Hugging Face's logo

AdaptiveChunking
/
emactl_noema_s2

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
f2bd1f6
·
verified ·
1 Parent(s): a2d0ed3

run records (split from bundled legacy repo)

Browse files
Files changed (2) hide show
  1. final.json +114 -0
  2. log.jsonl +0 -0
final.json ADDED
@@ -0,0 +1,114 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "event": "final",
3
+ "run": "emactl_noema_s2",
4
+ "condition": "A",
5
+ "step": 6000,
6
+ "params": 22516992,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.22863874044704935,
10
+ "cps": 31.0859375,
11
+ "bpb": 2.0440929497706244,
12
+ "bytes_per_chunk": 4.447226675488375
13
+ },
14
+ "de": {
15
+ "cpc": 0.23143232844689038,
16
+ "cps": 37.125,
17
+ "bpb": 2.229200510378607,
18
+ "bytes_per_chunk": 4.252719353033265
19
+ },
20
+ "vi": {
21
+ "cpc": 0.2374830485489558,
22
+ "cps": 34.203125,
23
+ "bpb": 1.9064757324673185,
24
+ "bytes_per_chunk": 5.540809316973674
25
+ },
26
+ "tr": {
27
+ "cpc": 0.2687952273103905,
28
+ "cps": 38.015625,
29
+ "bpb": 2.1740123956039623,
30
+ "bytes_per_chunk": 3.9388166000540914
31
+ },
32
+ "fi": {
33
+ "cpc": 0.23316144808888417,
34
+ "cps": 34.265625,
35
+ "bpb": 2.3927762408084456,
36
+ "bytes_per_chunk": 4.315731419210918
37
+ },
38
+ "ru": {
39
+ "cpc": 0.3329287598944591,
40
+ "cps": 49.2890625,
41
+ "bpb": 1.3761500732624246,
42
+ "bytes_per_chunk": 5.276809887577121
43
+ },
44
+ "el": {
45
+ "cpc": 0.34005749646737804,
46
+ "cps": 54.5234375,
47
+ "bpb": 1.4288364316573776,
48
+ "bytes_per_chunk": 5.19611496531219
49
+ },
50
+ "hi": {
51
+ "cpc": 0.5375334323922734,
52
+ "cps": 70.65625,
53
+ "bpb": 1.15970682221278,
54
+ "bytes_per_chunk": 4.935961889698545
55
+ },
56
+ "ta": {
57
+ "cpc": 0.4450922706606044,
58
+ "cps": 65.0078125,
59
+ "bpb": 0.9492078075249067,
60
+ "bytes_per_chunk": 5.968602361084232
61
+ },
62
+ "th": {
63
+ "cpc": 0.6685732048492384,
64
+ "cps": 84.015625,
65
+ "bpb": 1.0939795622231419,
66
+ "bytes_per_chunk": 4.236396838992227
67
+ },
68
+ "zh": {
69
+ "cpc": 0.47307822553335593,
70
+ "cps": 21.828125,
71
+ "bpb": 3.1189088345478835,
72
+ "bytes_per_chunk": 5.605319990591656
73
+ },
74
+ "my": {
75
+ "cpc": 0.5314545070719865,
76
+ "cps": 78.671875,
77
+ "bpb": 1.1001802838363859,
78
+ "bytes_per_chunk": 5.20814169489505
79
+ }
80
+ },
81
+ "gini_cpc": 0.2097257925277538,
82
+ "gini_cps": 0.2205780004871769,
83
+ "bpic": 5.008103664265581,
84
+ "seed": 2,
85
+ "mean_boundary_entropy": 0.6694104671478271,
86
+ "dechunk": "ste",
87
+ "router_noise": 0.0,
88
+ "transitions": [
89
+ {
90
+ "step": 0,
91
+ "phase": 1,
92
+ "objective": "global",
93
+ "router_trainable": true,
94
+ "encoder_trainable": true,
95
+ "frozen_router_params": 0
96
+ },
97
+ {
98
+ "step": 2400,
99
+ "phase": 2,
100
+ "objective": "global",
101
+ "router_trainable": true,
102
+ "encoder_trainable": true,
103
+ "frozen_router_params": 0
104
+ },
105
+ {
106
+ "step": 4200,
107
+ "phase": 3,
108
+ "objective": "global",
109
+ "router_trainable": true,
110
+ "encoder_trainable": true,
111
+ "frozen_router_params": 0
112
+ }
113
+ ]
114
+ }
log.jsonl ADDED
The diff for this file is too large to render. See raw diff