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
//! Decode backend abstraction.
//!
//! Different backends (CUDA, Metal, CPU/Candle) implement `DecodeBackend`
//! to provide optimized decode execution. The `GenericDecodeExecutor` uses
//! a DecodeBackend to execute the decode hot path.
use crateTensorRef;
use crateTransformerWeights;
use Result;
/// Decode-phase execution backend.
///
/// Implements the actual computation for single-token decode steps.
/// Different backends optimize for different hardware:
/// - `CudaDecodeBackend`: cuBLAS + custom CUDA kernels, pre-allocated buffers
/// - `MetalDecodeBackend`: Metal compute shaders
/// - `CandleDecodeBackend`: candle tensor ops (CPU/fallback)
///
/// The backend is initialized with model weights and manages its own
/// internal state (KV cache, buffers, cuBLAS handles, etc.).