telos-ml 1.0.0__tar.gz

telos_ml-1.0.0/.gitignore

@@ -0,0 +1,50 @@
+__pycache__/
+*.py[cod]
+*$py.class
+
+.env
+.venv
+.hf_venv
+env/
+venv/
+
+.vscode/
+.idea/
+.DS_Store
+
+/data/
+/checkpoints/
+/checkpoints_*/
+.gradio/
+*.pt
+*.safetensors
+
+build/
+develop-eggs/
+dist/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+htmlcov/
+.tox/
+.coverage
+.cache
+.pytest_cache/
+
+# Compiled Swift AFM bridge. Built on demand into a user cache dir
+# (~/.cache/telos) rather than into the package; this pattern only catches
+# stray local builds that land next to the source.
+telos/**/afm_bridge
+/afm_bridge
+
+# Stray compiled binary at repo root (not a source file)
+/main

telos_ml-1.0.0/CHANGELOG.md

@@ -0,0 +1,47 @@
+# Changelog
+
+All notable changes to this project are documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+## [1.0.0] - 2026-09-23
+
+First public release on PyPI as `telos-ml` (`pip install telos-ml`).
+
+### Added
+- **`telos-ml` PyPI Distribution** — Published package to PyPI under the name
+  `telos-ml`, preserving the `telos` Python module import and the `telos` CLI
+  command.
+- **`telos dataprep`** — prepares raw text, code directories, JSONL, or Hugging
+  Face datasets into memory-mapped binary token streams (`.bin`).
+- **`telos train`** — zero-config dimensional trainer. Specify parameters and
+  tokens; the geometry solver derives `d_model`, layers, and heads. Supports the
+  AR, MDLM, UNDLM, and COROSred paradigms on MLX (Apple Silicon), CUDA, TPU via
+  PyTorch-XLA, and CPU.
+- **COROSred** — the unified dual-paradigm trainer, combining causal AR drafting,
+  uniform masked-diffusion infilling, and confidence-guided selective
+  re-diffusion, with a Learned Reliability Head driving routing.
+- **`telos eval`** — multi-domain evaluation: contextual probes with bootstrap
+  confidence intervals, sandboxed functional execution (Pass@1, AST validity),
+  anti-cheat span masking, linguistic probes, and tool-use / function-calling
+  benchmarks.
+- **`telos bench`** — hardware throughput and step-latency benchmark, capped at
+  five minutes.
+- **`telos test`** — unified verification suite for model contracts, causality,
+  losses, and samplers.
+- **`telos.afm`** — Python access to Apple's on-device Foundation Models
+  (AFM 3 Core / Core Advanced) on Apple Silicon. Exposes `probe()`, `require()`,
+  `generate()`, and a persistent `AFMBridge`, plus a `telos afm status` /
+  `telos afm generate` CLI. The Swift bridge is compiled on first use into a user
+  cache directory.
+- Bundled default ByteLevel BPE tokenizer, so an installed package works with no
+  manual downloads.
+- Graceful degradation when MLX is installed but Metal is unavailable (headless,
+  virtualized, or sandboxed macOS sessions): `telos train`, `telos bench` and
+  `telos test` fall back to the PyTorch CPU path instead of aborting the process.
+
+[Unreleased]: https://github.com/kazenoko-git/telos/compare/v1.0.0...HEAD
+[1.0.0]: https://github.com/kazenoko-git/telos/releases/tag/v1.0.0

telos_ml-1.0.0/LICENSE

@@ -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 exemplary 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 Ivan Samuel / télos team
+
+   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.

telos_ml-1.0.0/PKG-INFO

@@ -0,0 +1,234 @@
+Metadata-Version: 2.5
+Name: telos-ml
+Version: 1.0.0
+Summary: Discrete diffusion and autoregressive language modeling: data preparation, training on MLX/CUDA/TPU, evaluation, and Python access to Apple's on-device Foundation Models.
+Project-URL: Homepage, https://telos.research.wingit.tech
+Project-URL: Repository, https://github.com/kazenoko-git/telos
+Project-URL: Issues, https://github.com/kazenoko-git/telos/issues
+Project-URL: Documentation, https://github.com/kazenoko-git/telos/blob/main/DEPLOYMENT.md
+Project-URL: Changelog, https://github.com/kazenoko-git/telos/blob/main/CHANGELOG.md
+Author: Ivan Samuel
+License-Expression: Apache-2.0
+License-File: LICENSE
+Keywords: apple-foundation-models,autoregressive,code-completion,cuda,discrete-diffusion,language-model,masked-diffusion,mlx,tpu
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.10
+Requires-Dist: datasets>=3.0.0
+Requires-Dist: huggingface-hub>=0.25.0
+Requires-Dist: numpy>=1.26.0
+Requires-Dist: pyyaml>=6.0.0
+Requires-Dist: safetensors>=0.4.0
+Requires-Dist: tokenizers>=0.21.0
+Requires-Dist: torch>=2.5.0
+Requires-Dist: tqdm>=4.66.0
+Provides-Extra: all
+Requires-Dist: gradio>=5.0.0; extra == 'all'
+Requires-Dist: matplotlib>=3.8.0; extra == 'all'
+Requires-Dist: mlx>=0.22.0; (sys_platform == 'darwin' and platform_machine == 'arm64') and extra == 'all'
+Requires-Dist: pytest>=8.0.0; extra == 'all'
+Requires-Dist: ruff>=0.8.0; extra == 'all'
+Provides-Extra: dev
+Requires-Dist: gradio>=5.0.0; extra == 'dev'
+Requires-Dist: matplotlib>=3.8.0; extra == 'dev'
+Requires-Dist: pytest>=8.0.0; extra == 'dev'
+Requires-Dist: ruff>=0.8.0; extra == 'dev'
+Provides-Extra: mlx
+Requires-Dist: mlx>=0.22.0; (sys_platform == 'darwin' and platform_machine == 'arm64') and extra == 'mlx'
+Description-Content-Type: text/markdown
+
+# télos (τέλος)
+
+<p align="center">
+  <img src="logos/telos_logo.png" alt="telos logo" width="340">
+</p>
+
+<p align="center">
+  <a href="https://telos.research.wingit.tech"><strong>Research Portal: telos.research.wingit.tech</strong></a>
+  <br>
+  <em>Note: The research portal is currently under active construction and is not finished yet.</em>
+</p>
+
+<p align="center">
+  <img src="https://img.shields.io/badge/organization-Wing%20It%20Research-blueviolet" alt="Wing It Research">
+  <img src="https://img.shields.io/badge/license-Apache--2.0-blue.svg" alt="License">
+  <img src="https://img.shields.io/badge/python-3.10%20%7C%203.11%20%7C%203.12%20%7C%203.13%20%7C%203.14-blue" alt="Python">
+  <img src="https://img.shields.io/badge/hardware-Apple%20Silicon%20(MLX)%20%7C%20CUDA%20%7C%20TPU%20(XLA)-orange" alt="Hardware Support">
+</p>
+
+## About Télos
+
+"Whatever is new — we wing it."
+
+Télos is an open research project by Wing It Research focused on AI research, benchmarks, and new architectures.
+
+We are working on:
+1. **COROSred** (COnfidence ROuted Selective ReDiffusion)
+2. **Inference Optimal Scaling Laws** for Masked Diffusion and Uniform Noise Diffusion models
+3. **Benchmarks for new models**
+
+## Latest Findings
+
+### Benchmarks (AFM-3 vs Granite vs LFM 8B)
+
+We tested three edge models: Apple AFM-3 Core Advanced, IBM Granite 4.2 3B, and LiquidAI LFM 8B across 7 evaluation suites.
+
+<p align="center">
+  <img src="figures/benchmark_radar_graph.png" alt="Benchmark Radar Graph Across Models" width="600">
+</p>
+
+| Suite | Tasks | Apple AFM-3 Core Adv. | IBM Granite 4.2 3B | LiquidAI LFM 8B A1B | Metric |
+| :--- | :---: | :---: | :---: | :---: | :--- |
+| **OpenAI HumanEval** | 164 | **62.80%** | 60.37% | 45.73% | Pass@1 (Code Execution) |
+| **BFCL Tool-Use** | 30 | **73.33%** | 70.00% | 63.33% | Pass@1 (JSON Match) |
+| **Cybersecurity Auditing** | 50 | 12.00% | 22.00% | **24.00%** | Pass@1 (Vulnerability Fix) |
+| **GPQA Diamond** | 198 | **39.39%** | 28.79% | 20.71% | Pass@1 (Multiple Choice) |
+| **MMLU Science** | 119 | 69.75% | 61.82% | **73.11%** | Pass@1 (Multiple Choice) |
+| **Competition MATH** | 140 | 52.00% | **64.71%** | 43.57% | Pass@1 (Math Solutions) |
+| **ARC-Challenge** | 150 | 85.32% | **86.60%** | 59.56% | Pass@1 (Reasoning Choice) |
+| **Macro Average** | **841** | **56.37%** | **56.33%** | **47.14%** | 7-Suite Average |
+
+#### Runaway Loops and Hardware Efficiency
+
+Edge models sometimes get caught repeating words over and over, or they use too much memory.
+
+<p align="center">
+  <img src="figures/benchmark_repetition_rate.png" alt="Degenerate Loop Rate Across Benchmarks" width="800">
+</p>
+
+<p align="center">
+  <img src="figures/benchmark_throughput_memory.png" alt="Throughput and Memory Comparison" width="800">
+</p>
+
+- **Runaway Loops**: IBM Granite 4.2 3B had the highest runaway loop rate at **36.8%** (1,129 tasks caught in repeat loops), followed by LiquidAI LFM 8B at **28.5%**, and Apple AFM-3 at **23.4%**.
+- **Speed and Memory**: Apple AFM-3 runs natively via Swift IPC at **58.7 tok/s** using **2.4 GB** RAM. IBM Granite 4.2 3B runs at **41.2 tok/s** using **3.2 GB** RAM. LiquidAI LFM 8B runs at **22.4 tok/s** using **7.8 GB** RAM.
+
+See [`benchmarks/README.md`](benchmarks/README.md) for full benchmark reports and details.
+
+### Research (COROSred)
+
+COROSred combines fast autoregressive token drafting with bidirectional diffusion fixing. High-confidence tokens are accepted in one step. Low-confidence tokens are fixed with bidirectional re-diffusion.
+
+```
++-------------------------------------------------------+
+|                 TÉLOS MODEL BACKBONE                  |
+|       (RoPE, SwiGLU, RMSNorm, Weight Tying, GQA)      |
++---------------------------+---------------------------+
+                            |
+             +--------------+--------------+
+             |                             |
+             v                             v
++-------------------------+   +-------------------------+
+|  Causal AR Next-Token   |   |   Learned Reliability   |
+|  Loss: L_causal (alpha) |   |   Head Gate             |
++------------+------------+   +------------+------------+
+             |                             |
+             +--------------+--------------+
+                            |
+                     Routing Decision
+                            |
+             +--------------+--------------+
+             |                             |
+             v                             v
++---------------------------+ +---------------------------+
+| High Confidence (>= 0.65) | | Low Confidence (< 0.65)   |
+| Accept Draft Token        | | Route to Bidirectional    |
+| (Fast 1-Step AR)          | | Masked Re-Diffusion       |
++---------------------------+ +---------------------------+
+```
+
+- **Perplexity**: Achieves **5.08 validation PPL** (1.29x lower than pure AR).
+- **Bidirectional Infilling**: Achieves **63.0% Top-1 accuracy** on fill-in-the-blank code tasks (compared to 7.6% for pure AR).
+- **Anti-Cheat Span Masking**: Achieves **52.0% exact match** with a low 2.0% copy rate when multiple tokens are masked.
+
+See [`research/README.md`](research/README.md) for research papers and scaling details.
+
+## Documentation Directory Routing
+
+| Directory | Topic & Contents |
+| :--- | :--- |
+| [Research (`research/`)](research/README.md) | Scaling laws, architecture notes, and papers |
+| [Benchmarks (`benchmarks/`)](benchmarks/README.md) | Model comparisons, benchmark charts, and hardware metrics |
+| [CLI (`cli/`)](cli/README.md) | Command line tools, training guide, and setup |
+
+## Basic CLI Usage
+
+### Installation
+
+```bash
+# Standard install
+pip install telos-ml
+
+# With Apple Silicon Metal support (MLX)
+pip install "telos-ml[mlx]"
+```
+
+### Core CLI Commands
+
+```bash
+# 1. Prepare token data
+telos dataprep --input raw_data/ --output data/python.bin --vocab-size 8192
+
+# 2. Train a 50M COROSred model on Apple Silicon
+telos train --paradigm corosred --params 50M --tokens 2.5B --hardware mlx
+
+# 3. Run evaluation
+telos eval --checkpoint checkpoints/corosred/model.safetensors --type all
+
+# 4. Hardware benchmark (capped at 5 minutes)
+telos bench --paradigm corosred --params 50M --hardware mlx
+
+# 5. Check Apple Foundation Models on Apple Silicon
+telos afm status
+telos afm generate "Explain binary search in two sentences."
+
+# 6. Verification suite
+telos test
+```
+
+### Python API
+
+```python
+import telos
+
+# Use on-device Apple Foundation Models
+if telos.afm.probe().available:
+    response = telos.afm.generate("Write a python function for quicksort.")
+    print(response)
+```
+
+See [`cli/README.md`](cli/README.md) for all CLI flags and options.
+
+## Citations & References
+
+### Primary Citation
+
+```bibtex
+@article{samuel2026telos,
+  title   = {télos: Exploring Scaling Laws, Hardware Optimizations, and Paradigm Trade-offs in Discrete Diffusion and Autoregressive Language Models},
+  author  = {Ivan Samuel},
+  journal = {Wing It Research},
+  year    = {2026},
+  url     = {https://telos.research.wingit.tech}
+}
+```
+
+### Framework & Related Research
+
+- **Apple MLX**: Hannun et al., *MLX: Efficient machine learning on Apple silicon*, 2023.
+- **Masked Diffusion Language Models (MDLM)**: Sahoo et al., *Masked Diffusion Language Models*, 2024.
+- **Compute-Optimal Scaling (Chinchilla)**: Hoffmann et al., *Training Compute-Optimal Large Language Models*, 2022.
+- **Large Language Diffusion Models (LLaDA)**: Nie et al., *LLaDA: Large Language Diffusion Models*, 2025.
+- **Discrete Diffusion Language Modeling (DiffusionGemma)**: Google DeepMind, *DiffusionGemma: An experimental discrete diffusion model based on Gemma*, 2026.
+
+## License
+
+Apache-2.0 License. See [LICENSE](LICENSE) for details.