Hugging Face's logo

AdaptiveChunking
/
spec_A_s2

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
1140868
·
verified ·
1 Parent(s): 043dba8

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_s2",
4
+ "condition": "A",
5
+ "step": 11000,
6
+ "params": 82470528,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.19272539217376314,
10
+ "cps": 26.203125,
11
+ "bpb": 1.6370897828816098,
12
+ "bytes_per_chunk": 5.164127062299926
13
+ },
14
+ "de": {
15
+ "cpc": 0.18857448984561437,
16
+ "cps": 30.25,
17
+ "bpb": 1.7686300899569196,
18
+ "bytes_per_chunk": 5.24277514449711
19
+ },
20
+ "vi": {
21
+ "cpc": 0.22196908055329537,
22
+ "cps": 31.96875,
23
+ "bpb": 1.4657194278770778,
24
+ "bytes_per_chunk": 5.878755599410205
25
+ },
26
+ "tr": {
27
+ "cpc": 0.19433243108876982,
28
+ "cps": 27.484375,
29
+ "bpb": 1.6620841612674058,
30
+ "bytes_per_chunk": 5.594046253567713
31
+ },
32
+ "fi": {
33
+ "cpc": 0.19525809366859817,
34
+ "cps": 28.6953125,
35
+ "bpb": 1.8530744099459817,
36
+ "bytes_per_chunk": 5.268776034208133
37
+ },
38
+ "ru": {
39
+ "cpc": 0.30854881266490763,
40
+ "cps": 45.6796875,
41
+ "bpb": 1.070778833898857,
42
+ "bytes_per_chunk": 5.86065124805777
43
+ },
44
+ "el": {
45
+ "cpc": 0.3208595234614822,
46
+ "cps": 51.4453125,
47
+ "bpb": 1.08748095053523,
48
+ "bytes_per_chunk": 5.744394348604955
49
+ },
50
+ "hi": {
51
+ "cpc": 0.44481426448737,
52
+ "cps": 58.46875,
53
+ "bpb": 0.8965628163319439,
54
+ "bytes_per_chunk": 5.600350363715992
55
+ },
56
+ "ta": {
57
+ "cpc": 0.4535437282695908,
58
+ "cps": 66.2421875,
59
+ "bpb": 0.7514792839599838,
60
+ "bytes_per_chunk": 5.935425440384006
61
+ },
62
+ "th": {
63
+ "cpc": 0.48436431457880014,
64
+ "cps": 60.8671875,
65
+ "bpb": 0.8530057592481629,
66
+ "bytes_per_chunk": 5.940031836489602
67
+ },
68
+ "zh": {
69
+ "cpc": 0.5414832373857095,
70
+ "cps": 24.984375,
71
+ "bpb": 2.475187059431537,
72
+ "bytes_per_chunk": 5.19207554083295
73
+ },
74
+ "my": {
75
+ "cpc": 0.5039582013932868,
76
+ "cps": 74.6015625,
77
+ "bpb": 0.876910052473754,
78
+ "bytes_per_chunk": 5.586476219905274
79
+ }
80
+ },
81
+ "gini_cpc": 0.21984466227883126,
82
+ "gini_cps": 0.2160448978381424,
83
+ "bpic": 5.690534088550161,
84
+ "seed": 2,
85
+ "mean_boundary_entropy": 0.19834211468696594,
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