Hugging Face's logo

AdaptiveChunking
/
pilot_E_s0

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
896f21f
·
verified ·
1 Parent(s): 3a84574

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_E_s0",
4
+ "condition": "E",
5
+ "step": 6000,
6
+ "params": 22516992,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.20898695627190714,
10
+ "cps": 28.4140625,
11
+ "bpb": 2.027534847106518,
12
+ "bytes_per_chunk": 4.792351075401505
13
+ },
14
+ "de": {
15
+ "cpc": 0.1766911800516242,
16
+ "cps": 28.34375,
17
+ "bpb": 2.209753539885548,
18
+ "bytes_per_chunk": 5.571071842225504
19
+ },
20
+ "vi": {
21
+ "cpc": 0.19528071602929212,
22
+ "cps": 28.125,
23
+ "bpb": 1.9175251448329642,
24
+ "bytes_per_chunk": 6.9727493981992525
25
+ },
26
+ "tr": {
27
+ "cpc": 0.19184665525051098,
28
+ "cps": 27.1328125,
29
+ "bpb": 2.1790023647295897,
30
+ "bytes_per_chunk": 5.453437210705333
31
+ },
32
+ "fi": {
33
+ "cpc": 0.18303120514592525,
34
+ "cps": 26.8984375,
35
+ "bpb": 2.3653980538947033,
36
+ "bytes_per_chunk": 5.287773194420631
37
+ },
38
+ "ru": {
39
+ "cpc": 0.19693931398416886,
40
+ "cps": 29.15625,
41
+ "bpb": 1.3694184250825936,
42
+ "bytes_per_chunk": 8.874054264484352
43
+ },
44
+ "el": {
45
+ "cpc": 0.16678848121619647,
46
+ "cps": 26.7421875,
47
+ "bpb": 1.4537136629856469,
48
+ "bytes_per_chunk": 10.662544996034248
49
+ },
50
+ "hi": {
51
+ "cpc": 0.19292719167904904,
52
+ "cps": 25.359375,
53
+ "bpb": 1.1981697263057156,
54
+ "bytes_per_chunk": 12.130960919965755
55
+ },
56
+ "ta": {
57
+ "cpc": 0.08558438085049479,
58
+ "cps": 12.5,
59
+ "bpb": 0.9989697245544682,
60
+ "bytes_per_chunk": 32.08224207563334
61
+ },
62
+ "th": {
63
+ "cpc": 0.20285980727385763,
64
+ "cps": 25.4921875,
65
+ "bpb": 1.1227061361153114,
66
+ "bytes_per_chunk": 13.105889411058895
67
+ },
68
+ "zh": {
69
+ "cpc": 0.6197087707416187,
70
+ "cps": 28.59375,
71
+ "bpb": 3.112132933082202,
72
+ "bytes_per_chunk": 4.639224153187272
73
+ },
74
+ "my": {
75
+ "cpc": 0.09927169094363521,
76
+ "cps": 14.6953125,
77
+ "bpb": 1.1179256096116879,
78
+ "bytes_per_chunk": 25.561308566135246
79
+ }
80
+ },
81
+ "gini_cpc": 0.2399013066286535,
82
+ "gini_cps": 0.09629140102627898,
83
+ "bpic": 9.946120354532733,
84
+ "seed": 0,
85
+ "mean_boundary_entropy": 0.3112594485282898,
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": "parity_b",
101
+ "router_trainable": true,
102
+ "encoder_trainable": true,
103
+ "frozen_router_params": 0
104
+ },
105
+ {
106
+ "step": 4200,
107
+ "phase": 3,
108
+ "objective": "none",
109
+ "router_trainable": false,
110
+ "encoder_trainable": true,
111
+ "frozen_router_params": 131584
112
+ }
113
+ ]
114
+ }
log.jsonl ADDED
The diff for this file is too large to render. See raw diff