File size: 1,465 Bytes
8b93567
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
Astro Emulators Toolkit Bundle

Summary:
  model: mlp
  task: regression
  fit_method: gradient
  x_names: ('teff', 'logg', 'feh')
  y_names: ('flux',)

Domain:
  input_domain: {'kind': 'box_v1', 'space': 'runtime_tree_v1', 'storage': {'format': 'safetensors_v1', 'filename': 'input_domain.safetensors', 'layout': 'tree_minmax_v1'}}

Provenance:
  toolkit_version: 0.1.0
  saved_at_utc: 2026-03-13T02:55:33.972802+00:00
  python_version: 3.12.11
  git_sha: 72e167d2c8d68abc1e2c6c4b2503e06f3a375537

spec:
spec_version: 2
x:
  names:
    - teff
    - logg
    - feh
  meaning:
    - effective temperature
    - surface gravity
    - metallicity [Fe/H]
  units:
    - K
    - dex
    - dex
y:
  names:
    - flux
  meaning:
    - continuum-normalized flux vector on the shared wavelength grid
  units:
    - dimensionless
  wavelength_angstrom:
    array(shape=(500,), dtype=float64)
input_domain:
  kind: box_v1
  space: runtime_tree_v1
  storage:
    format: safetensors_v1
    filename: input_domain.safetensors
    layout: tree_minmax_v1
reference_scaling:
  kind: affine_v1
  parametrization: minmax
  space: runtime_tree_v1
  storage:
    format: safetensors_v1
    filename: reference_scaling.safetensors
    layout: tree_minmax_v1
notes: Inputs are min-max scaled before training using the fixed bounds stored in this bundle. Output reference scaling is identity.

Note: this bundle is the canonical emulator artifact. Physical-space composition is external.