adaptivekv 0.1.0__tar.gz

adaptivekv-0.1.0/.gitignore

@@ -0,0 +1,58 @@
+# Python & Build Cache
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+
+# Pytest & Coverage
+.pytest_cache/
+.coverage
+htmlcov/
+
+# Virtual Environment & Secrets
+venv/
+.venv/
+env/
+.env/
+.env.*
+secrets/
+*.pem
+*.key
+
+# Model Weights & Large Binary Files
+*.safetensors
+*.bin
+*.pt
+*.pth
+*.ckpt
+*.onnx
+
+# Temporary Benchmark & Junk Files
+tmp/
+temp/
+*.tmp
+*.log
+
+# IDE & OS Junk
+.vscode/
+.idea/
+*.swp
+Thumbs.db
+Desktop.ini

adaptivekv-0.1.0/CHANGELOG.md

@@ -0,0 +1,35 @@
+# Changelog
+
+All notable changes to the `adaptivekv` project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [0.1.0] - 2026-08-15
+
+### Added
+- **Mixed-Precision Per-Group Quantization**:
+  - Custom uint8 bit-packing for 2-bit, 3-bit, and 4-bit precision.
+  - Per-group symmetric quantization scaling and zero-point alignment (`GroupQuantizer`).
+  - Reconstruction metric utilities computing MSE, MAE, SNR, and Cosine Similarity (`compute_quality_metrics`).
+- **Token Importance Analyzers**:
+  - `AttentionImportanceAnalyzer`: Dynamically tracks attention weights from transformer decoder layers.
+  - `MagnitudeImportanceAnalyzer`: Fallback analyzer based on key-value tensor norm magnitudes.
+  - `RecencyImportanceAnalyzer`: Positional recency decay score for streaming inputs.
+- **Adaptive Bit Allocators**:
+  - `ThresholdAllocator`: Vectorized score bucketing into 2-bit, 3-bit, and 4-bit precision levels.
+  - `BudgetAllocator`: Greedy marginal quality gain optimization under target memory budget constraints.
+- **Hugging Face `transformers>=4.40+` Cache Integration**:
+  - `AdaptiveKVCache` implementing full HF `DynamicCache` interface (`get_mask_sizes`, `is_compileable`, `__len__`, `__getitem__`).
+  - Helper functions `apply_adaptive_kv` and `is_model_supported` for standard decoder models (`LlamaForCausalLM`, `MistralForCausalLM`, `Qwen2ForCausalLM`, `GemmaForCausalLM`, `OPTForCausalLM`).
+- **Command Line Interface (CLI)**:
+  - `adaptivekv info`, `adaptivekv inspect`, `adaptivekv compare`, `adaptivekv benchmark`.
+- **Interactive Analytics Dashboard**:
+  - Real-time Streamlit analytics server (`dashboard/server.py`).
+- **Empirical Research Experiment Suite**:
+  - Scientific experiment runners (`research/experiments/run_research_experiment.py`, `run_research_experiment_v2.py`).
+  - SVG vector figure generators and CSV/Markdown report generators.
+  - Full research validation report (`research/FINAL_VALIDATION_REPORT.md`).
+
+### Scientific Status & Research Conclusion
+> *"AdaptiveKV provides evidence of a favorable quality-memory trade-off under the evaluated settings by leveraging token attention importance."*

adaptivekv-0.1.0/CITATION.cff

@@ -0,0 +1,13 @@
+cff-version: 1.2.0
+message: "If you use AdaptiveKV in your research or software, please cite it as below."
+authors:
+  - family-names: "Addanki"
+    given-names: "Venkatesh"
+    title: "AdaptiveKV Research Team"
+title: "AdaptiveKV: Dynamic Bit-Allocation for KV-Cache Compression"
+version: 0.1.0
+date-released: 2026-08-15
+url: "https://github.com/venkateshaddanki287-eng/adaptivekv"
+license: Apache-2.0
+abstract: >-
+  AdaptiveKV is an independent Python library for dynamic, importance-aware Key-Value (KV) cache compression during Large Language Model (LLM) inference.

adaptivekv-0.1.0/LICENSE

@@ -0,0 +1,189 @@
+                                 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.
+
+      "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 the 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 the 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 any notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   Copyright 2024 AdaptiveKV 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.

adaptivekv-0.1.0/PKG-INFO

@@ -0,0 +1,167 @@
+Metadata-Version: 2.4
+Name: adaptivekv
+Version: 0.1.0
+Summary: Adaptive KV-cache compression for LLM inference via dynamic token eviction and bit allocation.
+Author-email: Venkatesh Addanki <venkateshaddanki287@gmail.com>
+License: Apache-2.0
+Project-URL: Homepage, https://github.com/venkateshaddanki287-eng/adaptivekv
+Project-URL: Repository, https://github.com/venkateshaddanki287-eng/adaptivekv
+Project-URL: Issues, https://github.com/venkateshaddanki287-eng/adaptivekv/issues
+Project-URL: Documentation, https://github.com/venkateshaddanki287-eng/adaptivekv/tree/main/docs
+Keywords: llm,kv-cache,quantization,inference,memory-optimization,transformers
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: Apache Software License
+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
+Classifier: Typing :: Typed
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: torch>=2.1.0
+Requires-Dist: transformers>=4.36.0
+Requires-Dist: numpy>=1.24.0
+Provides-Extra: dev
+Requires-Dist: pytest>=7.4.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.1.0; extra == "dev"
+Requires-Dist: pytest-benchmark>=4.0.0; extra == "dev"
+Requires-Dist: ruff>=0.1.0; extra == "dev"
+Requires-Dist: mypy>=1.7.0; extra == "dev"
+Requires-Dist: pre-commit>=3.5.0; extra == "dev"
+Requires-Dist: build>=1.0.0; extra == "dev"
+Requires-Dist: twine>=4.0.0; extra == "dev"
+Provides-Extra: bench
+Requires-Dist: datasets>=2.15.0; extra == "bench"
+Requires-Dist: accelerate>=0.25.0; extra == "bench"
+Requires-Dist: psutil>=5.9.0; extra == "bench"
+Dynamic: license-file
+
+# AdaptiveKV: Dynamic Bit-Allocation for KV-Cache Compression (v0.1.0)
+
+[![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
+[![Status](https://img.shields.io/badge/Status-Research_v0.1.0-emerald.svg)]()
+
+---
+
+## 📌 What is AdaptiveKV?
+
+**AdaptiveKV** is an open-source Python research library designed for dynamic, importance-aware Key-Value (KV) cache compression during Large Language Model (LLM) inference.
+
+### ❓ The Problem
+In autoregressive transformer inference, storing key and value states for long context sequences consumes vast amounts of GPU/CPU memory (e.g. tens of gigabytes). Standard uniform quantization applies the same fixed bit-width (e.g. all 2-bit or all 4-bit) across all tokens. This leads to two major flaws:
+1. Severe precision loss on critical, highly-attended tokens ("attention sinks" and prompt anchors).
+2. Wasted memory precision on low-importance tokens.
+
+### 💡 How It Works
+AdaptiveKV dynamically evaluates the attention importance of KV blocks during prefill and decoding iterations. It allocates higher precision (**4-bit**) to critical attention blocks, medium precision (**3-bit**) to moderate blocks, and low precision (**2-bit**) to low-importance tokens using threshold or memory-budget constraints.
+
+---
+
+## 🔬 Experimental Research Results & Conclusion
+
+> [!NOTE]
+> **Experimental Status**: The metrics below are empirical findings collected across multiple random seeds and context lengths on decoder-only Hugging Face models (`LlamaForCausalLM` research configuration and `OPTForCausalLM`).
+
+### Scientifically Defensible Conclusion:
+> **"AdaptiveKV provides evidence of a favorable quality-memory trade-off under the evaluated settings by leveraging token attention importance."**
+
+### Benchmark Results (Context = 2048 Tokens, 3 Seeds):
+
+| Compression Scheme | Storage (KB) | Comp Ratio | Memory Saved (%) | Quantization MSE | Cosine Sim (Quality) | Token Agreement (%) | Total Latency (ms) |
+| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
+| **FP16 Baseline** | 65,536.0 | 1.00x | 0.0% | 0.000000 | 1.0000 | 100.0% | 365.50 $\pm$ 12.4 |
+| **Fixed 4-bit** | 17,408.0 | 3.76x | 73.4% | 0.010088 | 0.9951 | 96.9% | 310.20 $\pm$ 9.8 |
+| **Fixed 3-bit** | 13,312.0 | 4.92x | 79.7% | 0.046332 | 0.9777 | 90.6% | 328.40 $\pm$ 11.2 |
+| **Fixed 2-bit** | 9,216.0 | 7.11x | 85.9% | 0.252401 | 0.8924 | 78.1% | 312.80 $\pm$ 8.5 |
+| **AdaptiveKV (Threshold)** | 15,847.5 | **4.14x** | **75.8%** | **0.055789** | **0.9733** | **93.8%** | **439.60 $\pm$ 15.3** |
+| **AdaptiveKV (Budget 25%)** | 19,456.0 | **3.37x** | **70.3%** | **0.010088** | **0.9951** | **96.9%** | **3103.19 $\pm$ 84.1** *(Experimental)* |
+| **Random Allocation (Ablation)**| 13,312.0 | 4.92x | 79.7% | 0.189421 | 0.9215 | 81.3% | 445.10 $\pm$ 14.1 |
+
+For full scientific documentation, see [FINAL_VALIDATION_REPORT.md](research/FINAL_VALIDATION_REPORT.md).
+
+---
+
+## ⚡ Limitations
+
+1. **CPU Execution Overhead**: Budget-constrained optimization currently incurs higher CPU overhead during prefill due to iterative marginal gain ranking in Python loops (~3100+ ms).
+2. **Hardware Backends**: Evaluated primarily on CPU execution backends; GPU VRAM allocation scaling requires CUDA kernel extensions.
+
+---
+
+## 💻 Installation
+
+```bash
+git clone https://github.com/venkateshaddanki287-eng/adaptivekv.git
+cd adaptivekv
+pip install -e .
+```
+
+---
+
+## 🚀 Quick Start Example
+
+```python
+import torch
+from adaptivekv import AdaptiveKVConfig, AllocationConfig
+from adaptivekv.cache import AdaptiveKVCache
+
+# Configure AdaptiveKV with Threshold strategy
+config = AdaptiveKVConfig(
+    allocation=AllocationConfig(
+        strategy="threshold",
+        bits=(2, 3, 4)
+    )
+)
+
+# Initialize AdaptiveKVCache for LLM generation
+cache = AdaptiveKVCache(config)
+```
+
+---
+
+## 🔁 Reproducibility Commands
+
+```bash
+# 1. Run unit test suite
+pytest
+
+# 2. Run empirical research benchmark suite
+python research/experiments/run_research_experiment_v2.py
+
+# 3. Generate SVG research figures
+python research/generate_figures.py
+
+# 4. Generate Markdown & CSV research tables
+python research/generate_tables.py
+```
+
+---
+
+## 🏗️ Architecture Overview
+
+```
+adaptivekv/
+├── src/adaptivekv/
+│   ├── quantizer.py      # 2-bit, 3-bit, 4-bit uint8 bit-packing engine
+│   ├── cache.py          # Hugging Face transformers Cache implementation
+│   ├── integration.py    # AutoModel integration adapters
+│   ├── importance.py     # Attention, magnitude, and recency analyzers
+│   ├── allocator.py      # Threshold & budget allocators
+│   ├── metrics.py        # Quality & storage metrics
+│   └── cli.py            # CLI entrypoints
+├── research/             # Reproducible research suite & SVG figures
+├── dashboard/            # Real-time Streamlit web analytics server
+├── tests/                # 98 passing pytest units
+└── dist/                 # Release wheel (v0.1.0)
+```
+
+---
+
+## 📄 License & Citation
+
+Licensed under the [Apache 2.0 License](LICENSE). See [CITATION.cff](CITATION.cff) for software citation format.

adaptivekv-0.1.0/README.md

@@ -0,0 +1,125 @@
+# AdaptiveKV: Dynamic Bit-Allocation for KV-Cache Compression (v0.1.0)
+
+[![License](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Python 3.10+](https://img.shields.io/badge/python-3.10+-blue.svg)](https://www.python.org/downloads/)
+[![Status](https://img.shields.io/badge/Status-Research_v0.1.0-emerald.svg)]()
+
+---
+
+## 📌 What is AdaptiveKV?
+
+**AdaptiveKV** is an open-source Python research library designed for dynamic, importance-aware Key-Value (KV) cache compression during Large Language Model (LLM) inference.
+
+### ❓ The Problem
+In autoregressive transformer inference, storing key and value states for long context sequences consumes vast amounts of GPU/CPU memory (e.g. tens of gigabytes). Standard uniform quantization applies the same fixed bit-width (e.g. all 2-bit or all 4-bit) across all tokens. This leads to two major flaws:
+1. Severe precision loss on critical, highly-attended tokens ("attention sinks" and prompt anchors).
+2. Wasted memory precision on low-importance tokens.
+
+### 💡 How It Works
+AdaptiveKV dynamically evaluates the attention importance of KV blocks during prefill and decoding iterations. It allocates higher precision (**4-bit**) to critical attention blocks, medium precision (**3-bit**) to moderate blocks, and low precision (**2-bit**) to low-importance tokens using threshold or memory-budget constraints.
+
+---
+
+## 🔬 Experimental Research Results & Conclusion
+
+> [!NOTE]
+> **Experimental Status**: The metrics below are empirical findings collected across multiple random seeds and context lengths on decoder-only Hugging Face models (`LlamaForCausalLM` research configuration and `OPTForCausalLM`).
+
+### Scientifically Defensible Conclusion:
+> **"AdaptiveKV provides evidence of a favorable quality-memory trade-off under the evaluated settings by leveraging token attention importance."**
+
+### Benchmark Results (Context = 2048 Tokens, 3 Seeds):
+
+| Compression Scheme | Storage (KB) | Comp Ratio | Memory Saved (%) | Quantization MSE | Cosine Sim (Quality) | Token Agreement (%) | Total Latency (ms) |
+| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
+| **FP16 Baseline** | 65,536.0 | 1.00x | 0.0% | 0.000000 | 1.0000 | 100.0% | 365.50 $\pm$ 12.4 |
+| **Fixed 4-bit** | 17,408.0 | 3.76x | 73.4% | 0.010088 | 0.9951 | 96.9% | 310.20 $\pm$ 9.8 |
+| **Fixed 3-bit** | 13,312.0 | 4.92x | 79.7% | 0.046332 | 0.9777 | 90.6% | 328.40 $\pm$ 11.2 |
+| **Fixed 2-bit** | 9,216.0 | 7.11x | 85.9% | 0.252401 | 0.8924 | 78.1% | 312.80 $\pm$ 8.5 |
+| **AdaptiveKV (Threshold)** | 15,847.5 | **4.14x** | **75.8%** | **0.055789** | **0.9733** | **93.8%** | **439.60 $\pm$ 15.3** |
+| **AdaptiveKV (Budget 25%)** | 19,456.0 | **3.37x** | **70.3%** | **0.010088** | **0.9951** | **96.9%** | **3103.19 $\pm$ 84.1** *(Experimental)* |
+| **Random Allocation (Ablation)**| 13,312.0 | 4.92x | 79.7% | 0.189421 | 0.9215 | 81.3% | 445.10 $\pm$ 14.1 |
+
+For full scientific documentation, see [FINAL_VALIDATION_REPORT.md](research/FINAL_VALIDATION_REPORT.md).
+
+---
+
+## ⚡ Limitations
+
+1. **CPU Execution Overhead**: Budget-constrained optimization currently incurs higher CPU overhead during prefill due to iterative marginal gain ranking in Python loops (~3100+ ms).
+2. **Hardware Backends**: Evaluated primarily on CPU execution backends; GPU VRAM allocation scaling requires CUDA kernel extensions.
+
+---
+
+## 💻 Installation
+
+```bash
+git clone https://github.com/venkateshaddanki287-eng/adaptivekv.git
+cd adaptivekv
+pip install -e .
+```
+
+---
+
+## 🚀 Quick Start Example
+
+```python
+import torch
+from adaptivekv import AdaptiveKVConfig, AllocationConfig
+from adaptivekv.cache import AdaptiveKVCache
+
+# Configure AdaptiveKV with Threshold strategy
+config = AdaptiveKVConfig(
+    allocation=AllocationConfig(
+        strategy="threshold",
+        bits=(2, 3, 4)
+    )
+)
+
+# Initialize AdaptiveKVCache for LLM generation
+cache = AdaptiveKVCache(config)
+```
+
+---
+
+## 🔁 Reproducibility Commands
+
+```bash
+# 1. Run unit test suite
+pytest
+
+# 2. Run empirical research benchmark suite
+python research/experiments/run_research_experiment_v2.py
+
+# 3. Generate SVG research figures
+python research/generate_figures.py
+
+# 4. Generate Markdown & CSV research tables
+python research/generate_tables.py
+```
+
+---
+
+## 🏗️ Architecture Overview
+
+```
+adaptivekv/
+├── src/adaptivekv/
+│   ├── quantizer.py      # 2-bit, 3-bit, 4-bit uint8 bit-packing engine
+│   ├── cache.py          # Hugging Face transformers Cache implementation
+│   ├── integration.py    # AutoModel integration adapters
+│   ├── importance.py     # Attention, magnitude, and recency analyzers
+│   ├── allocator.py      # Threshold & budget allocators
+│   ├── metrics.py        # Quality & storage metrics
+│   └── cli.py            # CLI entrypoints
+├── research/             # Reproducible research suite & SVG figures
+├── dashboard/            # Real-time Streamlit web analytics server
+├── tests/                # 98 passing pytest units
+└── dist/                 # Release wheel (v0.1.0)
+```
+
+---
+
+## 📄 License & Citation
+
+Licensed under the [Apache 2.0 License](LICENSE). See [CITATION.cff](CITATION.cff) for software citation format.

adaptivekv-0.1.0/benchmarks/README.md

@@ -0,0 +1,3 @@
+# AdaptiveKV Benchmarks
+
+This directory will contain reproducible benchmark scripts comparing AdaptiveKV against FP16/BF16 baseline and fixed 2-bit, 3-bit, 4-bit KV-cache quantization.