model-unfolder 0.2.0__tar.gz

model_unfolder-0.2.0/LICENSE

@@ -0,0 +1,201 @@
+                                 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 describing the origin of the Work and
+      reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 unfold contributors
+
+   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.

model_unfolder-0.2.0/PKG-INFO

@@ -0,0 +1,122 @@
+Metadata-Version: 2.4
+Name: model-unfolder
+Version: 0.2.0
+Summary: Unfold any HuggingFace transformer into an interactive architecture diagram, inline in Jupyter.
+Author: model-unfolder contributors
+License: Apache-2.0
+Keywords: transformers,visualization,llm,architecture,jupyter
+Classifier: Programming Language :: Python :: 3
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Framework :: Jupyter
+Classifier: Topic :: Scientific/Engineering :: Visualization
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: hf
+Requires-Dist: transformers>=4.40; extra == "hf"
+Provides-Extra: dev
+Requires-Dist: pytest>=7; extra == "dev"
+Requires-Dist: transformers>=4.40; extra == "dev"
+Dynamic: license-file
+
+# MODEL UNFOLDER
+
+> your one click model unfolder
+
+```python
+from model_unfolder import unfold
+unfold("meta-llama/Meta-Llama-3-8B")
+```
+
+<p align="center">
+  <a href="examples/llama-3-8b.html">
+    <img src="examples/images/llama-3-8b.png" width="540" alt="Meta-Llama-3-8B architecture diagram">
+  </a>
+</p>
+
+---
+
+## Install
+
+```bash
+pip install model-unfolder
+
+# for local development
+pip install -e .
+pip install transformers   # only required to load by model ID
+```
+
+## Three ways to call it
+
+```python
+from model_unfolder import unfold
+
+# 1) by HuggingFace model ID — only config.json is downloaded, never weights
+unfold("meta-llama/Meta-Llama-3-8B")
+unfold("deepseek-ai/DeepSeek-V3")
+
+# 2) from a transformers AutoConfig
+from transformers import AutoConfig
+unfold(AutoConfig.from_pretrained("Qwen/Qwen2.5-7B", trust_remote_code=True))
+
+# 3) from a raw config.json dict — no transformers install needed
+import json
+unfold(json.load(open("config.json")))
+```
+
+## Built on `transformers`
+
+Pass a model ID and `unfold` calls `transformers.AutoConfig.from_pretrained(model_id)` under the hood ([parser.py](model_unfolder/parser.py)). It only retries with `trust_remote_code=True` when Transformers says the config requires remote code.
+
+## Auth-token from your environment
+
+Gated models (Llama-3, Mistral, Gemma, …) need a HuggingFace token. `unfold` reuses whatever `transformers` / `huggingface_hub` already see:
+
+```bash
+# Either set an env var
+export HF_TOKEN="hf_xxxxxxxx"            # also accepted: HUGGING_FACE_HUB_TOKEN
+
+# or use the CLI cache (persists across sessions)
+huggingface-cli login
+
+# or load a .env in your notebook
+# >>> from dotenv import load_dotenv; load_dotenv()
+```
+
+No extra config in `model_unfolder` itself.
+
+## Save / export
+
+```python
+diagram = unfold(cfg)
+diagram.save("model.html")   # standalone interactive HTML
+diagram.save("model.json")   # IR (no rendering)
+diagram.param_count()        # {"total": ..., "active": ..., "per_layer": [...]}
+diagram.to_ir()              # full IR dict
+```
+
+Param estimates are close to published numbers — DeepSeek-V3 reports `~675B (~41B active)`, Llama-3-8B reports `8.03B`.
+
+## Live demos
+
+Open in any browser to interact (click blocks, expand sub-blocks, toggle layer types):
+
+| Model | Highlights | Demo |
+|---|---|---|
+| Llama-3-8B | GQA + dense baseline | [examples/llama-3-8b.html](examples/llama-3-8b.html) |
+| Mistral-7B-v0.3 | GQA + dense, 32k context | [examples/mistral-7b-v0.3.html](examples/mistral-7b-v0.3.html) |
+| DeepSeek-V3 | MLA + dense → MoE phase change | [examples/deepseek-v3.html](examples/deepseek-v3.html) |
+| Kimi K2 | MLA + 384-expert MoE, ~1T params | [examples/kimi-k2.html](examples/kimi-k2.html) |
+
+## Supported architectures
+
+| Family | Adapter | Notes |
+|---|---|---|
+| DeepSeek-V2 / V3 / Kimi K2 | [families/deepseek.py](model_unfolder/adapters/transformer/families/deepseek.py) | MLA + dense → MoE phase change |
+| Llama / Mistral / Qwen2 / Qwen3 / Phi-3 | [families/llama.py](model_unfolder/adapters/transformer/families/llama.py) | GQA / MQA / MHA + dense FFN |
+| Gemma 4 | [families/gemma4.py](model_unfolder/adapters/transformer/families/gemma4.py) | sliding/global layers, KV sharing, PLE |
+
+
+## License
+
+[Apache 2.0](LICENSE).

model_unfolder-0.2.0/README.md

@@ -0,0 +1,101 @@
+# MODEL UNFOLDER
+
+> your one click model unfolder
+
+```python
+from model_unfolder import unfold
+unfold("meta-llama/Meta-Llama-3-8B")
+```
+
+<p align="center">
+  <a href="examples/llama-3-8b.html">
+    <img src="examples/images/llama-3-8b.png" width="540" alt="Meta-Llama-3-8B architecture diagram">
+  </a>
+</p>
+
+---
+
+## Install
+
+```bash
+pip install model-unfolder
+
+# for local development
+pip install -e .
+pip install transformers   # only required to load by model ID
+```
+
+## Three ways to call it
+
+```python
+from model_unfolder import unfold
+
+# 1) by HuggingFace model ID — only config.json is downloaded, never weights
+unfold("meta-llama/Meta-Llama-3-8B")
+unfold("deepseek-ai/DeepSeek-V3")
+
+# 2) from a transformers AutoConfig
+from transformers import AutoConfig
+unfold(AutoConfig.from_pretrained("Qwen/Qwen2.5-7B", trust_remote_code=True))
+
+# 3) from a raw config.json dict — no transformers install needed
+import json
+unfold(json.load(open("config.json")))
+```
+
+## Built on `transformers`
+
+Pass a model ID and `unfold` calls `transformers.AutoConfig.from_pretrained(model_id)` under the hood ([parser.py](model_unfolder/parser.py)). It only retries with `trust_remote_code=True` when Transformers says the config requires remote code.
+
+## Auth-token from your environment
+
+Gated models (Llama-3, Mistral, Gemma, …) need a HuggingFace token. `unfold` reuses whatever `transformers` / `huggingface_hub` already see:
+
+```bash
+# Either set an env var
+export HF_TOKEN="hf_xxxxxxxx"            # also accepted: HUGGING_FACE_HUB_TOKEN
+
+# or use the CLI cache (persists across sessions)
+huggingface-cli login
+
+# or load a .env in your notebook
+# >>> from dotenv import load_dotenv; load_dotenv()
+```
+
+No extra config in `model_unfolder` itself.
+
+## Save / export
+
+```python
+diagram = unfold(cfg)
+diagram.save("model.html")   # standalone interactive HTML
+diagram.save("model.json")   # IR (no rendering)
+diagram.param_count()        # {"total": ..., "active": ..., "per_layer": [...]}
+diagram.to_ir()              # full IR dict
+```
+
+Param estimates are close to published numbers — DeepSeek-V3 reports `~675B (~41B active)`, Llama-3-8B reports `8.03B`.
+
+## Live demos
+
+Open in any browser to interact (click blocks, expand sub-blocks, toggle layer types):
+
+| Model | Highlights | Demo |
+|---|---|---|
+| Llama-3-8B | GQA + dense baseline | [examples/llama-3-8b.html](examples/llama-3-8b.html) |
+| Mistral-7B-v0.3 | GQA + dense, 32k context | [examples/mistral-7b-v0.3.html](examples/mistral-7b-v0.3.html) |
+| DeepSeek-V3 | MLA + dense → MoE phase change | [examples/deepseek-v3.html](examples/deepseek-v3.html) |
+| Kimi K2 | MLA + 384-expert MoE, ~1T params | [examples/kimi-k2.html](examples/kimi-k2.html) |
+
+## Supported architectures
+
+| Family | Adapter | Notes |
+|---|---|---|
+| DeepSeek-V2 / V3 / Kimi K2 | [families/deepseek.py](model_unfolder/adapters/transformer/families/deepseek.py) | MLA + dense → MoE phase change |
+| Llama / Mistral / Qwen2 / Qwen3 / Phi-3 | [families/llama.py](model_unfolder/adapters/transformer/families/llama.py) | GQA / MQA / MHA + dense FFN |
+| Gemma 4 | [families/gemma4.py](model_unfolder/adapters/transformer/families/gemma4.py) | sliding/global layers, KV sharing, PLE |
+
+
+## License
+
+[Apache 2.0](LICENSE).

model_unfolder-0.2.0/model_unfolder/__init__.py

@@ -0,0 +1,58 @@
+"""model_unfolder — turn any HuggingFace transformer into a clear architecture diagram.
+
+Quick start in a Jupyter notebook::
+
+    from model_unfolder import unfold
+    unfold("moonshotai/Kimi-K2-Instruct")
+
+Outside Jupyter::
+
+    diagram = unfold(cfg)
+    diagram.save("kimi_k2.html")
+"""
+from .diagram import Diagram
+from .parser import config_to_ir
+from .ir import ModelIR, LayerSpec, AttentionSpec, FFNSpec, CrossLayerEdge
+from .params import estimate_params
+
+__version__ = "0.2.0"
+
+__all__ = [
+    "unfold",
+    "show",
+    "Diagram",
+    "ModelIR",
+    "LayerSpec",
+    "AttentionSpec",
+    "FFNSpec",
+    "CrossLayerEdge",
+    "config_to_ir",
+    "estimate_params",
+]
+
+
+def unfold(cfg_or_id, token=None) -> Diagram:
+    """Unfold a transformer into a renderable architecture diagram.
+
+    Parameters
+    ----------
+    cfg_or_id
+        A HuggingFace ``PretrainedConfig`` instance, a model ID string
+        (e.g. ``"moonshotai/Kimi-K2-Instruct"``), or a plain ``dict`` of
+        ``config.json`` contents.
+    token
+        Optional Hugging Face token used only when ``cfg_or_id`` is a model ID.
+        If omitted, ``HF_TOKEN`` and legacy Hugging Face token env vars are used
+        when present.
+
+    Returns
+    -------
+    Diagram
+        Renders inline in Jupyter; otherwise call ``.save()`` or ``.to_html()``.
+    """
+    ir = config_to_ir(cfg_or_id, token=token)
+    return Diagram(ir)
+
+
+# friendly alias
+show = unfold

model_unfolder-0.2.0/model_unfolder/adapters/__init__.py

@@ -0,0 +1,15 @@
+"""Adapter registry. Order matters: more specific adapters come first."""
+from . import custom, diffusor, transformer
+
+ADAPTERS = [
+    *custom.ADAPTERS,
+    *diffusor.ADAPTERS,
+    *transformer.ADAPTERS,
+]
+
+
+def find_adapter(cfg):
+    for a in ADAPTERS:
+        if a.matches(cfg):
+            return a
+    return None

model_unfolder-0.2.0/model_unfolder/adapters/custom/__init__.py

@@ -0,0 +1,8 @@
+"""Project-local and experimental adapters.
+
+Drop custom adapters here when a model needs bespoke parsing before it is
+general enough to move into a domain package such as ``transformer``.
+"""
+
+ADAPTERS = []
+

model_unfolder-0.2.0/model_unfolder/adapters/diffusor/__init__.py

@@ -0,0 +1,8 @@
+"""Diffusor-family adapters.
+
+This package is intentionally empty for now.  It gives diffusion-style models
+their own adapter namespace instead of stretching transformer assumptions.
+"""
+
+ADAPTERS = []
+

model_unfolder-0.2.0/model_unfolder/adapters/transformer/__init__.py

@@ -0,0 +1,5 @@
+"""Transformer-family adapters and reusable block vocabularies."""
+
+from . import families
+
+ADAPTERS = families.ADAPTERS

model_unfolder-0.2.0/model_unfolder/adapters/transformer/assembly.py

@@ -0,0 +1,57 @@
+"""Assembly helpers for transformer-family adapters."""
+from __future__ import annotations
+
+from collections.abc import Iterable, Mapping
+from typing import Any
+
+from ...ir import AttentionSpec, FFNSpec, LayerSpec
+from .blocks import decoder_layer_blocks, decoder_only_render_spec
+
+
+def decoder_layer(
+    index: int,
+    attention: AttentionSpec,
+    ffn: FFNSpec,
+    hidden_size: int,
+    *,
+    extra_blocks: Iterable[dict] | None = None,
+) -> LayerSpec:
+    """Build a decoder layer from parsed specs plus optional reusable parts."""
+    blocks = decoder_layer_blocks(attention, ffn, hidden_size)
+    if extra_blocks:
+        blocks.extend(extra_blocks)
+    return LayerSpec(
+        index=index,
+        attention=attention,
+        ffn=ffn,
+        blocks=blocks,
+    )
+
+
+def decoder_extras(
+    vocab_size: int,
+    hidden_size: int,
+    tie_word_embeddings: bool,
+    *extra_maps: Mapping[str, Any] | None,
+) -> dict:
+    """Build top-level extras shared by decoder-only transformer models."""
+    extras = {
+        "render": decoder_only_render_spec(
+            vocab_size,
+            hidden_size,
+            tie_word_embeddings,
+        )
+    }
+    for extra in extra_maps:
+        if not extra:
+            continue
+        _merge_extras(extras, extra)
+    return extras
+
+
+def _merge_extras(target: dict, extra: Mapping[str, Any]) -> None:
+    for key, value in extra.items():
+        if key == "external_pathways" and key in target:
+            target[key].extend(value)
+        else:
+            target[key] = value