Hugging Face's logo

AdaptiveChunking
/
spec_A_s0

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
1a75fb0
·
verified ·
1 Parent(s): 132352a

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": "spec_A_s0",
4
+ "condition": "A",
5
+ "step": 11000,
6
+ "params": 82470528,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.18743894730793542,
10
+ "cps": 25.484375,
11
+ "bpb": 1.624982773953399,
12
+ "bytes_per_chunk": 5.10015223957548
13
+ },
14
+ "de": {
15
+ "cpc": 0.18136658062630887,
16
+ "cps": 29.09375,
17
+ "bpb": 1.76161577754148,
18
+ "bytes_per_chunk": 5.566783285464767
19
+ },
20
+ "vi": {
21
+ "cpc": 0.22755627881746676,
22
+ "cps": 32.7734375,
23
+ "bpb": 1.4094545136020074,
24
+ "bytes_per_chunk": 5.748173161786875
25
+ },
26
+ "tr": {
27
+ "cpc": 0.18008064961608572,
28
+ "cps": 25.46875,
29
+ "bpb": 1.6799797300116097,
30
+ "bytes_per_chunk": 5.93300742350172
31
+ },
32
+ "fi": {
33
+ "cpc": 0.17085747700813353,
34
+ "cps": 25.109375,
35
+ "bpb": 1.868181290755361,
36
+ "bytes_per_chunk": 5.955483361636631
37
+ },
38
+ "ru": {
39
+ "cpc": 0.398311345646438,
40
+ "cps": 58.96875,
41
+ "bpb": 1.0388261127015617,
42
+ "bytes_per_chunk": 4.674444211643136
43
+ },
44
+ "el": {
45
+ "cpc": 0.3343078497295717,
46
+ "cps": 53.6015625,
47
+ "bpb": 1.0367840659476264,
48
+ "bytes_per_chunk": 5.490760377230051
49
+ },
50
+ "hi": {
51
+ "cpc": 0.41997028231797917,
52
+ "cps": 55.203125,
53
+ "bpb": 0.9135630702680455,
54
+ "bytes_per_chunk": 5.897270636139185
55
+ },
56
+ "ta": {
57
+ "cpc": 0.44118748328430063,
58
+ "cps": 64.4375,
59
+ "bpb": 0.7907793082315361,
60
+ "bytes_per_chunk": 5.881030634106945
61
+ },
62
+ "th": {
63
+ "cpc": 0.4851103512589369,
64
+ "cps": 60.9609375,
65
+ "bpb": 0.8860151257289858,
66
+ "bytes_per_chunk": 5.926769989034716
67
+ },
68
+ "zh": {
69
+ "cpc": 0.47409414155096513,
70
+ "cps": 21.875,
71
+ "bpb": 2.5161213810896603,
72
+ "bytes_per_chunk": 5.786491990000607
73
+ },
74
+ "my": {
75
+ "cpc": 0.45925691365843363,
76
+ "cps": 67.984375,
77
+ "bpb": 0.8805026195654585,
78
+ "bytes_per_chunk": 6.253696182450365
79
+ }
80
+ },
81
+ "gini_cpc": 0.20958090744294489,
82
+ "gini_cps": 0.2201348169698425,
83
+ "bpic": 5.755304950287179,
84
+ "seed": 0,
85
+ "mean_boundary_entropy": 0.14483880996704102,
86
+ "dechunk": "ema",
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": 4400,
99
+ "phase": 2,
100
+ "objective": "global",
101
+ "router_trainable": true,
102
+ "encoder_trainable": true,
103
+ "frozen_router_params": 0
104
+ },
105
+ {
106
+ "step": 7700,
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