File size: 2,659 Bytes
45b798c
 
 
86b06f4
 
 
 
45b798c
 
 
86b06f4
45b798c
 
 
 
 
 
 
 
 
 
 
 
 
 
541e34f
 
 
0d3a2ba
45b798c
86b06f4
0d3a2ba
13a0c89
0d3a2ba
 
 
 
 
 
 
 
 
 
 
 
 
45b798c
ec7f6ce
 
0d3a2ba
2751c6b
 
0d3a2ba
541e34f
 
 
 
 
 
 
 
 
 
 
 
0d3a2ba
 
ec7f6ce
45b798c
86b06f4
0d3a2ba
 
45b798c
0d3a2ba
45b798c
 
86b06f4
0d3a2ba
 
45b798c
0d3a2ba
45b798c
 
86b06f4
0d3a2ba
 
45b798c
0d3a2ba
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
# targets.yaml — declarative fleet spec for LEM-Eval workers.
#
# Each target is a (base, this) model pair that gets benchmarked together in
# a paired A/B run. The `type` field says which inference backend the target
# needs — 'mlx' runs on Apple Silicon via mlx_lm, 'gguf' runs on any machine
# that can serve GGUF (via Ollama or llama.cpp). Workers filter by type so
# partitioning across machines falls out of capability, not hostnames.
#
# Each target is an independent canon. Workers writing to different targets
# don't race because each target's .eval_results/ lives in a different model
# repo (and its results/<target>/ path in LEM-benchmarks is also disjoint).
#
# Editing this file is the way to change the fleet. After an edit, commit
# and push — workers pick up the new config on their next `git pull` cycle
# (handled by `lem-eval.sh maintain`).

task: mmlu_pro

defaults:
  n_questions: 1        # one new question per run
  rounds: 8             # full 8-PAC batch per question
  samples_start: auto   # derive from canon max(question_index) + 1

targets:

  # Lemer — Gemma 4 E2B. Six fully paired quant slots:
  #   mlx  Q4 / 8bit / BF16
  #   gguf Q4_K_M / Q8_0 / BF16

  - name: lemer
    type: mlx
    quant: Q4
    base: LetheanNetwork/lemer-mlx
    this: lthn/lemer-mlx

  - name: lemer
    type: mlx
    quant: 8bit
    base: LetheanNetwork/lemer-mlx-8bit
    this: lthn/lemer-mlx-8bit

  - name: lemer
    type: mlx
    quant: BF16
    base: LetheanNetwork/lemer-mlx-bf16
    this: lthn/lemer-mlx-bf16

  - name: lemer
    type: gguf
    quant: Q4_K_M
    base: hf.co/LetheanNetwork/lemer:Q4_K_M
    this: hf.co/lthn/lemer:Q4_K_M

  - name: lemer
    type: gguf
    quant: Q8_0
    base: hf.co/LetheanNetwork/lemer:Q8_0
    this: hf.co/lthn/lemer:Q8_0

  - name: lemer
    type: gguf
    quant: BF16
    base: hf.co/LetheanNetwork/lemer:BF16
    this: hf.co/lthn/lemer:BF16

  # Lemma / Lemmy / Lemrd — pending the same treatment as lemer.
  # (LetheanNetwork bases in matching formats, naming alignment, etc.)

  - name: lemma
    type: mlx
    quant: Q4
    base: mlx-community/gemma-4-e4b-it-4bit     # TODO: LetheanNetwork/lemma-mlx
    this: lthn/lemma
    notes: Gemma 4 E4B  pending family rollout

  - name: lemmy
    type: gguf
    quant: Q4_K_M
    base: mlx-community/gemma-4-26b-a4b-it-4bit # TODO: LetheanNetwork/lemmy gguf
    this: lthn/lemmy
    notes: Gemma 4 26B A4B MoE  pending family rollout

  - name: lemrd
    type: gguf
    quant: Q4_K_M
    base: mlx-community/gemma-4-31b-it-4bit     # TODO: LetheanNetwork/lemrd gguf
    this: lthn/lemrd
    notes: Gemma 4 31B  pending family rollout