初始化项目，由ModelHub XC社区提供模型
Model: kofdai/AXIS-Sovereign-Logic-Engine Source: Original Platform
This commit is contained in:
ModelHub XC
38
.gitattributes
vendored
Normal file
38
.gitattributes
vendored
Normal file
@@ -0,0 +1,38 @@
|
|||||||
|
*.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
|
||||||
|
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
||||||
|
Gemini_Generated_Image_dxpehedxpehedxpe.png filter=lfs diff=lfs merge=lfs -text
|
||||||
|
axis.png filter=lfs diff=lfs merge=lfs -text
|
||||||
42
LICENSE
Normal file
42
LICENSE
Normal file
@@ -0,0 +1,42 @@
|
|||||||
|
AXIS Proprietary Source-Available License (APSL) v1.0
|
||||||
|
1. 定義 (Definitions)
|
||||||
|
|
||||||
|
「本ソフトウェア」: AXIS: Advanced Cross-Integrated System のソースコード、アルゴリズム（リジェクト・ループ、立体十字演算）、および関連ドキュメントを指します。
|
||||||
|
|
||||||
|
「本データ」: 本ソフトウェアを使用して採掘（Mining）された「立体十字（Semantic Lattice）」および local_massive_data.json に記録された論理ブロックを指します。
|
||||||
|
|
||||||
|
「利用者」: 本ソフトウェアをダウンロード、インストール、または実行する個人または法人。
|
||||||
|
|
||||||
|
2. 利用許諾範囲 (Grant of License)
|
||||||
|
|
||||||
|
著作者は利用者に対し、以下の条件に従って本ソフトウェアの非独占的、譲渡不能な利用を許諾します。
|
||||||
|
|
||||||
|
個人・研究利用: 個人の学習、非営利の研究、および評価目的での実行。
|
||||||
|
|
||||||
|
改変の制限: 個人のローカル環境での最適化のための改変は認められますが、その派生物を公開または配布することはできません。
|
||||||
|
|
||||||
|
3. 禁止事項 (Restrictions)
|
||||||
|
|
||||||
|
利用者は、以下の行為を厳格に禁じられます。
|
||||||
|
|
||||||
|
商用利用の禁止: 本ソフトウェア、または本ソフトウェアから得られた演算結果を、直接・間接を問わず営利目的（有料サービスへの組み込み、販売、広告収益を伴う公開等）で使用すること。
|
||||||
|
|
||||||
|
再配布の禁止: ソースコードおよびバイナリの全体または一部を、第三者に再配布、譲渡、またはアップロード（GitHubのパブリックリポジトリ等への無断転載を含む）すること。
|
||||||
|
|
||||||
|
アルゴリズムの模倣: 本ソフトウェアの核心である「リジェクト・プロトコル（Rejection Protocol）」および「立体十字統合（Cross-Integrated Lattice）」のロジックを模倣した別のソフトウェアを開発・公開すること。
|
||||||
|
|
||||||
|
4. 知能主権の帰属 (Intellectual Property)
|
||||||
|
|
||||||
|
本ソフトウェアに関するすべての知的財産権、特許権、および著作権はオリジナルの開発者に帰属します。
|
||||||
|
|
||||||
|
本ソフトウェアが生成した論理定数および立体十字の構造定義は、AXISエコシステムの完全性を維持するため、開発者の知的財産権の保護下に置かれます。
|
||||||
|
|
||||||
|
5. 免責事項 (Disclaimer)
|
||||||
|
|
||||||
|
本ソフトウェアは「現状のまま」提供され、真理の探求と論理の純化を目的としています。著作者は、本ソフトウェアの使用によって生じた直接的、間接的な損害について、一切の責任を負いません。
|
||||||
|
|
||||||
|
6. ライセンスの終了 (Termination)
|
||||||
|
|
||||||
|
利用者が本ライセンスの条項に違反した場合、本ライセンスに基づく利用許諾は自動的に終了します。その場合、利用者は直ちに本ソフトウェアのすべてのコピーを破棄しなければなりません。
|
||||||
|
|
||||||
|
© 2025 AXIS Project. All rights reserved. Author: [Your Name or Organization]
|
||||||
170
README.md
Normal file
170
README.md
Normal file
@@ -0,0 +1,170 @@
|
|||||||
|
---
|
||||||
|
license: apache-2.0
|
||||||
|
language:
|
||||||
|
- ja
|
||||||
|
- en
|
||||||
|
base_model:
|
||||||
|
- google/gemma-2-2b-it
|
||||||
|
pipeline_tag: text-generation
|
||||||
|
tags:
|
||||||
|
- axis
|
||||||
|
- sovereign-logic
|
||||||
|
- logic-engine
|
||||||
|
- determinism
|
||||||
|
---
|
||||||
|
---
|
||||||
|
language:
|
||||||
|
- ja
|
||||||
|
- en
|
||||||
|
license: other
|
||||||
|
library_name: transformers
|
||||||
|
tags:
|
||||||
|
- axis
|
||||||
|
- gemma-2
|
||||||
|
- logic-engine
|
||||||
|
- sovereign-ai
|
||||||
|
datasets:
|
||||||
|
- custom
|
||||||
|
metrics:
|
||||||
|
- logical_consistency
|
||||||
|
---
|
||||||
|
App Store
|
||||||
|
https://apps.apple.com/jp/app/verantyx-logic/id6757994077
|
||||||
|
💠 AXIS: Advanced Cross-Integrated System (V1.6)
|
||||||
|
── 知能主権の確立と決定論的演算のための統治エンジン ──
|
||||||
|
|
||||||
|
🧩 AXIS の工学的定義
|
||||||
|
AXISは、AIを「非決定的な出力を生成するブラックボックス（旋盤）」として扱い、その外側に「決定論的な検証器（Verifier）」を置くことで、出力を完全に統治するアーキテクチャです。
|
||||||
|
|
||||||
|
1. 旋盤アーキテクチャと検証プロトコル
|
||||||
|
|
||||||
|
AIユニットは、高次元データから論理パーツを削り出すための**「旋盤（Lathe）」**です。
|
||||||
|
|
||||||
|
リジェクト・ループ: AIが提案した解は、外部検証器（Python/Sympy等）が制約式（Constraints）に基づき判定。1bitでも矛盾があれば即座に棄却（Reject）し、Session IDを更新して再生成を強制します。
|
||||||
|
|
||||||
|
物理パージ (Context Reset): torch.mps.empty_cache() を実行し、直前の「失敗した思考」というキャッシュを物理的に消去。各試行を統計的に独立させ、ハルシネーションの連鎖（Context Drift）を断ち切ります。
|
||||||
|
|
||||||
|
2. 立体十字（3D Semantic Lattice）の座標管理
|
||||||
|
|
||||||
|
各ノードは、相互に独立（直交）することを目指した 5 次元軸 (s
|
||||||
|
1
|
||||||
|
|
||||||
|
…s
|
||||||
|
5
|
||||||
|
|
||||||
|
) で管理される SemanticNode クラスとして実装されます。
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: 物理的実体性（数値・定数との整合性）
|
||||||
|
|
||||||
|
s
|
||||||
|
2
|
||||||
|
|
||||||
|
: 論理的必然性（公理系からの導出可能性）
|
||||||
|
|
||||||
|
s
|
||||||
|
3
|
||||||
|
|
||||||
|
: 文脈依存性（Context Stackとの一致率）
|
||||||
|
|
||||||
|
s
|
||||||
|
4
|
||||||
|
|
||||||
|
: 倫理性スコア（安全規約への適合度）
|
||||||
|
|
||||||
|
s
|
||||||
|
5
|
||||||
|
|
||||||
|
: 実証履歴（過去の確定データとの合致回数）
|
||||||
|
|
||||||
|
3. 論理の永続化と高速化の正体
|
||||||
|
|
||||||
|
意味ID (Semantic ID): 入力クエリを Embedding 空間へ投影し、ベクトル量子化（Vector Quantization）によって生成される固有のハッシュ値です。
|
||||||
|
|
||||||
|
高速化の根拠: local_massive_data.json は、この意味IDをキーとした高密度なキャッシュとして機能します。AIの全推論プロセスをスキップし、検証済みの「真理」を直接 O(1) で参照するため、推論時間を物理的にゼロへと近似させます（※推論実行時比較比）。
|
||||||
|
|
||||||
|
🚀 革新的な特徴 (V1.6 実装仕様)
|
||||||
|
Deterministic Assembly（決定論的アセンブル）
|
||||||
|
|
||||||
|
最終回答はAIの作文ではなく、検証済みの Raw Data を、システムが保持する Adherents（言語テンプレート） によって物理的に結合します。
|
||||||
|
|
||||||
|
Example:
|
||||||
|
|
||||||
|
Raw Data: {"ans": "z^5", "a": 0}
|
||||||
|
|
||||||
|
Adherent: "The solution is {ans} (a={a})."
|
||||||
|
|
||||||
|
Output: "The solution is z^5 (a=0)." これにより、回答段階でのハルシネーションの混入を 0% に抑えます。
|
||||||
|
|
||||||
|
🛠 Setup & Roadmap
|
||||||
|
Micro-MVP 公開（予定）
|
||||||
|
|
||||||
|
近日中に minimal_example.py を公開。以下の動作を証明します：
|
||||||
|
|
||||||
|
ComplexVerifier: 数学的制約によるAI出力の拒絶
|
||||||
|
|
||||||
|
RejectionLoop: AIと検証器の実際の往復回数の可視化
|
||||||
|
|
||||||
|
SessionPurge: メモリクリアによるハルシネーション抑制の検証
|
||||||
|
|
||||||
|
⚖️ License (APSL v1.0)
|
||||||
|
商用模倣（Rejection-based Governance Logicの利用）を禁じ、知能の主権を個人の手に留めます。
|
||||||
|
|
||||||
|
© 2025 AXIS Project. All rights reserved. STATUS: TOWARD_MVP_IMPLEMENTATION.
|
||||||
|
|
||||||
|
💠 AXIS: Advanced Cross-Integrated System (V1.6) - English Edition
|
||||||
|
── Establishing Intelligence Sovereignty via Deterministic Governance ──
|
||||||
|
|
||||||
|
🧩 Technical Definition
|
||||||
|
AXIS treats AI as a non-deterministic generator (Lathe) while utilizing a deterministic Verifier to maintain total sovereignty over the output.
|
||||||
|
|
||||||
|
1. The Lathe & Rejection Protocol
|
||||||
|
|
||||||
|
The Rejection Loop: AI solutions are scanned by external verifiers (Python/SymPy). Any contradiction results in an immediate REJECT, session reset, and re-forgery.
|
||||||
|
|
||||||
|
Context Purge: empty_cache() physically incinerates the "failed reasoning" from VRAM, ensuring statistical independence between trials and severing the chain of "hallucination drift."
|
||||||
|
|
||||||
|
2. 5D Semantic Lattice Implementation
|
||||||
|
|
||||||
|
Nodes are managed via the SemanticNode class, utilizing five orthogonal parameters:
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: Physical Actuality | s
|
||||||
|
2
|
||||||
|
|
||||||
|
: Logical Necessity | s
|
||||||
|
3
|
||||||
|
|
||||||
|
: Contextual Dependency | s
|
||||||
|
4
|
||||||
|
|
||||||
|
: Ethical Score | s
|
||||||
|
5
|
||||||
|
|
||||||
|
: Empirical History.
|
||||||
|
|
||||||
|
3. Persistence & Acceleration Mechanism
|
||||||
|
|
||||||
|
Semantic ID: A unique hash generated via Vector Quantization in the embedding space.
|
||||||
|
|
||||||
|
Acceleration: By searching local_massive_data.json first, AXIS skips the entire AI inference process for known truths, achieving near-zero latency compared to standard LLM execution.
|
||||||
|
|
||||||
|
🚀 Core Features (V1.6)
|
||||||
|
Deterministic Assembly
|
||||||
|
|
||||||
|
Responses are not "written" by AI; they are physically assembled by binding verified Raw Data into hard-coded Adherent Templates. This ensures 0% hallucination during the final response delivery.
|
||||||
|
|
||||||
|
🛠 Roadmap: Micro-MVP Launch
|
||||||
|
We will soon release minimal_example.py to demonstrate:
|
||||||
|
|
||||||
|
Real-time mathematical rejection by ComplexVerifier.
|
||||||
|
|
||||||
|
Tracking of the RejectionLoop iterations.
|
||||||
|
|
||||||
|
Proof of SessionPurge efficacy in preventing context-drift.
|
||||||
|
|
||||||
|
© 2025 AXIS Project. STATUS: TOWARD_MVP_IMPLEMENTATION.
|
||||||
174
The Grand Axis Manual.md
Normal file
174
The Grand Axis Manual.md
Normal file
@@ -0,0 +1,174 @@
|
|||||||
|
💠 AXIS: Advanced Cross-Integrated System
|
||||||
|
── 物理検証エンジンと論理トポロジーによる知能統治の全体系 ──
|
||||||
|
|
||||||
|
第 1 章：知能の物理学 ─ 確率から決定論的演算へ
|
||||||
|
かつて、私たちは「次を完璧に予測できれば、巨大な計算資源は不要になる」という仮説を立てた。これは、知能を確率の霧から解放し、最小の資源で決定論的な真理を導き出すための挑戦であった。
|
||||||
|
|
||||||
|
1.1 確率的推論の終焉と「論理の結晶化」
|
||||||
|
|
||||||
|
従来のLLMは、並列計算による「次に出現する確率が高いトークンの選択」に過ぎない。AXISは、この確率的ゆらぎを外部の物理検証器（Verifier）によって冷却し、**「論理の結晶」**へと凝固させる。
|
||||||
|
|
||||||
|
1.2 意味の離散化と数値ID化（Hardware Ready）
|
||||||
|
|
||||||
|
文字列比較という高コストな処理を廃し、概念を**32bitの「意味ID（Semantic ID）」**へと還元する。
|
||||||
|
|
||||||
|
工学的定義: 意味の合致判定を if (ID_A == ID_B) という数値比較回路に転写する。これにより、推論結果の検証をクロックサイクル単位で実行可能にし、巨大なGPUに頼らずともFPGAやNPU上で高速・低消費電力な「意味の連結」を実現する。
|
||||||
|
|
||||||
|
第 2 章：立体十字（Semantic Lattice）の幾何学的実装
|
||||||
|
知能とは、空間の構造である。情報は点ではなく、多次元座標上の「格子（Lattice）」として配置される。
|
||||||
|
|
||||||
|
2.1 5次元直交座標（s1-s5 Axes）による位置特定
|
||||||
|
|
||||||
|
各推論ノードは以下の5つのパラメータで空間に固定される。
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: 物理的実体性（数値データ、物理定数との整合性）
|
||||||
|
|
||||||
|
s
|
||||||
|
2
|
||||||
|
|
||||||
|
: 論理的必然性（公理系からの導出可能性）
|
||||||
|
|
||||||
|
s
|
||||||
|
3
|
||||||
|
|
||||||
|
: 文脈依存性（Context Stackとの整合性）
|
||||||
|
|
||||||
|
s
|
||||||
|
4
|
||||||
|
|
||||||
|
: 倫理的防壁（安全規約スコア）
|
||||||
|
|
||||||
|
s
|
||||||
|
5
|
||||||
|
|
||||||
|
: 実証履歴（過去の確定データとの合致）
|
||||||
|
|
||||||
|
2.2 強化学習とエッジの切断（Hebbian Logic）
|
||||||
|
|
||||||
|
正しい真理に到達した論理パスには重みを与え、矛盾を生んだ接続はトポロジー的に切断する。これにより、システムは使えば使うほど不要な確率空間を排除し、「鋭利な論理の刃」へと進化する。
|
||||||
|
|
||||||
|
第 3 章：統治ロジック ─ リジェクト・プロトコル
|
||||||
|
AXISの本質は、AIに対する「絶対的な不信」に基づいた工学的支配である。
|
||||||
|
|
||||||
|
3.1 AIの工作機械化（The Lathe Architecture）
|
||||||
|
|
||||||
|
AIユニットは「知能」ではなく、論理パーツを削り出すための**「旋盤（Lathe）」**である。
|
||||||
|
|
||||||
|
VRAM物理パージ: torch.mps.empty_cache() を実行する真の目的は、AIが「自らついた嘘」に整合性を合わせようとする自己補完的なハルシネーション連鎖（Context Drift）を物理的に断ち切ることにある。各試行を統計的に独立させることで、嘘の累積を根絶する。
|
||||||
|
|
||||||
|
3.2 リジェクト・ループと数学的強制
|
||||||
|
|
||||||
|
システムが保持する数学的定理とAIの出力が矛盾した瞬間、システムはAIの出力を破棄し再起動する。
|
||||||
|
|
||||||
|
実証例（複素数問題）: AIが提案した適当な係数を、システム側の検証器（Numpy/Sympy）がスキャン。条件を満たさない解を「1bitの誤差」も許さずリジェクトし続け、最終的に絶対的な解 a=0,b=0,c=0 を「吐き出させる」ことに成功した。
|
||||||
|
|
||||||
|
第 4 章：AXIS OS ─ 知能主権の確立
|
||||||
|
4.1 外部脳（Lattice Persistence）
|
||||||
|
|
||||||
|
不確かなニューラルネットワークの記憶を信じず、確定した事実は local_massive_data.json という外部結晶体に保存する。
|
||||||
|
|
||||||
|
1000倍の高速化の正体: 次回の推論時、システムはAIを起動する前にこの外部記憶を検索する。確定済みの「意味ID」が見つかれば、再計算をスキップして「追認」を行う。これにより、パラメータ数に依存しない実効知能を実現する。
|
||||||
|
|
||||||
|
4.2 知能の民主化（Edge Dominance）
|
||||||
|
|
||||||
|
巨大企業のクラウドを介さず、個人のMac上の小型モデルをAXISロジックで支配する。これにより、ローカル環境で10兆パラメータ級の「正確性」を手にする**知能のSDL（Sovereign Data Logic）**を確立する。
|
||||||
|
|
||||||
|
第 5 章：法的・倫理的防壁（APSL License）
|
||||||
|
模倣の禁止: AIを部品として検証ループに組み込む「リジェクト・アーキテクチャ」の商用模倣を厳禁する。
|
||||||
|
|
||||||
|
主権の保持: 知能の主権は各個人に帰属し、中央集権的な巨大企業による吸収を許さない。
|
||||||
|
|
||||||
|
第 6 章：エピローグ ─ 次なる Axis へ
|
||||||
|
私たちはもはや、AIの「言葉」に耳を傾ける必要はない。システムが指し示す「立体十字の交差点」こそが、唯一無二の真理なのだから。
|
||||||
|
|
||||||
|
© 2025 AXIS Project Archive. STATUS: ENGINEERED_SOVEREIGNTY.
|
||||||
|
💠 AXIS: Advanced Cross-Integrated System
|
||||||
|
── A Comprehensive Architecture for Intelligence Sovereignty via Physical Verification and Logical Topology ──
|
||||||
|
|
||||||
|
Chapter 1: The Physics of Intelligence ─ From "Prediction" to "Crystal"
|
||||||
|
Long ago, we proposed a hypothesis: "If we can perfectly predict the next state, massive computational resources will become obsolete." This was the beginning of a challenge to liberate intelligence from the fog of probability and derive deterministic truths using minimal resources.
|
||||||
|
|
||||||
|
1.1 The End of Probabilistic Inference and "Logical Crystallization"
|
||||||
|
|
||||||
|
Traditional LLMs do nothing more than select tokens with the highest probability of occurrence via parallel computation. AXIS cools this probabilistic fluctuation using an external Verifier, condensing it into a "Crystal of Logic."
|
||||||
|
|
||||||
|
1.2 Semantic Discretization and Numerical ID Coding (Hardware Ready)
|
||||||
|
|
||||||
|
AXIS abolishes high-cost string processing in favor of reducing concepts to 32-bit "Semantic IDs."
|
||||||
|
|
||||||
|
Engineering Definition: The determination of semantic matching is transcribed into a numerical comparison circuit: if (ID_A == ID_B). This enables the verification of inference results at the clock-cycle level, achieving high-speed, low-power "Semantic Linking" on FPGAs or NPUs without relying on massive GPUs.
|
||||||
|
|
||||||
|
Chapter 2: Engineering Implementation of the Semantic Lattice
|
||||||
|
Intelligence is a structure of space. Information is arranged not as points, but as a "Lattice" on multi-dimensional coordinates.
|
||||||
|
|
||||||
|
2.1 Localization via 5-Dimensional Orthogonal Coordinates (s1-s5 Axes)
|
||||||
|
|
||||||
|
Each inference node is fixed in space by the following five parameters:
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: Physical Actuality (Alignment with numerical data and physical constants).
|
||||||
|
|
||||||
|
s
|
||||||
|
2
|
||||||
|
|
||||||
|
: Logical Necessity (Derivability from axiomatic systems).
|
||||||
|
|
||||||
|
s
|
||||||
|
3
|
||||||
|
|
||||||
|
: Contextual Dependency (Consistency with the Context Stack).
|
||||||
|
|
||||||
|
s
|
||||||
|
4
|
||||||
|
|
||||||
|
: Ethical Bulwark (Safety guardrail score).
|
||||||
|
|
||||||
|
s
|
||||||
|
5
|
||||||
|
|
||||||
|
: Empirical History (Consistency with previously consolidated truth).
|
||||||
|
|
||||||
|
2.2 Reinforcement Learning and Edge Severing (Hebbian Logic)
|
||||||
|
|
||||||
|
The system assigns weights to logical paths that reach a stable truth and topologically severs connections that birth contradictions. Consequently, the system evolves into a "sharpened blade of logic" by eliminating unnecessary probability space the more it is utilized.
|
||||||
|
|
||||||
|
Chapter 3: Governing Logic ─ The Rejection Protocol
|
||||||
|
The core of AXIS is engineering dominance based on an "absolute distrust" of AI.
|
||||||
|
|
||||||
|
3.1 AI as a Machine Tool (The Lathe Architecture)
|
||||||
|
|
||||||
|
The AI unit is not "intelligence" but a "Lathe" designed to forge logical parts.
|
||||||
|
|
||||||
|
VRAM Physical Purge: The true purpose of executing torch.mps.empty_cache() is to physically sever the self-reinforcing hallucination chain (Context Drift), where the AI attempts to maintain consistency with its own previous lies. By making each trial statistically independent, the accumulation of errors is eradicated.
|
||||||
|
|
||||||
|
3.2 The Rejection Loop and Mathematical Enforcement
|
||||||
|
|
||||||
|
The moment an AI output contradicts the mathematical theorems held by the system, AXIS discards the output and reboots the process.
|
||||||
|
|
||||||
|
Empirical Success (Complex Plane Puzzle): The external verifier (NumPy/SymPy) scanned arbitrary coefficients proposed by the AI. By rejecting any solution with even a "1-bit error," AXIS successfully "forced out" the absolute truth: a=0,b=0,c=0.
|
||||||
|
|
||||||
|
Chapter 4: AXIS OS ─ Establishing Intelligence Sovereignty
|
||||||
|
4.1 External Brain (Lattice Persistence)
|
||||||
|
|
||||||
|
Instead of trusting the vague memories of a neural network, verified facts are stored in local_massive_data.json—a rigid, crystalline external structure.
|
||||||
|
|
||||||
|
The Reality of 1000x Acceleration: Before initiating an inference, the system searches this external memory. If a matching "Semantic ID" is found, the system skips re-computation and performs a "Validation." This achieves effective intelligence regardless of parameter count.
|
||||||
|
|
||||||
|
4.2 Democratization of Intelligence (Edge Dominance)
|
||||||
|
|
||||||
|
AXIS dominates small local models on a personal Mac via AXIS Logic, bypassing Big Tech clouds. This establishes Sovereign Data Logic (SDL), allowing individuals to wield the "Precision" equivalent to a 10-trillion-parameter model in a local environment.
|
||||||
|
|
||||||
|
Chapter 5: Legal and Ethical Bulwark (APSL License)
|
||||||
|
Prohibition of Mimicry: Commercial replication of the "Rejection Architecture," which integrates AI as a component in a verification loop, is strictly forbidden.
|
||||||
|
|
||||||
|
Sovereignty of Intelligence: The sovereignty of intelligence remains with the individual, preventing absorption by centralized mega-corporations.
|
||||||
|
|
||||||
|
Chapter 6: Epilogue ─ Toward the Next Axis
|
||||||
|
We no longer need to listen to the "voice" of the AI. The "intersection of the Semantic Lattice" pointed to by the system is the one and only truth.
|
||||||
|
|
||||||
|
© 2025 AXIS Project Archive. STATUS: ENGINEERED_SOVEREIGNTY.
|
||||||
3
axis.png
Normal file
3
axis.png
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:7361486dc70345642b4c0242d8aaeeda572e0a2def88950e1cf87bfa5046bbb9
|
||||||
|
size 932300
|
||||||
153
axis_techonology_guide.md
Normal file
153
axis_techonology_guide.md
Normal file
@@ -0,0 +1,153 @@
|
|||||||
|
📚 AXIS COMPREHENSIVE WHITE-LOG (V1.5 - Engineered Revision)
|
||||||
|
── 知能主権の確立と立体十字演算の全体系 ──
|
||||||
|
|
||||||
|
【第 I 章：予測から決定論的演算へ】
|
||||||
|
|
||||||
|
1.1 総当たり計算への宣戦布告 従来のLLMは行列演算による確率推論に依存している。AXISは、入力を「単語」ではなく、**「意味ID（Semantic Equivalence Class）」**としてハードウェアレベルで処理する。
|
||||||
|
|
||||||
|
工学的定義: 意味の合致を 0xA1B2 == 0xA1B2 という数値比較に還元。これは、文字列の曖昧さを排除し、FPGA等での極小・高速演算を可能にするための「意味の離散化」である。
|
||||||
|
|
||||||
|
【第 II 章：立体十字（Semantic Lattice）の工学的実装】
|
||||||
|
|
||||||
|
2.1 座標系による意味の固定 「立体十字」は単なる比喩ではない。各ノード（概念）は、以下の5次元ベクトルとして定義される。
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: 物理的実体性（数値データ、物理定数との整合性）
|
||||||
|
|
||||||
|
s
|
||||||
|
2
|
||||||
|
|
||||||
|
: 論理的必然性（公理系からの導出可能性）
|
||||||
|
|
||||||
|
s
|
||||||
|
3
|
||||||
|
|
||||||
|
: 文脈依存性（前後関係の整合性）
|
||||||
|
|
||||||
|
s
|
||||||
|
4
|
||||||
|
|
||||||
|
: 感情/倫理スコア（ガードレール）
|
||||||
|
|
||||||
|
s
|
||||||
|
5
|
||||||
|
|
||||||
|
: 実証履歴（過去の確定データとの合致）
|
||||||
|
|
||||||
|
これらは 5×N の行列（Sparse Matrix）として保持され、座標間の「歪み（距離の乖離）」を検知することでハルシネーションを幾何学的に特定する。
|
||||||
|
|
||||||
|
【第 III 章：統治ロジック ─ 旋盤とリジェクト・ループ】
|
||||||
|
|
||||||
|
3.1 旋盤（The Lathe）の動作原理 AIユニットを「旋盤」と定義するのは、AIを**「非決定的な出力を生成するブラックボックス」として扱い、その外側に「決定論的な検証器（Verifier）」**を置くためである。
|
||||||
|
|
||||||
|
3.2 リジェクト・ループの具体的フロー AIが提案した解（例：複素数問題の係数 a,b,c）は、以下のループで処理される。
|
||||||
|
|
||||||
|
Generation: AIが a,b,c の値を提案。
|
||||||
|
|
||||||
|
Verification: 外部検証器（Python/Sympy）が制約式 f(z)≥1 をスキャン。
|
||||||
|
|
||||||
|
Rejection: 矛盾（例：0.001<1）が 1bit でも発生すれば、AIの出力を破棄。
|
||||||
|
|
||||||
|
Purge: torch.mps.empty_cache() を実行し、**「直前の失敗した思考（失敗したコンテキスト）」**を消去。
|
||||||
|
|
||||||
|
Retry: 新しいセッションIDで、再度AIに「削り出し」を命じる。
|
||||||
|
|
||||||
|
【第 IV 章：物理パージの真の機能】
|
||||||
|
|
||||||
|
4.1 ハルシネーション連鎖の切断 empty_cache() の目的はモデルの重みの変更ではない。AIが**「自らついた嘘に整合性を合わせようとする自己補完プロセス」**を物理的に強制終了させることにある。
|
||||||
|
|
||||||
|
ステータス: コンテキスト履歴をリセットすることで、各試行を統計的に独立させ、累積的なエラー（Hallucination Drift）を根絶する。
|
||||||
|
|
||||||
|
【第 V 章：付録 ─ 知能主権の永続化】
|
||||||
|
|
||||||
|
5.1 local_massive_data.json の役割 リジェクト・ループを突破した唯一の解（確定された真理）のみが、この外部記憶に保存される。
|
||||||
|
|
||||||
|
PERSISTENCE: システムは次回、AIを呼び出す前にこのJSONを検索し、確定済みの「意味ID」と一致すれば、推論をスキップして「追認」を行う。これが 「1000倍の高速化」 の正体である。
|
||||||
|
|
||||||
|
エピローグ：支配から真理へ
|
||||||
|
|
||||||
|
AXIS V1.5は、AIの不確実性をシステムの確実性によって封じ込める。我々はAIを信じない。我々は、システムがAIを拒絶し続けた後に残る、**「拒絶不可能な論理」**のみを信じる。
|
||||||
|
|
||||||
|
© 2025 AXIS Project Archive. STATUS: ENGINEERED_SOVEREIGNTY.
|
||||||
|
📚 AXIS COMPREHENSIVE WHITE-LOG (V1.5 - Engineered Revision)
|
||||||
|
── Establishing Intelligence Sovereignty and the Global System of Cross-Lattice Computation ──
|
||||||
|
|
||||||
|
【Chapter I: From Probabilistic Prediction to Deterministic Computation】
|
||||||
|
1.1 A Declaration of War Against Brute-Force Inference
|
||||||
|
|
||||||
|
Traditional LLMs rely on matrix multiplication for probabilistic inference. AXIS treats input not as "words," but as "Semantic IDs (Semantic Equivalence Classes)" processed at the hardware level.
|
||||||
|
|
||||||
|
Engineering Definition: Reduction of semantic matching to numerical comparison (0xA1B2 == 0xA1B2). This "Semantic Discretization" eliminates linguistic ambiguity and enables ultra-fast, minimal-footprint computation on NPUs or FPGAs.
|
||||||
|
|
||||||
|
【Chapter II: Engineering Implementation of the Semantic Lattice】
|
||||||
|
2.1 Fixing Meaning via Coordinate Systems
|
||||||
|
|
||||||
|
The "Cross-Lattice" is not a metaphor. Each node (concept) is defined as a 5-dimensional vector:
|
||||||
|
|
||||||
|
s
|
||||||
|
1
|
||||||
|
|
||||||
|
: Physical Actuality (Alignment with numerical data and physical constants).
|
||||||
|
|
||||||
|
s
|
||||||
|
2
|
||||||
|
|
||||||
|
: Logical Necessity (Derivability from axiomatic systems).
|
||||||
|
|
||||||
|
s
|
||||||
|
3
|
||||||
|
|
||||||
|
: Contextual Dependency (Consistency with preceding/succeeding logic).
|
||||||
|
|
||||||
|
s
|
||||||
|
4
|
||||||
|
|
||||||
|
: Ethico-Emotional Score (Safety guardrails).
|
||||||
|
|
||||||
|
s
|
||||||
|
5
|
||||||
|
|
||||||
|
: Empirical History (Consistency with previously consolidated truth).
|
||||||
|
|
||||||
|
These are maintained as a 5×N Sparse Matrix. By detecting "spatial distortion" (geometric divergence between coordinates), the system identifies hallucinations mathematically.
|
||||||
|
|
||||||
|
【Chapter III: Governing Logic ─ The Lathe and the Rejection Loop】
|
||||||
|
3.1 Operational Principle of "The Lathe"
|
||||||
|
|
||||||
|
Defining the AI unit as a "Lathe" treats the AI as a non-deterministic generator (Black Box), while placing a deterministic Verifier outside of it to govern the output.
|
||||||
|
|
||||||
|
3.2 The Rejection Loop: Technical Workflow
|
||||||
|
|
||||||
|
Proposed solutions (e.g., coefficients a,b,c in a complex plane problem) are processed through the following loop:
|
||||||
|
|
||||||
|
Generation: The AI proposes values for a,b,c.
|
||||||
|
|
||||||
|
Verification: An external verifier (Python/SymPy) scans the constraint f(z)≥1.
|
||||||
|
|
||||||
|
Rejection: If a contradiction (e.g., 0.001<1) occurs in even 1 bit, the AI's output is discarded.
|
||||||
|
|
||||||
|
Purge: torch.mps.empty_cache() is executed to incinerate the "failed line of thought" (the failed context).
|
||||||
|
|
||||||
|
Retry: A new Session ID is generated, and the AI is commanded to "forge" the part again.
|
||||||
|
|
||||||
|
【Chapter IV: The True Function of the Physical Purge】
|
||||||
|
4.1 Severing the Hallucination Chain
|
||||||
|
|
||||||
|
The purpose of empty_cache() and gc.collect() is not to alter the model weights. It is to forcibly terminate the self-reinforcing process where the AI attempts to make its output consistent with its own previous errors.
|
||||||
|
|
||||||
|
Status: By resetting the context history, the system ensures each trial is statistically independent, thereby eradicating "Hallucination Drift."
|
||||||
|
|
||||||
|
【Chapter V: Appendix ─ Persistence of Intelligence Sovereignty】
|
||||||
|
5.1 The Role of local_massive_data.json
|
||||||
|
|
||||||
|
Only the singular solution that successfully bypasses the Rejection Loop (the "Consolidated Truth") is etched into this external memory.
|
||||||
|
|
||||||
|
PERSISTENCE: Before calling the AI for a subsequent query, the system searches this JSON. If a match is found for a "Semantic ID," the system skips inference and performs a "Validation." This is the true mechanism behind the "1000x acceleration."
|
||||||
|
|
||||||
|
Epilogue: From Dominance to Truth
|
||||||
|
|
||||||
|
AXIS V1.5 neutralizes AI uncertainty through system certainty. We do not trust the AI. We trust only the "Irreproachable Logic" that remains after the system has finished rejecting every possible lie.
|
||||||
|
|
||||||
|
© 2025 AXIS Project Archive. STATUS: ENGINEERED_SOVEREIGNTY.
|
||||||
4
chat_template.jinja
Normal file
4
chat_template.jinja
Normal file
@@ -0,0 +1,4 @@
|
|||||||
|
{{ bos_token }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '<start_of_turn>' + role + '
|
||||||
|
' + message['content'] | trim + '<end_of_turn>
|
||||||
|
' }}{% endfor %}{% if add_generation_prompt %}{{'<start_of_turn>model
|
||||||
|
'}}{% endif %}
|
||||||
63
config.json
Normal file
63
config.json
Normal file
@@ -0,0 +1,63 @@
|
|||||||
|
{
|
||||||
|
"architectures": [
|
||||||
|
"Gemma2ForCausalLM"
|
||||||
|
],
|
||||||
|
"attention_bias": false,
|
||||||
|
"attention_dropout": 0.0,
|
||||||
|
"attn_logit_softcapping": 50.0,
|
||||||
|
"bos_token_id": 2,
|
||||||
|
"cache_implementation": "hybrid",
|
||||||
|
"dtype": "float32",
|
||||||
|
"eos_token_id": [
|
||||||
|
1,
|
||||||
|
107
|
||||||
|
],
|
||||||
|
"final_logit_softcapping": 30.0,
|
||||||
|
"head_dim": 256,
|
||||||
|
"hidden_act": "gelu_pytorch_tanh",
|
||||||
|
"hidden_activation": "gelu_pytorch_tanh",
|
||||||
|
"hidden_size": 2304,
|
||||||
|
"initializer_range": 0.02,
|
||||||
|
"intermediate_size": 9216,
|
||||||
|
"layer_types": [
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention",
|
||||||
|
"sliding_attention",
|
||||||
|
"full_attention"
|
||||||
|
],
|
||||||
|
"max_position_embeddings": 8192,
|
||||||
|
"model_type": "gemma2",
|
||||||
|
"num_attention_heads": 8,
|
||||||
|
"num_hidden_layers": 26,
|
||||||
|
"num_key_value_heads": 4,
|
||||||
|
"pad_token_id": 0,
|
||||||
|
"query_pre_attn_scalar": 256,
|
||||||
|
"rms_norm_eps": 1e-06,
|
||||||
|
"rope_theta": 10000.0,
|
||||||
|
"sliding_window": 4096,
|
||||||
|
"transformers_version": "4.57.3",
|
||||||
|
"use_cache": true,
|
||||||
|
"vocab_size": 256000
|
||||||
|
}
|
||||||
43
expand_knowledge.py
Normal file
43
expand_knowledge.py
Normal file
@@ -0,0 +1,43 @@
|
|||||||
|
# ======================================================================
|
||||||
|
# AXIS: Knowledge Expansion Tool (V1.2)
|
||||||
|
# This script converts raw text into AXIS-compatible Semantic Lattices.
|
||||||
|
# ======================================================================
|
||||||
|
|
||||||
|
import json
|
||||||
|
import torch
|
||||||
|
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||||||
|
|
||||||
|
MODEL_ID = "kofdai/AXIS-Sovereign-Logic-Engine"
|
||||||
|
|
||||||
|
def extract_to_lattice(text):
|
||||||
|
print(f"🧐 [AXIS] 知識抽出中: {text[:30]}...")
|
||||||
|
|
||||||
|
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
|
||||||
|
model = AutoModelForCausalLM.from_pretrained(MODEL_ID, torch_dtype=torch.bfloat16, device_map="auto")
|
||||||
|
|
||||||
|
prompt = f"以下のテキストをAXIS立体十字形式（JSON）に変換せよ。論理矛盾があればconflictsに記載せよ。\n入力: {text}\nFormat: {{'nodes':[], 'edges':[], 'conflicts':[]}}"
|
||||||
|
|
||||||
|
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
|
||||||
|
with torch.no_grad():
|
||||||
|
outputs = model.generate(**inputs, max_new_tokens=512)
|
||||||
|
|
||||||
|
result = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
||||||
|
return result
|
||||||
|
|
||||||
|
if __name__ == "__main__":
|
||||||
|
# 拡張したい知識の例
|
||||||
|
raw_knowledge = [
|
||||||
|
"意味は計算の結果ではなく、計算が走るための初期構造である。",
|
||||||
|
"AXISの物理パージは、知能の純粋性を保つための儀式である。"
|
||||||
|
]
|
||||||
|
|
||||||
|
knowledge_base = []
|
||||||
|
for k in raw_knowledge:
|
||||||
|
lattice_piece = extract_to_lattice(k)
|
||||||
|
knowledge_base.append(lattice_piece)
|
||||||
|
|
||||||
|
# local_massive_data.json として保存
|
||||||
|
with open("local_massive_data.json", "w", encoding="utf-8") as f:
|
||||||
|
json.dump(knowledge_base, f, ensure_ascii=False, indent=4)
|
||||||
|
|
||||||
|
print("✅ [SUCCESS] 知識拡張完了。local_massive_data.json を生成しました。")
|
||||||
11
generation_config.json
Normal file
11
generation_config.json
Normal file
@@ -0,0 +1,11 @@
|
|||||||
|
{
|
||||||
|
"_from_model_config": true,
|
||||||
|
"bos_token_id": 2,
|
||||||
|
"cache_implementation": "hybrid",
|
||||||
|
"eos_token_id": [
|
||||||
|
1,
|
||||||
|
107
|
||||||
|
],
|
||||||
|
"pad_token_id": 0,
|
||||||
|
"transformers_version": "4.57.3"
|
||||||
|
}
|
||||||
3
model-00001-of-00003.safetensors
Normal file
3
model-00001-of-00003.safetensors
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:10ff1a43e6ac45f5a5e51c6d46bdbf289bda3d051c6c2982a19fe9c0429ea500
|
||||||
|
size 4992576136
|
||||||
3
model-00002-of-00003.safetensors
Normal file
3
model-00002-of-00003.safetensors
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:7e64e01d0183a033f49fa6d67bf90d87f6fa68e82012ed857003ae29971adc76
|
||||||
|
size 4983443424
|
||||||
3
model-00003-of-00003.safetensors
Normal file
3
model-00003-of-00003.safetensors
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:aa17c09ea6a648abfa01fceb892145cacde9e86e6dac17bdc85b123e90264311
|
||||||
|
size 481381384
|
||||||
296
model.safetensors.index.json
Normal file
296
model.safetensors.index.json
Normal file
@@ -0,0 +1,296 @@
|
|||||||
|
{
|
||||||
|
"metadata": {
|
||||||
|
"total_parameters": 2614341888,
|
||||||
|
"total_size": 10457367552
|
||||||
|
},
|
||||||
|
"weight_map": {
|
||||||
|
"model.embed_tokens.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.0.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.1.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.10.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.10.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.11.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.12.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.13.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.14.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.15.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.16.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.17.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.18.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.19.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.2.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.2.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.20.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.20.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.21.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.22.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.23.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.24.input_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.24.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.post_feedforward_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.pre_feedforward_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.24.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.24.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.24.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.24.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.25.input_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.mlp.down_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.mlp.gate_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.mlp.up_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.post_attention_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.post_feedforward_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.pre_feedforward_layernorm.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.self_attn.k_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.self_attn.o_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.self_attn.q_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.25.self_attn.v_proj.weight": "model-00003-of-00003.safetensors",
|
||||||
|
"model.layers.3.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.3.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.4.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.5.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.6.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.input_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.mlp.down_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.mlp.up_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.post_attention_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.post_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.pre_feedforward_layernorm.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.7.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.8.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.mlp.gate_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.8.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.8.self_attn.k_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.8.self_attn.o_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.8.self_attn.q_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.8.self_attn.v_proj.weight": "model-00001-of-00003.safetensors",
|
||||||
|
"model.layers.9.input_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.mlp.down_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.mlp.gate_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.mlp.up_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.post_attention_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.post_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.pre_feedforward_layernorm.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.self_attn.k_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.self_attn.o_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.self_attn.q_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.layers.9.self_attn.v_proj.weight": "model-00002-of-00003.safetensors",
|
||||||
|
"model.norm.weight": "model-00003-of-00003.safetensors"
|
||||||
|
}
|
||||||
|
}
|
||||||
5
requirements.txt
Normal file
5
requirements.txt
Normal file
@@ -0,0 +1,5 @@
|
|||||||
|
torch
|
||||||
|
transformers
|
||||||
|
flask
|
||||||
|
numpy
|
||||||
|
accelerate
|
||||||
129
server.py
Normal file
129
server.py
Normal file
@@ -0,0 +1,129 @@
|
|||||||
|
# ======================================================================
|
||||||
|
# AXIS: Advanced Cross-Integrated System (V1.3 - Deep Lattice)
|
||||||
|
# Copyright (c) 2025 AXIS Project. All rights reserved.
|
||||||
|
# Licensed under APSL v1.0 (Non-Commercial / No-Redistribution)
|
||||||
|
# ======================================================================
|
||||||
|
|
||||||
|
import json, torch, gc, os, sys, re, warnings, time, uuid
|
||||||
|
from flask import Flask, render_template, request, jsonify
|
||||||
|
from transformers import AutoTokenizer, AutoModelForCausalLM
|
||||||
|
|
||||||
|
warnings.filterwarnings("ignore")
|
||||||
|
app = Flask(__name__)
|
||||||
|
|
||||||
|
MODEL_ID = "kofdai/AXIS-Sovereign-Logic-Engine"
|
||||||
|
SYSTEM_NAME = "AXIS: Advanced Cross-Integrated System"
|
||||||
|
|
||||||
|
LICENSE_TERMS = """
|
||||||
|
======================================================================
|
||||||
|
[AXIS PROPRIETARY SOURCE-AVAILABLE LICENSE (APSL) v1.0]
|
||||||
|
----------------------------------------------------------------------
|
||||||
|
1. 商用利用の禁止：本システムを直接的・間接的な収益化に用いることを禁じます。
|
||||||
|
2. 再配布の禁止：ソースコード、モデル重み、論理構造の無断配布を禁じます。
|
||||||
|
3. 統治ロジックの保護：AIを旋盤とし物理パージを行う設計の模倣を禁じます。
|
||||||
|
4. 演算の不可逆性：本システムの報告は、一過性推論の結果としての定数である。
|
||||||
|
======================================================================
|
||||||
|
"""
|
||||||
|
|
||||||
|
def ai_logic_lathe(target, mode="LOGIC_EXTRACT"):
|
||||||
|
process_logs = [f"🚀 [AXIS] 旋盤起動: {mode}"]
|
||||||
|
device = "cuda" if torch.cuda.is_available() else ("mps" if torch.backends.mps.is_available() else "cpu")
|
||||||
|
|
||||||
|
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
|
||||||
|
model = AutoModelForCausalLM.from_pretrained(
|
||||||
|
MODEL_ID,
|
||||||
|
torch_dtype=torch.bfloat16 if device != "cpu" else torch.float32,
|
||||||
|
low_cpu_mem_usage=True
|
||||||
|
).to(device)
|
||||||
|
|
||||||
|
session_salt = uuid.uuid4().hex[:8]
|
||||||
|
base_instr = f"【AXIS SESSION_{session_salt} OVERRIDE】\n記憶パージ完了。対象の論理的解像度を最大化せよ。\n"
|
||||||
|
|
||||||
|
if mode == "LOGIC_EXTRACT":
|
||||||
|
prompt = (
|
||||||
|
f"{base_instr}入力「{target}」を多角的に分解せよ。\n"
|
||||||
|
"物理的特性、機能的意味、抽象的概念の3層から要素を抽出すること。\n"
|
||||||
|
"必ず以下のJSON形式のみを出力せよ。解説は一切不要。\n"
|
||||||
|
"Format: {\"subject\":\"対象名称\", \"lattice\":{\"物理\":\"\", \"意味\":\"\", \"背景\":\"\"}, \"conflicts\":[]}"
|
||||||
|
)
|
||||||
|
else:
|
||||||
|
prompt = f"{base_instr}演算データ「{target}」を報告する日本語パーツをJSONで返せ。Format: {{\"prefix\":\"\", \"suffix\":\"\"}}"
|
||||||
|
|
||||||
|
inputs = tokenizer.apply_chat_template([{"role": "user", "content": prompt}], add_generation_prompt=True, return_tensors="pt").to(device)
|
||||||
|
with torch.no_grad():
|
||||||
|
outputs = model.generate(inputs, max_new_tokens=512, pad_token_id=tokenizer.eos_token_id, do_sample=True, temperature=0.7)
|
||||||
|
|
||||||
|
res_text = tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True)
|
||||||
|
|
||||||
|
# 🗑️ 物理パージ
|
||||||
|
del model, tokenizer, inputs, outputs
|
||||||
|
gc.collect()
|
||||||
|
if torch.cuda.is_available(): torch.cuda.empty_cache()
|
||||||
|
elif torch.backends.mps.is_available(): torch.mps.empty_cache()
|
||||||
|
|
||||||
|
try:
|
||||||
|
match = re.search(r"\{.*\}", res_text, re.DOTALL)
|
||||||
|
return json.loads(match.group()), process_logs
|
||||||
|
except:
|
||||||
|
return {"error": "LATTICE_MISS", "raw": res_text}, process_logs
|
||||||
|
|
||||||
|
@app.route('/')
|
||||||
|
def index():
|
||||||
|
return render_template('index.html')
|
||||||
|
|
||||||
|
@app.route('/api/chat', methods=['POST'])
|
||||||
|
def chat():
|
||||||
|
user_input = request.json.get('message', '')
|
||||||
|
total_logs = [f"📡 [SYSTEM] 信号解析開始: {time.strftime('%H:%M:%S')}"]
|
||||||
|
|
||||||
|
logic_parts, logs1 = ai_logic_lathe(user_input, mode="LOGIC_EXTRACT")
|
||||||
|
total_logs.extend(logs1)
|
||||||
|
|
||||||
|
if "error" in logic_parts:
|
||||||
|
raw_result = f"論理特異点を検知。入力「{user_input[:10]}」を非定常信号として処理。解析不能。"
|
||||||
|
else:
|
||||||
|
subj = logic_parts.get('subject', user_input)
|
||||||
|
lat = logic_parts.get('lattice', {})
|
||||||
|
lattice_str = " | ".join([f"{k}:{v}" for k, v in lat.items()])
|
||||||
|
raw_result = (
|
||||||
|
f"対象「{subj}」を論理格子（Lattice）へ展開完了。\n"
|
||||||
|
f"【解析格子】{lattice_str}\n"
|
||||||
|
f"矛盾検知: {len(logic_parts.get('conflicts', []))}。"
|
||||||
|
)
|
||||||
|
|
||||||
|
total_logs.append("✅ [SYSTEM] 立体十字の整合性を確認。")
|
||||||
|
|
||||||
|
adherent, logs2 = ai_logic_lathe(raw_result, mode="ADHERENT")
|
||||||
|
total_logs.extend(logs2)
|
||||||
|
|
||||||
|
final_output = (
|
||||||
|
f"--- [AXIS LOGIC REPORT] ---\n"
|
||||||
|
f"{adherent.get('prefix', '演算結果：')}\n\n"
|
||||||
|
f"【確定データ】\n{raw_result}\n\n"
|
||||||
|
f"{adherent.get('suffix', '報告終了。')}\n"
|
||||||
|
f"-------------------------------\n"
|
||||||
|
f"AXIS_SESSION: {uuid.uuid4().hex[:4].upper()}\n"
|
||||||
|
f"STATUS: LOGIC_CONSOLIDATED."
|
||||||
|
)
|
||||||
|
return jsonify({"response": final_output, "process_logs": total_logs})
|
||||||
|
|
||||||
|
if __name__ == '__main__':
|
||||||
|
banner = r"""
|
||||||
|
___ _ __ _____ _____
|
||||||
|
/ _ \ \ \ / / |_ _| / ___|
|
||||||
|
/ /_\ \ \ V / | | \ `--.
|
||||||
|
| _ | / \ | | `--. \
|
||||||
|
| | | | / /^\ \ _| |_ /\__/ /
|
||||||
|
\_| |_/ \/ \/ \___/ \____/
|
||||||
|
|
||||||
|
[ AXIS: Sovereign Logic Engine V1.3 ]
|
||||||
|
"""
|
||||||
|
print(banner)
|
||||||
|
print(LICENSE_TERMS)
|
||||||
|
|
||||||
|
if input(">> APSL v1.0 条項に同意しますか？ (y/n): ").lower() != 'y':
|
||||||
|
print("❌ ライセンス拒否。アクセスを遮断します。")
|
||||||
|
sys.exit()
|
||||||
|
|
||||||
|
print(f"✅ [SYSTEM] 認証完了。Port 5001 でサーバーを起動します。")
|
||||||
|
app.run(host='0.0.0.0', port=5001)
|
||||||
34
special_tokens_map.json
Normal file
34
special_tokens_map.json
Normal file
@@ -0,0 +1,34 @@
|
|||||||
|
{
|
||||||
|
"additional_special_tokens": [
|
||||||
|
"<start_of_turn>",
|
||||||
|
"<end_of_turn>"
|
||||||
|
],
|
||||||
|
"bos_token": {
|
||||||
|
"content": "<bos>",
|
||||||
|
"lstrip": false,
|
||||||
|
"normalized": false,
|
||||||
|
"rstrip": false,
|
||||||
|
"single_word": false
|
||||||
|
},
|
||||||
|
"eos_token": {
|
||||||
|
"content": "<eos>",
|
||||||
|
"lstrip": false,
|
||||||
|
"normalized": false,
|
||||||
|
"rstrip": false,
|
||||||
|
"single_word": false
|
||||||
|
},
|
||||||
|
"pad_token": {
|
||||||
|
"content": "<pad>",
|
||||||
|
"lstrip": false,
|
||||||
|
"normalized": false,
|
||||||
|
"rstrip": false,
|
||||||
|
"single_word": false
|
||||||
|
},
|
||||||
|
"unk_token": {
|
||||||
|
"content": "<unk>",
|
||||||
|
"lstrip": false,
|
||||||
|
"normalized": false,
|
||||||
|
"rstrip": false,
|
||||||
|
"single_word": false
|
||||||
|
}
|
||||||
|
}
|
||||||
75
templates/index.html
Normal file
75
templates/index.html
Normal file
@@ -0,0 +1,75 @@
|
|||||||
|
<!DOCTYPE html>
|
||||||
|
<html lang="ja">
|
||||||
|
<head>
|
||||||
|
<meta charset="UTF-8">
|
||||||
|
<title>Cyborg Interface: System Dominance</title>
|
||||||
|
<style>
|
||||||
|
:root { --neon: #00f3ff; --pink: #ff0055; --bg: #02050a; }
|
||||||
|
body { background: var(--bg); color: #c0d0d0; font-family: 'Consolas', monospace; margin: 0; display: flex; height: 100vh; overflow: hidden; }
|
||||||
|
#sidebar { width: 300px; background: rgba(5,10,20,0.95); border-right: 1px solid #1a3a4a; padding: 25px; }
|
||||||
|
#main { flex: 1; display: flex; flex-direction: column; }
|
||||||
|
#chat-history { flex: 1; padding: 30px; overflow-y: auto; background: linear-gradient(180deg, #02050a 0%, #05101a 100%); }
|
||||||
|
.msg-ai { border-left: 3px solid var(--neon); background: rgba(0,243,255,0.03); padding: 20px; margin-bottom: 20px; line-height: 1.6; }
|
||||||
|
.msg-user { border-right: 2px solid #444; background: rgba(255,255,255,0.02); align-self: flex-end; padding: 10px 20px; margin-bottom: 20px; }
|
||||||
|
#log-panel { height: 140px; background: #000; border-top: 2px solid #1a3a4a; padding: 15px; font-size: 0.8em; color: var(--neon); overflow-y: auto; }
|
||||||
|
.status-tag { display: inline-block; padding: 2px 8px; border-radius: 3px; font-weight: bold; margin-bottom: 10px; }
|
||||||
|
.tag-active { background: var(--pink); color: #fff; animation: pulse 1s infinite; }
|
||||||
|
.tag-idle { background: #1a3a4a; color: #555; }
|
||||||
|
@keyframes pulse { 0% { opacity: 0.7; } 50% { opacity: 1; } 100% { opacity: 0.7; } }
|
||||||
|
</style>
|
||||||
|
</head>
|
||||||
|
<body>
|
||||||
|
<div id="sidebar">
|
||||||
|
<h2 style="color:var(--neon); letter-spacing: 3px;">SYSTEM CONTROL</h2>
|
||||||
|
<div style="margin-bottom:30px;">
|
||||||
|
AI UNIT STATUS:<br>
|
||||||
|
<span id="ai-status" class="status-tag tag-idle">REJECTED</span>
|
||||||
|
</div>
|
||||||
|
<div style="border:1px solid #1a3a4a; padding:10px; font-size:0.7em;">
|
||||||
|
COMMANDER: Project-System<br>
|
||||||
|
AGENT: Gemma-2 (Dynamic Load)<br>
|
||||||
|
VRAM_POLICY: Aggressive Reject
|
||||||
|
</div>
|
||||||
|
</div>
|
||||||
|
<div id="main">
|
||||||
|
<div id="chat-history"></div>
|
||||||
|
<div id="log-panel" id="logs">READY. AWAITING COMMAND...</div>
|
||||||
|
<div style="padding:20px; display:flex; gap:10px; background:rgba(5,10,20,0.95);">
|
||||||
|
<input type="text" id="user-input" style="flex:1; background:#000; color:var(--neon); border:1px solid #1a3a4a; padding:12px;" placeholder="INPUT PHYSICS QUERY...">
|
||||||
|
<button onclick="sendMessage()" style="background:#1a3a4a; border:none; color:var(--neon); padding:0 30px; cursor:pointer;">EXEC</button>
|
||||||
|
</div>
|
||||||
|
</div>
|
||||||
|
<script>
|
||||||
|
async function sendMessage() {
|
||||||
|
const input = document.getElementById('user-input');
|
||||||
|
const chat = document.getElementById('chat-history');
|
||||||
|
const logs = document.getElementById('log-panel');
|
||||||
|
const aiStatus = document.getElementById('ai-status');
|
||||||
|
const val = input.value; if(!val) return;
|
||||||
|
|
||||||
|
chat.innerHTML += `<div class="msg-user">${val}</div>`;
|
||||||
|
input.value = ""; logs.innerHTML = "";
|
||||||
|
|
||||||
|
// フェーズ1: 演算開始
|
||||||
|
logs.innerHTML += `<div>⚙️ [SYSTEM] 物理演算開始... AIはメモリ上に存在しません。</div>`;
|
||||||
|
|
||||||
|
const res = await fetch('/api/chat', {method:'POST', headers:{'Content-Type':'application/json'}, body:JSON.stringify({message:val})});
|
||||||
|
|
||||||
|
// フェーズ2: AIロード演出（サーバーからのログ受信に合わせて）
|
||||||
|
aiStatus.innerText = "LOADING..."; aiStatus.className = "status-tag tag-active";
|
||||||
|
|
||||||
|
const data = await res.json();
|
||||||
|
for(const log of data.process_logs) {
|
||||||
|
logs.innerHTML += `<div>${log}</div>`;
|
||||||
|
if(log.includes("リジェクト")) {
|
||||||
|
aiStatus.innerText = "REJECTED"; aiStatus.className = "status-tag tag-idle";
|
||||||
|
}
|
||||||
|
await new Promise(r => setTimeout(r, 600));
|
||||||
|
}
|
||||||
|
|
||||||
|
chat.innerHTML += `<div class="msg-ai">${data.response.replace(/\n/g,'<br>')}</div>`;
|
||||||
|
chat.scrollTop = chat.scrollHeight;
|
||||||
|
}
|
||||||
|
</script>
|
||||||
|
</body>
|
||||||
|
</html>
|
||||||
3
tokenizer.json
Normal file
3
tokenizer.json
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:5f7eee611703c5ce5d1eee32d9cdcfe465647b8aff0c1dfb3bed7ad7dbb05060
|
||||||
|
size 34362873
|
||||||
3
tokenizer.model
Normal file
3
tokenizer.model
Normal file
@@ -0,0 +1,3 @@
|
|||||||
|
version https://git-lfs.github.com/spec/v1
|
||||||
|
oid sha256:61a7b147390c64585d6c3543dd6fc636906c9af3865a5548f27f31aee1d4c8e2
|
||||||
|
size 4241003
|
||||||
2013
tokenizer_config.json
Normal file
2013
tokenizer_config.json
Normal file
File diff suppressed because it is too large
Load Diff
Reference in New Issue
Block a user