amass-kv 0.1.0__tar.gz

amass_kv-0.1.0/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 reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   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 [yyyy] [name of copyright owner]
+
+   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.

amass_kv-0.1.0/MANIFEST.in

@@ -0,0 +1,23 @@
+# MANIFEST.in -- controls what goes into the sdist (.tar.gz). The wheel's
+# contents are governed by pyproject's package discovery (pure `.py` modules).
+
+# Project metadata.
+include README.md
+include LICENSE
+include MANIFEST.in
+include pyproject.toml
+
+# Ship every Python source under the package (the runtime CUDA kernels are
+# `.py` files that `load_inline` at import time, so they MUST be included).
+recursive-include src/amass *.py
+
+# Keep dev-only notes and byte-compiled / scratch junk out of the release.
+recursive-exclude src *.md
+global-exclude *.pyc
+global-exclude *.pyo
+global-exclude *.so
+global-exclude *~
+global-exclude __pycache__/*
+prune **/__pycache__
+prune **/scratch_*
+prune **/tmp_*

amass_kv-0.1.0/PKG-INFO

@@ -0,0 +1,179 @@
+Metadata-Version: 2.4
+Name: amass-kv
+Version: 0.1.0
+Summary: AMASS: decode-time sparse KV-cache attention for vLLM (quad page-score selector + DRAM value tier)
+Author-email: Junsung Hwang <junsung.k.hwang@gmail.com>
+License-Expression: Apache-2.0
+Project-URL: Homepage, https://github.com/Js-Hwang1/amass-kv
+Project-URL: Repository, https://github.com/Js-Hwang1/amass-kv
+Project-URL: Issues, https://github.com/Js-Hwang1/amass-kv/issues
+Keywords: vllm,kv-cache,sparse-attention,llm-inference,decode,long-context,flash-attention,cuda
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Environment :: GPU :: NVIDIA CUDA
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: torch>=2.4
+Requires-Dist: triton>=3.0
+Provides-Extra: vllm
+Requires-Dist: vllm>=0.8; extra == "vllm"
+Provides-Extra: yaml
+Requires-Dist: pyyaml>=6.0; extra == "yaml"
+Provides-Extra: bench
+Requires-Dist: transformers>=4.40; extra == "bench"
+Provides-Extra: all
+Requires-Dist: vllm>=0.8; extra == "all"
+Requires-Dist: transformers>=4.40; extra == "all"
+Requires-Dist: pyyaml>=6.0; extra == "all"
+Dynamic: license-file
+
+# AMASS
+
+**Decode-time sparse KV-cache attention for vLLM.** AMASS keeps LLM decoding
+fast and memory-light at long context by attending, at every decode step, to
+only the pages that actually carry attention mass: a compact per-page score
+selects a small working set (typically 5 to 10 percent of the KV cache), and a
+sparse paged-attention kernel reads just those pages. Prefill is untouched, so
+time-to-first-token is unchanged.
+
+> **License note (provisional).** This release is distributed under Apache-2.0
+> as a placeholder pending final confirmation by the authors. See `LICENSE`.
+
+## What is in the box
+
+AMASS has two parts that share one selection core:
+
+1. **The quad selector (Stage A).** A quadratic, Gaussian-MGF page score built
+   from a tiny per-page summary. "quad" drops the per-key coordinates and stores
+   only `r'` eigenvalues per page, which shrinks the resident selector state by
+   **3.28x** (from 15.6 percent down to 4.7 percent of the KV cache; measured
+   1201 MiB vs 3943 MiB) while remaining accurate enough to pick the right pages.
+2. **The DRAM value tier (Stage B).** A pinned, device-mapped host-DRAM pool for
+   the value cache, with a bounded resident hot buffer plus staging pool. Only
+   the ~5 percent quad summary and the pages actually called each step stay in
+   VRAM; the rest of V (or K+V) lives in host memory.
+
+### Two variants, one algorithm
+
+| Variant       | KV residency                                  | The play          |
+| ------------- | --------------------------------------------- | ----------------- |
+| **AMASS-fast** | K+V resident in VRAM                          | speed             |
+| **AMASS-mem**  | V (mem-v) or K+V (mem-kv) offloaded to DRAM; only the quad summary + called pages resident | memory + speed |
+
+## Install
+
+```bash
+pip install amass-kv
+```
+
+The import name is `amass`:
+
+```python
+import amass
+print(amass.__version__)
+```
+
+`amass-kv` declares `torch` and `triton` as its runtime dependencies. vLLM is an
+**optional extra** (see [Dependencies](#dependencies)); install AMASS into your
+existing vLLM environment, or pull a compatible engine with:
+
+```bash
+pip install "amass-kv[vllm]"     # engine + plugin
+pip install "amass-kv[all]"      # + transformers (bench) + pyyaml (yaml configs)
+```
+
+Requires Python >= 3.10 and an NVIDIA GPU. The hot-path CUDA kernels are
+compiled once at first use via `torch.utils.cpp_extension` (targeting Hopper,
+`sm_90a`); the reference Triton kernels are the automatic fallback.
+
+## Quickstart
+
+AMASS installs as a vLLM **general plugin**, so once the wheel is present vLLM
+auto-registers it at engine startup with no code change: AMASS patches itself in
+wherever vLLM would have chosen the FlashAttention backend. Configuration is a
+single `AmassConfig`, sourced from the `AMASS_CONFIG` environment variable (an
+inline JSON string, or a path to a JSON/YAML file):
+
+```bash
+# Run any vLLM entry point (serve / offline) with AMASS active.
+export AMASS_CONFIG='{"variant": "fast", "coverage": 0.95}'
+vllm serve meta-llama/Llama-3.1-8B-Instruct
+
+# The memory play: offload V to DRAM, keep K resident.
+export AMASS_CONFIG='{"variant": "mem-v", "coverage": 0.95}'
+
+# Turn AMASS off (stock FlashAttention / FullKV reference line):
+export AMASS_DISABLE=1
+```
+
+Key `AmassConfig` fields: `variant` (`fast` / `mem-v` / `mem-kv`), `coverage`
+(adaptive per-head nucleus coverage over exact page masses; default 0.95) or
+`budget` (a fixed selected-page fraction), `score` (`quad` default, or the `r8`
+low-rank fallback), and `use_cuda` (hand-CUDA hot path vs Triton reference).
+
+## Headline results (measured)
+
+- **3.28x smaller selector state.** The quad summary is 4.7 percent of the KV
+  cache vs 15.6 percent for the low-rank `r8` baseline (1201 vs 3943 MiB
+  measured), with a cheaper tail-free hot-path kernel.
+- **Near-lossless on LongBench.** quad at a 5 percent per-step budget lands
+  within noise of FullKV (LongBench-v1 delta about -0.22 for quad vs -0.38 for
+  r8 at the same budget; about -0.29 at coverage 0.95 across 14 subsets).
+- **Faster in the regime that matters.** AMASS-fast beats dense FlashAttention-3
+  in the **long-context, high-batch** decode regime by up to about **2x**, and
+  is at **parity at batch size 1** (the hot path is bandwidth-bound and only
+  pulls ahead once the dense KV read dominates). It does not claim >=2x
+  everywhere.
+- **76 to 92 percent VRAM saved** in the mem variant, by holding only the quad
+  summary plus the called pages resident and serving the rest of V from DRAM.
+
+Numbers are per-kernel and end-to-end reproducible with `amass-bench` (below).
+See the paper for the full protocol, benchmarks (RULER, LongBench v1/v2,
+SCBench, a reasoning suite), and ablations.
+
+- Paper: see the project page linked in this repository.
+
+## Benchmarking: `amass-bench`
+
+The wheel ships a benchmarker so the latency claims are reproducible on any
+Hopper box with one command:
+
+```bash
+amass-bench kernels --quick    # per-kernel: r8_score / topb / decode, CUDA vs Triton vs dense-FA
+amass-bench e2e --ctx 16384    # end-to-end decode-step TPOT in a real vLLM run
+amass-bench all --json out.json
+python -m amass.bench kernels  # identical to the console entry point
+```
+
+`amass-bench e2e` needs the `bench` + `vllm` extras (it tokenizes a real model
+and spins up an engine). Timing is CUDA-event based with warmup and medians, and
+reports both eager and CUDA-graph-replay speedups plus an HBM roofline for the
+bandwidth-bound score kernel. Full flag reference: `amass-bench --help`.
+
+## Dependencies
+
+| Dependency     | Why                                              | How it is declared        |
+| -------------- | ------------------------------------------------ | ------------------------- |
+| `torch`        | tensors + runtime CUDA (`load_inline`) kernels   | hard (`>=2.4`)            |
+| `triton`       | reference / fallback kernels                     | hard (`>=3.0`)            |
+| `vllm`         | plugin host + attention backend                  | extra `[vllm]` (`>=0.8`)  |
+| `transformers` | `amass-bench e2e` tokenizer (lazy import)         | extra `[bench]`           |
+| `pyyaml`       | YAML `AMASS_CONFIG` files (JSON needs no dep)      | extra `[yaml]`            |
+
+`numpy` is **not** a direct dependency: nothing under `amass/` imports it; it
+arrives transitively through torch. vLLM is an extra rather than a hard floor
+because it is a heavy, CUDA-ABI-sensitive wheel that production deployments pin
+themselves, and a hard floor could silently upgrade a pinned engine.
+
+## License
+
+Apache-2.0 (provisional; see the note at the top). See `LICENSE` for the full
+text.

amass_kv-0.1.0/README.md

@@ -0,0 +1,142 @@
+# AMASS
+
+**Decode-time sparse KV-cache attention for vLLM.** AMASS keeps LLM decoding
+fast and memory-light at long context by attending, at every decode step, to
+only the pages that actually carry attention mass: a compact per-page score
+selects a small working set (typically 5 to 10 percent of the KV cache), and a
+sparse paged-attention kernel reads just those pages. Prefill is untouched, so
+time-to-first-token is unchanged.
+
+> **License note (provisional).** This release is distributed under Apache-2.0
+> as a placeholder pending final confirmation by the authors. See `LICENSE`.
+
+## What is in the box
+
+AMASS has two parts that share one selection core:
+
+1. **The quad selector (Stage A).** A quadratic, Gaussian-MGF page score built
+   from a tiny per-page summary. "quad" drops the per-key coordinates and stores
+   only `r'` eigenvalues per page, which shrinks the resident selector state by
+   **3.28x** (from 15.6 percent down to 4.7 percent of the KV cache; measured
+   1201 MiB vs 3943 MiB) while remaining accurate enough to pick the right pages.
+2. **The DRAM value tier (Stage B).** A pinned, device-mapped host-DRAM pool for
+   the value cache, with a bounded resident hot buffer plus staging pool. Only
+   the ~5 percent quad summary and the pages actually called each step stay in
+   VRAM; the rest of V (or K+V) lives in host memory.
+
+### Two variants, one algorithm
+
+| Variant       | KV residency                                  | The play          |
+| ------------- | --------------------------------------------- | ----------------- |
+| **AMASS-fast** | K+V resident in VRAM                          | speed             |
+| **AMASS-mem**  | V (mem-v) or K+V (mem-kv) offloaded to DRAM; only the quad summary + called pages resident | memory + speed |
+
+## Install
+
+```bash
+pip install amass-kv
+```
+
+The import name is `amass`:
+
+```python
+import amass
+print(amass.__version__)
+```
+
+`amass-kv` declares `torch` and `triton` as its runtime dependencies. vLLM is an
+**optional extra** (see [Dependencies](#dependencies)); install AMASS into your
+existing vLLM environment, or pull a compatible engine with:
+
+```bash
+pip install "amass-kv[vllm]"     # engine + plugin
+pip install "amass-kv[all]"      # + transformers (bench) + pyyaml (yaml configs)
+```
+
+Requires Python >= 3.10 and an NVIDIA GPU. The hot-path CUDA kernels are
+compiled once at first use via `torch.utils.cpp_extension` (targeting Hopper,
+`sm_90a`); the reference Triton kernels are the automatic fallback.
+
+## Quickstart
+
+AMASS installs as a vLLM **general plugin**, so once the wheel is present vLLM
+auto-registers it at engine startup with no code change: AMASS patches itself in
+wherever vLLM would have chosen the FlashAttention backend. Configuration is a
+single `AmassConfig`, sourced from the `AMASS_CONFIG` environment variable (an
+inline JSON string, or a path to a JSON/YAML file):
+
+```bash
+# Run any vLLM entry point (serve / offline) with AMASS active.
+export AMASS_CONFIG='{"variant": "fast", "coverage": 0.95}'
+vllm serve meta-llama/Llama-3.1-8B-Instruct
+
+# The memory play: offload V to DRAM, keep K resident.
+export AMASS_CONFIG='{"variant": "mem-v", "coverage": 0.95}'
+
+# Turn AMASS off (stock FlashAttention / FullKV reference line):
+export AMASS_DISABLE=1
+```
+
+Key `AmassConfig` fields: `variant` (`fast` / `mem-v` / `mem-kv`), `coverage`
+(adaptive per-head nucleus coverage over exact page masses; default 0.95) or
+`budget` (a fixed selected-page fraction), `score` (`quad` default, or the `r8`
+low-rank fallback), and `use_cuda` (hand-CUDA hot path vs Triton reference).
+
+## Headline results (measured)
+
+- **3.28x smaller selector state.** The quad summary is 4.7 percent of the KV
+  cache vs 15.6 percent for the low-rank `r8` baseline (1201 vs 3943 MiB
+  measured), with a cheaper tail-free hot-path kernel.
+- **Near-lossless on LongBench.** quad at a 5 percent per-step budget lands
+  within noise of FullKV (LongBench-v1 delta about -0.22 for quad vs -0.38 for
+  r8 at the same budget; about -0.29 at coverage 0.95 across 14 subsets).
+- **Faster in the regime that matters.** AMASS-fast beats dense FlashAttention-3
+  in the **long-context, high-batch** decode regime by up to about **2x**, and
+  is at **parity at batch size 1** (the hot path is bandwidth-bound and only
+  pulls ahead once the dense KV read dominates). It does not claim >=2x
+  everywhere.
+- **76 to 92 percent VRAM saved** in the mem variant, by holding only the quad
+  summary plus the called pages resident and serving the rest of V from DRAM.
+
+Numbers are per-kernel and end-to-end reproducible with `amass-bench` (below).
+See the paper for the full protocol, benchmarks (RULER, LongBench v1/v2,
+SCBench, a reasoning suite), and ablations.
+
+- Paper: see the project page linked in this repository.
+
+## Benchmarking: `amass-bench`
+
+The wheel ships a benchmarker so the latency claims are reproducible on any
+Hopper box with one command:
+
+```bash
+amass-bench kernels --quick    # per-kernel: r8_score / topb / decode, CUDA vs Triton vs dense-FA
+amass-bench e2e --ctx 16384    # end-to-end decode-step TPOT in a real vLLM run
+amass-bench all --json out.json
+python -m amass.bench kernels  # identical to the console entry point
+```
+
+`amass-bench e2e` needs the `bench` + `vllm` extras (it tokenizes a real model
+and spins up an engine). Timing is CUDA-event based with warmup and medians, and
+reports both eager and CUDA-graph-replay speedups plus an HBM roofline for the
+bandwidth-bound score kernel. Full flag reference: `amass-bench --help`.
+
+## Dependencies
+
+| Dependency     | Why                                              | How it is declared        |
+| -------------- | ------------------------------------------------ | ------------------------- |
+| `torch`        | tensors + runtime CUDA (`load_inline`) kernels   | hard (`>=2.4`)            |
+| `triton`       | reference / fallback kernels                     | hard (`>=3.0`)            |
+| `vllm`         | plugin host + attention backend                  | extra `[vllm]` (`>=0.8`)  |
+| `transformers` | `amass-bench e2e` tokenizer (lazy import)         | extra `[bench]`           |
+| `pyyaml`       | YAML `AMASS_CONFIG` files (JSON needs no dep)      | extra `[yaml]`            |
+
+`numpy` is **not** a direct dependency: nothing under `amass/` imports it; it
+arrives transitively through torch. vLLM is an extra rather than a hard floor
+because it is a heavy, CUDA-ABI-sensitive wheel that production deployments pin
+themselves, and a hard floor could silently upgrade a pinned engine.
+
+## License
+
+Apache-2.0 (provisional; see the note at the top). See `LICENSE` for the full
+text.

amass_kv-0.1.0/pyproject.toml

@@ -0,0 +1,89 @@
+# pyproject.toml -- standalone release metadata for the AMASS pip package.
+#
+# This tree is SELF-CONTAINED: it is not the research-repo root pyproject (that
+# one builds the separate `kvcomp` harness). `packaging/build.sh` stages a
+# snapshot of the live `src/amass` into `src/amass` next to this file, then
+# builds the sdist + wheel from here.
+#
+# Distribution name = `amass-kv` (PyPI). Import name = `amass` (`import amass`).
+
+[build-system]
+requires = ["setuptools>=77", "wheel"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "amass-kv"
+version = "0.1.0"
+description = "AMASS: decode-time sparse KV-cache attention for vLLM (quad page-score selector + DRAM value tier)"
+readme = "README.md"
+requires-python = ">=3.10"
+license = "Apache-2.0"
+license-files = ["LICENSE"]
+authors = [
+  { name = "Junsung Hwang", email = "junsung.k.hwang@gmail.com" },
+]
+keywords = [
+  "vllm", "kv-cache", "sparse-attention", "llm-inference",
+  "decode", "long-context", "flash-attention", "cuda",
+]
+classifiers = [
+  "Development Status :: 4 - Beta",
+  "Intended Audience :: Developers",
+  "Intended Audience :: Science/Research",
+  "Operating System :: POSIX :: Linux",
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Topic :: Scientific/Engineering :: Artificial Intelligence",
+  "Environment :: GPU :: NVIDIA CUDA",
+]
+
+# Runtime deps that are ALWAYS needed once you touch a kernel (the whole point
+# of the package). Kept minimal + floors-only so AMASS drops into an existing
+# CUDA/torch deployment without fighting the resolver. `numpy` is intentionally
+# absent: nothing under `amass/` imports it directly (it arrives transitively
+# via torch). See README "Dependencies".
+dependencies = [
+  "torch>=2.4",
+  "triton>=3.0",
+]
+
+[project.optional-dependencies]
+# vLLM is the plugin host. It is REQUIRED for the backend/plugin path but is a
+# heavy, CUDA-ABI-sensitive wheel that real deployments pin themselves, so we
+# ship it as an extra rather than a hard floor that could upgrade a pinned
+# engine out from under the user. Install AMASS into your existing vLLM env, or
+# `pip install amass-kv[vllm]` to pull a compatible engine.
+vllm = ["vllm>=0.8"]
+# YAML `AMASS_CONFIG` files (JSON configs work with no extra dep).
+yaml = ["pyyaml>=6.0"]
+# `amass-bench e2e` tokenizes a real model (lazy `transformers` import); it also
+# needs the `vllm` extra to spin up an engine.
+bench = ["transformers>=4.40"]
+# Everything for the full plugin + bench + yaml experience.
+all = ["vllm>=0.8", "transformers>=4.40", "pyyaml>=6.0"]
+
+[project.urls]
+Homepage = "https://github.com/Js-Hwang1/amass-kv"
+Repository = "https://github.com/Js-Hwang1/amass-kv"
+Issues = "https://github.com/Js-Hwang1/amass-kv/issues"
+
+# Console script: `amass-bench <kernels|e2e|all>`.
+[project.scripts]
+amass-bench = "amass.bench.cli:main"
+
+# vLLM general-plugin auto-registration: with the wheel installed, vLLM calls
+# this at engine startup with NO import needed in user code, and AMASS patches
+# itself in as the attention backend.
+[project.entry-points."vllm.general_plugins"]
+amass = "amass.backend.register:register"
+
+[tool.setuptools.packages.find]
+where = ["src"]
+include = ["amass*"]
+
+[tool.setuptools.package-data]
+# The CUDA kernels are `.py` sources compiled at runtime via `load_inline`, so
+# they ship as ordinary modules (no data files). Nothing extra to include.
+amass = []

amass_kv-0.1.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

amass_kv-0.1.0/src/amass/__init__.py

@@ -0,0 +1,31 @@
+"""AMASS — decode-time certified page selection for KV-cache attention, with an
+optional DRAM offload tier. Two build targets share one algorithm core:
+
+  * AMASS-fast  — K+V resident (the speed play).
+  * AMASS-mem   — V (or K+V) offloaded to DRAM (default; the memory play).
+
+Public API:
+  from amass import AmassConfig, register        # register() = vLLM plugin hook
+
+Architecture (ours_doc/AMASS_DESIGN.md):
+  config.py     AmassConfig — single source of truth (no scattered env vars)
+  selection/    Stage A: decode-time page selection (SHARED by both variants)
+  attention/    Stage B: sparse paged attention (V source injected as a seam)
+  tier/         DRAM offload state machine (mem variants only)
+  backend/      vLLM integration: attention impl, metadata builder, register()
+"""
+from __future__ import annotations
+
+from .config import AmassConfig  # noqa: F401
+
+__all__ = ["AmassConfig", "register"]
+
+__version__ = "0.1.0"
+
+
+def register() -> None:
+    """vLLM general-plugin entry point. Thin re-export of
+    :func:`amass.backend.register.register` so the entry point is stable while
+    the backend is ported."""
+    from .backend.register import register as _register
+    _register()