Kakeya-LLM-Inference-engine/results/platform-tests/bench_gemma4_26b_mac.json at main · FluffyAIcode/Kakeya-LLM-Inference-engine · GitHub

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
{
  "kind": "mlx_kakeya_deployment_benchmark",
  "config": {
    "model_id": "models/gemma-4-26B-A4B-it-mlx-4bit",
    "context_lengths": [
      512,
      2048,
      8192
    ],
    "gen_tokens": 64,
    "sink_size": 4,
    "window_size": 64
  },
  "env": {
    "mlx_version": "0.31.2"
  },
  "results": [
    {
      "context_length": 512,
      "kakeya": {
        "error": "TypeError: make_sink_window_cache() takes 1 positional argument but 3 were given"
      },
      "vanilla": {
        "prefill_s": 9.1962,
        "decode_s": 4.4363,
        "decode_tokens": 63,
        "decode_tokens_per_s": 14.201,
        "kv_bytes": 129536000,
        "peak_memory_bytes": 16024951996
      }
    },
    {
      "context_length": 2048,
      "kakeya": {
        "error": "TypeError: make_sink_window_cache() takes 1 positional argument but 3 were given"
      },
      "vanilla": {
        "prefill_s": 7.6443,
        "decode_s": 5.9676,
        "decode_tokens": 63,
        "decode_tokens_per_s": 10.557,
        "kv_bytes": 475566080,
        "peak_memory_bytes": 17317280424
      }
    },
    {
      "context_length": 8192,
      "kakeya": {
        "error": "TypeError: make_sink_window_cache() takes 1 positional argument but 3 were given"
      },
      "vanilla": {
        "prefill_s": 45.3409,
        "decode_s": 20.705,
        "decode_tokens": 63,
        "decode_tokens_per_s": 3.043,
        "kv_bytes": 1859686400,
        "peak_memory_bytes": 22526543070
      }
    }
  ]
}