17459d91ff475",
+  "max_state_tokens": 16384,
+  "max_question_tokens": 16384,
+  "max_request_tokens": 32768,
+  "temperature_by_options": {
+    "a": 1.25,
+    "b": -0.05,
+    "min": 0.05
+  },
+  "version": "agr-1"
+}