murabcd commited on
Commit
77eb3d0
·
verified ·
1 Parent(s): bdceab0

docs: center fev in model comparison

Browse files
Files changed (3) hide show
  1. MODEL_CARD.md +3 -3
  2. README.md +3 -3
  3. manifest.json +4 -4
MODEL_CARD.md CHANGED
@@ -114,10 +114,10 @@ Checkpoint and adapter files retain their original precision and hashes. Exact a
114
 
115
  The new common native typed-decision comparison against pinned Kev-4B and Jev is documented in [BENCHMARKS.md](BENCHMARKS.md), with separate public-development and authored-Russian denominators, probability metrics and reproducibility commands. It is distinct from the routing-composition evidence below.
116
 
117
- | Native comparison | Fev‑4B | Kev‑4B | Jev‑1.14.0 |
118
  |---|---:|---:|---:|
119
- | Public development | 83.6% | 80.2% | 84.6% |
120
- | Russian probes | 100.0% | 99.3% | 100.0% |
121
 
122
  Public development: 764 questions. Authored Russian probes: 144 questions.
123
 
 
114
 
115
  The new common native typed-decision comparison against pinned Kev-4B and Jev is documented in [BENCHMARKS.md](BENCHMARKS.md), with separate public-development and authored-Russian denominators, probability metrics and reproducibility commands. It is distinct from the routing-composition evidence below.
116
 
117
+ | Native comparison | Kev‑4B | **Fev‑4B** | Jev‑1.14.0 |
118
  |---|---:|---:|---:|
119
+ | Public development | 80.2% | **83.6%** | 84.6% |
120
+ | Russian probes | 99.3% | **100.0%** | 100.0% |
121
 
122
  Public development: 764 questions. Authored Russian probes: 144 questions.
123
 
README.md CHANGED
@@ -114,10 +114,10 @@ Checkpoint and adapter files retain their original precision and hashes. Exact a
114
 
115
  The new common native typed-decision comparison against pinned Kev-4B and Jev is documented in [BENCHMARKS.md](BENCHMARKS.md), with separate public-development and authored-Russian denominators, probability metrics and reproducibility commands. It is distinct from the routing-composition evidence below.
116
 
117
- | Native comparison | Fev‑4B | Kev‑4B | Jev‑1.14.0 |
118
  |---|---:|---:|---:|
119
- | Public development | 83.6% | 80.2% | 84.6% |
120
- | Russian probes | 100.0% | 99.3% | 100.0% |
121
 
122
  Public development: 764 questions. Authored Russian probes: 144 questions.
123
 
 
114
 
115
  The new common native typed-decision comparison against pinned Kev-4B and Jev is documented in [BENCHMARKS.md](BENCHMARKS.md), with separate public-development and authored-Russian denominators, probability metrics and reproducibility commands. It is distinct from the routing-composition evidence below.
116
 
117
+ | Native comparison | Kev‑4B | **Fev‑4B** | Jev‑1.14.0 |
118
  |---|---:|---:|---:|
119
+ | Public development | 80.2% | **83.6%** | 84.6% |
120
+ | Russian probes | 99.3% | **100.0%** | 100.0% |
121
 
122
  Public development: 764 questions. Authored Russian probes: 144 questions.
123
 
manifest.json CHANGED
@@ -27,16 +27,16 @@
27
  "sha256": "bbedc3fda3305820b977265f01b8619d87570a6739de3a5582c3464840f1e57a"
28
  },
29
  "MODEL_CARD.md": {
30
- "bytes": 11694,
31
- "sha256": "303f4bc9ef027af06619a0e200257d2f08c67c1d340513c13889039dcc7d014a"
32
  },
33
  "NOTICE": {
34
  "bytes": 132,
35
  "sha256": "381fb81e814716a5ef8e5eae26a08607d052cf12f019dba9da9e81298ef5cb75"
36
  },
37
  "README.md": {
38
- "bytes": 11694,
39
- "sha256": "303f4bc9ef027af06619a0e200257d2f08c67c1d340513c13889039dcc7d014a"
40
  },
41
  "RUNNING.md": {
42
  "bytes": 7334,
 
27
  "sha256": "bbedc3fda3305820b977265f01b8619d87570a6739de3a5582c3464840f1e57a"
28
  },
29
  "MODEL_CARD.md": {
30
+ "bytes": 11706,
31
+ "sha256": "48e68df982a747d2ffde857f03ec16bcfb942eb80b8033e8233c050238b13961"
32
  },
33
  "NOTICE": {
34
  "bytes": 132,
35
  "sha256": "381fb81e814716a5ef8e5eae26a08607d052cf12f019dba9da9e81298ef5cb75"
36
  },
37
  "README.md": {
38
+ "bytes": 11706,
39
+ "sha256": "48e68df982a747d2ffde857f03ec16bcfb942eb80b8033e8233c050238b13961"
40
  },
41
  "RUNNING.md": {
42
  "bytes": 7334,