Hugging Face's logo

AdaptiveChunking
/
exp32_local_s2

PyTorch
Safetensors
dynamic-chunking
tokenization
hnet
byte-level
beetle-hnet
legacy-trainer
Model card Files
xet
 Community
suchirsalhan commited on
Commit
043c4ee
·
verified ·
1 Parent(s): fcefb5c

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": "exp32_local_s2",
4
+ "condition": "A",
5
+ "step": 11000,
6
+ "params": 82470528,
7
+ "per_lang": {
8
+ "en": {
9
+ "cpc": 0.17048784692294433,
10
+ "cps": 23.1796875,
11
+ "bpb": 1.6931716484971258,
12
+ "bytes_per_chunk": 5.338519578244246
13
+ },
14
+ "de": {
15
+ "cpc": 0.1685579311352457,
16
+ "cps": 27.0390625,
17
+ "bpb": 1.8315910466286345,
18
+ "bytes_per_chunk": 5.687531188301404
19
+ },
20
+ "vi": {
21
+ "cpc": 0.21106590724165988,
22
+ "cps": 30.3984375,
23
+ "bpb": 1.5223879424380602,
24
+ "bytes_per_chunk": 6.167332270719499
25
+ },
26
+ "tr": {
27
+ "cpc": 0.18599127216483455,
28
+ "cps": 26.3046875,
29
+ "bpb": 1.7367852878730408,
30
+ "bytes_per_chunk": 5.575693122480884
31
+ },
32
+ "fi": {
33
+ "cpc": 0.187230875551539,
34
+ "cps": 27.515625,
35
+ "bpb": 1.899413071429155,
36
+ "bytes_per_chunk": 5.327128537825714
37
+ },
38
+ "ru": {
39
+ "cpc": 0.2920316622691293,
40
+ "cps": 43.234375,
41
+ "bpb": 1.101538844663503,
42
+ "bytes_per_chunk": 5.922518624787489
43
+ },
44
+ "el": {
45
+ "cpc": 0.2878234176290016,
46
+ "cps": 46.1484375,
47
+ "bpb": 1.079825282567976,
48
+ "bytes_per_chunk": 6.0586114449477675
49
+ },
50
+ "hi": {
51
+ "cpc": 0.4279346210995542,
52
+ "cps": 56.25,
53
+ "bpb": 0.9105429205574486,
54
+ "bytes_per_chunk": 6.009307016940605
55
+ },
56
+ "ta": {
57
+ "cpc": 0.46113934207007223,
58
+ "cps": 67.3515625,
59
+ "bpb": 0.7750441210539766,
60
+ "bytes_per_chunk": 5.947342748567864
61
+ },
62
+ "th": {
63
+ "cpc": 0.49400062169723347,
64
+ "cps": 62.078125,
65
+ "bpb": 0.8569683953833194,
66
+ "bytes_per_chunk": 5.815310044200917
67
+ },
68
+ "zh": {
69
+ "cpc": 0.7167287504232983,
70
+ "cps": 33.0703125,
71
+ "bpb": 2.4890658228800207,
72
+ "bytes_per_chunk": 3.7616266555697457
73
+ },
74
+ "my": {
75
+ "cpc": 0.4506016466117796,
76
+ "cps": 66.703125,
77
+ "bpb": 0.8467113339624879,
78
+ "bytes_per_chunk": 6.260865411598927
79
+ }
80
+ },
81
+ "gini_cpc": 0.2639468220288741,
82
+ "gini_cps": 0.2120399772960868,
83
+ "bpic": 5.887385521653091,
84
+ "seed": 2,
85
+ "mean_boundary_entropy": 0.14645597338676453,
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