bitpolar 0.3.3

BitPolar: near-optimal vector quantization with zero training overhead — 3-bit precision, provably unbiased inner products (ICLR 2026)
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154

# BitPolar Examples

Comprehensive examples for every language, integration, and use case — **30 Python examples** across 58 integrations.

## Rust Examples

Run with `cargo run --example <name>`:

### Search Patterns

| Example | Description |
|---|---|
| `search_vector_database` | Database compression + approximate nearest neighbor search |
| `search_brute_force` | Simplest pattern: encode all, scan all, return top-k |
| `search_oversampled` | Two-phase: approximate scoring + exact re-ranking for high recall |

### LLM / KV Cache

| Example | Description |
|---|---|
| `llm_kv_cache` | Single-head KV cache compression for transformers |
| `llm_multi_head_cache` | Multi-head attention KV cache (Llama-style) |

### Architecture Patterns

| Example | Description |
|---|---|
| `pattern_multi_tenant` | Per-tenant seed isolation for SaaS vector databases |
| `pattern_federated_search` | Independent nodes quantize separately, codes are comparable |

### Integrations

| Example | Description |
|---|---|
| `integration_tiered_storage` | Hot/Warm/Cold tiers with different bit-widths |
| `integration_adaptive` | Per-vector adaptive bit-width with promote/demote |

---

## Python Examples

Run with `python examples/python/<file>`:

### Core & Existing Integrations (01-11)

| # | Example | Integration | Prerequisites |
|---|---|---|---|
| 01 | [`01_quickstart.py`](python/01_quickstart.py) | Core BitPolar | `pip install bitpolar numpy` |
| 02 | [`02_huggingface_embeddings.py`](python/02_huggingface_embeddings.py) | HuggingFace sentence-transformers | `pip install bitpolar sentence-transformers` |
| 03 | [`03_langchain_vectorstore.py`](python/03_langchain_vectorstore.py) | LangChain VectorStore | `pip install bitpolar langchain-core` |
| 04 | [`04_qdrant_integration.py`](python/04_qdrant_integration.py) | Qdrant two-phase search | `pip install bitpolar qdrant-client` |
| 05 | [`05_vllm_kv_cache.py`](python/05_vllm_kv_cache.py) | vLLM KV cache + HF DynamicCache | `pip install bitpolar numpy` |
| 06 | [`06_agent_memory.py`](python/06_agent_memory.py) | AI agent episodic memory | `pip install bitpolar numpy` |
| 07 | [`07_mcp_tool_server.py`](python/07_mcp_tool_server.py) | MCP Tool for AI agents | `pip install bitpolar numpy` |
| 08 | [`08_finetuning.py`](python/08_finetuning.py) | Quantization-aware fine-tuning | `pip install bitpolar sentence-transformers torch` |
| 09 | [`09_vector_index.py`](python/09_vector_index.py) | In-memory vector search index | `pip install bitpolar numpy` |
| 10 | [`10_grpc_client.py`](python/10_grpc_client.py) | gRPC compression service | `pip install grpcio grpcio-tools` |
| 11 | [`11_cli_usage.sh`](python/11_cli_usage.sh) | CLI tool (compress/search/bench) | `cargo install --path bitpolar-cli` |

### ML Frameworks & Search (12-15)

| # | Example | Integration | Prerequisites |
|---|---|---|---|
| 12 | [`12_pytorch_quantizer.py`](python/12_pytorch_quantizer.py) | PyTorch torchao quantizer | `pip install bitpolar torch` |
| 13 | [`13_llamaindex_vectorstore.py`](python/13_llamaindex_vectorstore.py) | LlamaIndex VectorStore | `pip install bitpolar llama-index-core` |
| 14 | [`14_faiss_dropin.py`](python/14_faiss_dropin.py) | FAISS drop-in replacement | `pip install bitpolar numpy` |
| 15 | [`15_chromadb_integration.py`](python/15_chromadb_integration.py) | ChromaDB integration | `pip install bitpolar chromadb` |

### Agentic AI Frameworks (16-22)

| # | Example | Integration | Prerequisites |
|---|---|---|---|
| 16 | [`16_langgraph_checkpointer.py`](python/16_langgraph_checkpointer.py) | LangGraph checkpointer | `pip install bitpolar langgraph-checkpoint` |
| 17 | [`17_crewai_memory.py`](python/17_crewai_memory.py) | CrewAI memory backend | `pip install bitpolar` |
| 18 | [`18_openai_agents_tool.py`](python/18_openai_agents_tool.py) | OpenAI Agents SDK tool | `pip install bitpolar` |
| 19 | [`19_haystack_pipeline.py`](python/19_haystack_pipeline.py) | Haystack DocumentStore | `pip install bitpolar haystack-ai` |
| 20 | [`20_dspy_retriever.py`](python/20_dspy_retriever.py) | DSPy retriever module | `pip install bitpolar dspy` |
| 21 | [`21_mem0_backend.py`](python/21_mem0_backend.py) | Mem0 vector store | `pip install bitpolar` |
| 22 | [`22_sklearn_pipeline.py`](python/22_sklearn_pipeline.py) | scikit-learn Pipeline | `pip install bitpolar scikit-learn` |

### Comprehensive Integration Demos (23-30)

| # | Example | Coverage | Prerequisites |
|---|---|---|---|
| 23 | [`23_vector_databases.py`](python/23_vector_databases.py) | DuckDB, SQLite, Milvus, Weaviate, Pinecone, Redis, ES | `pip install bitpolar duckdb` |
| 24 | [`24_llm_inference.py`](python/24_llm_inference.py) | llama.cpp, SGLang, HF Transformers KV cache | `pip install bitpolar numpy` |
| 25 | [`25_ml_frameworks.py`](python/25_ml_frameworks.py) | PyTorch, JAX, TensorFlow, scikit-learn patterns | `pip install bitpolar numpy` |
| 26 | [`26_agent_memory.py`](python/26_agent_memory.py) | Zep, Letta (MemGPT), Mem0 memory backends | `pip install bitpolar numpy` |
| 27 | [`27_cloud_enterprise.py`](python/27_cloud_enterprise.py) | Vercel AI SDK, AWS Bedrock, Triton | `pip install bitpolar` |
| 28 | [`28_agentic_tools.py`](python/28_agentic_tools.py) | OpenAI, Google ADK, Anthropic, SmolAgents, PydanticAI | `pip install bitpolar` |
| 29 | [`29_supabase_neon.py`](python/29_supabase_neon.py) | Supabase, Neon serverless Postgres | `pip install bitpolar` |
| 30 | [`30_complete_rag.py`](python/30_complete_rag.py) | End-to-end RAG pipeline | `pip install bitpolar numpy` |

---

## JavaScript Examples

| # | Example | Runtime | Prerequisites |
|---|---|---|---|
| 01 | [`01_quickstart.js`](javascript/01_quickstart.js) | Browser (WASM) | `wasm-pack build bitpolar-wasm --target web` |
| 02 | [`02_node_bindings.js`](javascript/02_node_bindings.js) | Node.js (NAPI-RS) | `cd bitpolar-node && npm run build` |

---

## Go Example

| Example | Prerequisites |
|---|---|
| [`main.go`](go/main.go) | `cargo build --release --features ffi` |

```bash
cd examples/go
CGO_LDFLAGS="-L../../target/release -lbitpolar" go run main.go
```

---

## Java Example

| Example | Prerequisites |
|---|---|
| [`BitPolarDemo.java`](java/BitPolarDemo.java) | `cargo build --release -p bitpolar-jni` |

```bash
javac examples/java/BitPolarDemo.java -d build
java -Djava.library.path=target/release -cp build BitPolarDemo
```

---

## Browser Demo

```bash
cd bitpolar-demo && python3 -m http.server 9090
# Open http://localhost:9090
```

---

## PostgreSQL

```sql
SELECT bitpolar_compress(ARRAY[0.1, 0.2, ...]::float4[], 4);
SELECT bitpolar_inner_product(compressed_col, query_array, 128, 4, 42)
FROM items ORDER BY 1 DESC LIMIT 10;
```

---

## gRPC Service

```bash
BITPOLAR_DIM=384 BITPOLAR_BITS=4 cargo run -p bitpolar-server
```