aprender 0.66.0

Next-generation ML framework in pure Rust — `cargo install aprender` for the `apr` CLI
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

<!-- PCU: examples-showcase-benchmark | contract: contracts/apr-page-examples-showcase-benchmark-v1.yaml -->
<!-- Example: cargo run -p aprender-core --example none -->
<!-- Status: enforced -->

# Showcase Benchmark

This example demonstrates the Qwen2.5-Coder showcase benchmark harness for measuring inference performance against baselines like Ollama and llama.cpp.

## 🏆 SHOWCASE COMPLETE (2026-01-18)

**CORRECTNESS-012 fixed.** Comparator ratios are deliberately absent from this
page: every one that used to appear here was derived from a hardcoded comparator
baseline that no run ever measured (`PP-LLAMA-001-MASTER.md`, formerly §9). A ratio is published
only from a signed receipt that recorded both sides.

### Qwen2.5-Coder-1.5B Results

| Format | M=8 | M=16 | M=32 | Status |
|--------|-----|------|------|--------|
| **GGUF** | 770.0 tok/s | **851.8 tok/s** | 812.8 tok/s | ✅ PASS |

### Key Achievements

- **GGUF GPU**: 851.8 tok/s measured. No comparator was run, so no ratio is stated.
- **CPU/GPU Parity**: Verified - outputs match exactly
- **APR Format**: Quantization preserved (Q4_K, Q6_K) through GGUF → APR conversion
- **File Size**: 1.9GB APR file with full model fidelity

### Run the Showcase

```bash
# APR GPU Benchmark (FEATURED)
MODEL_PATH=/path/to/qwen2.5-coder-1.5b-instruct-q4_k_m.gguf \
  cargo run --example apr_gpu_benchmark --release --features cuda

# Full showcase benchmark suite
cargo run --release --example showcase_benchmark
```

## Overview

The `showcase_benchmark` example implements:

- Automated model downloading from Hugging Face
- Side-by-side benchmarking against Ollama
- Performance visualization
- Regression detection
- GGUF → APR conversion with quantization preservation

## Test Matrix

| Model | Size | GPU Target | GPU Achieved | CPU Target |
|-------|------|------------|--------------|------------|
| Qwen2.5-Coder-0.5B | 490MB | 500+ tok/s | — | 150+ tok/s |
| Qwen2.5-Coder-1.5B | 1.1GB | 350+ tok/s | **824.7 tok/s** ✅ | 75+ tok/s |
| Qwen2.5-Coder-7B | 4.4GB | 150+ tok/s | — | 25+ tok/s |
| Qwen2.5-Coder-32B | 19GB | 40+ tok/s | — | 6+ tok/s |

## Metrics

- **Throughput**: Tokens per second (decode phase)
- **Prefill**: Prompt processing speed
- **TTFT**: Time to first token
- **Memory**: Peak VRAM/RAM usage

## See Also

- Qwen2.5-Coder Showcase Spec
- [Benchmark Comparison](./bench-comparison.md)