Hugging Face's logo

AdaptiveChunking
/
pilot_A_s0

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
f498035
·
verified ·
1 Parent(s): 09991ee

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": "pilot_A_s0",
4
+ "condition": "A",
5
+ "step": 6000,
6
+ "params": 22516992,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.2076653450554502,
10
+ "cps": 28.234375,
11
+ "bpb": 2.0273781279159193,
12
+ "bytes_per_chunk": 4.917766459370984
13
+ },
14
+ "de": {
15
+ "cpc": 0.20089611844348124,
16
+ "cps": 32.2265625,
17
+ "bpb": 2.194651560581718,
18
+ "bytes_per_chunk": 4.990082424380866
19
+ },
20
+ "vi": {
21
+ "cpc": 0.2209926769731489,
22
+ "cps": 31.828125,
23
+ "bpb": 1.8993434507023028,
24
+ "bytes_per_chunk": 5.929249977381707
25
+ },
26
+ "tr": {
27
+ "cpc": 0.20388885820029828,
28
+ "cps": 28.8359375,
29
+ "bpb": 2.172621765386917,
30
+ "bytes_per_chunk": 5.211869377205627
31
+ },
32
+ "fi": {
33
+ "cpc": 0.19892616022540002,
34
+ "cps": 29.234375,
35
+ "bpb": 2.362985137303098,
36
+ "bytes_per_chunk": 4.885778452879069
37
+ },
38
+ "ru": {
39
+ "cpc": 0.3211081794195251,
40
+ "cps": 47.5390625,
41
+ "bpb": 1.3424836631873538,
42
+ "bytes_per_chunk": 5.437995270298303
43
+ },
44
+ "el": {
45
+ "cpc": 0.3130146664717634,
46
+ "cps": 50.1875,
47
+ "bpb": 1.3926614235177859,
48
+ "bytes_per_chunk": 5.5650401757756525
49
+ },
50
+ "hi": {
51
+ "cpc": 0.43910846953937593,
52
+ "cps": 57.71875,
53
+ "bpb": 1.1374732168277908,
54
+ "bytes_per_chunk": 5.7044250291048755
55
+ },
56
+ "ta": {
57
+ "cpc": 0.42326825354372827,
58
+ "cps": 61.8203125,
59
+ "bpb": 0.9469699101528437,
60
+ "bytes_per_chunk": 6.453172502923257
61
+ },
62
+ "th": {
63
+ "cpc": 0.5020826857320485,
64
+ "cps": 63.09375,
65
+ "bpb": 1.093262663291936,
66
+ "bytes_per_chunk": 5.682784335403592
67
+ },
68
+ "zh": {
69
+ "cpc": 0.482390789028107,
70
+ "cps": 22.2578125,
71
+ "bpb": 3.1054285549014486,
72
+ "bytes_per_chunk": 5.831448051875827
73
+ },
74
+ "my": {
75
+ "cpc": 0.513299556681444,
76
+ "cps": 75.984375,
77
+ "bpb": 1.0876047296737308,
78
+ "bytes_per_chunk": 6.01137405980554
79
+ }
80
+ },
81
+ "gini_cpc": 0.2046874384551851,
82
+ "gini_cps": 0.2124337217717518,
83
+ "bpic": 5.668261775001108,
84
+ "seed": 0,
85
+ "mean_boundary_entropy": 0.32596713304519653,
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": 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