Squish42 commited on
Commit
8267d8a
β€’
1 Parent(s): 9e890c9

Adjust directory structure for multiple splits and configurations

Browse files
.gitattributes CHANGED
@@ -56,4 +56,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
56
  *.webp filter=lfs diff=lfs merge=lfs -text
57
  *.json filter=lfs diff=lfs merge=lfs -text
58
  *.gz filter=lfs diff=lfs merge=lfs -text
59
- stats/stats.json filter= diff= merge= text
 
 
56
  *.webp filter=lfs diff=lfs merge=lfs -text
57
  *.json filter=lfs diff=lfs merge=lfs -text
58
  *.gz filter=lfs diff=lfs merge=lfs -text
59
+ # Stats
60
+ **/stats/stats.json filter= diff= merge= text
.gitignore CHANGED
@@ -2,6 +2,6 @@
2
  out
3
  original
4
  # Generated data
5
- pack/gen
6
  *.pyc
7
  scripts
 
2
  out
3
  original
4
  # Generated data
5
+ resources/gen
6
  *.pyc
7
  scripts
bluemoon.json β†’ data/full/train/bluemoon.json RENAMED
File without changes
{stats β†’ data/full/train/stats}/conversations.json.gz RENAMED
File without changes
{stats β†’ data/full/train/stats}/stats.json RENAMED
File without changes
{stats β†’ data/full/train/stats}/vocabulary.json.gz RENAMED
File without changes
dataset.yaml CHANGED
@@ -27,7 +27,7 @@ _output_steps:
27
  type: FastChatOutput
28
  sync: main
29
  config:
30
- path: "out/bluemoon.json"
31
  indent: 2
32
  #includeMessageId: true
33
 
@@ -153,10 +153,10 @@ _steps:
153
  - step: &spelling
154
  type: ExactReplace
155
  description: "Spelling mistakes and typos."
156
- pack: [ 'pack/spelling.csv' ]
157
  config:
158
  caseSensitive: false
159
- packedReplacements: [ 'pack/spelling.csv' ]
160
 
161
  - step: &empty
162
  type: FullMatch
@@ -182,19 +182,19 @@ _steps:
182
  - step: &edge_cases
183
  type: ExactReplace
184
  description: "Remove a few edge cases."
185
- pack: [ 'pack/edge_cases.csv' ]
186
  config:
187
- packedReplacements: [ 'pack/edge_cases.csv' ]
188
 
189
  - step: &stats_output
190
  type: StatsOutput
191
  description: "Output statistics"
192
  sync: main
193
  config:
194
- path: 'stats/stats.json'
195
- vocabPath: 'stats/vocabulary.json.gz'
196
  gZipVocabulary: true
197
- conversationsPath: 'stats/conversations.json.gz'
198
  gZipConversations: true
199
  _passes:
200
  - pass: &clean
@@ -211,7 +211,7 @@ _passes:
211
  *bluemoon_fastchat, *stats_output
212
  ]
213
  artifacts:
214
- - file: 'out/bluemoon.json'
215
  sha512: 70ddecc11f8b280a5d641ce502e91a2392aa9dc12e0704617edb1599e4c235b6e8015987f109a09c83746f717cec715fcad04587cc066897583eaf2bd439616b
216
 
217
  passes: [ *clean ]
 
27
  type: FastChatOutput
28
  sync: main
29
  config:
30
+ path: "out/data/full/train/bluemoon.json"
31
  indent: 2
32
  #includeMessageId: true
33
 
 
153
  - step: &spelling
154
  type: ExactReplace
155
  description: "Spelling mistakes and typos."
156
+ pack: [ 'resources/spelling.csv' ]
157
  config:
158
  caseSensitive: false
159
+ packedReplacements: [ 'resources/spelling.csv' ]
160
 
161
  - step: &empty
162
  type: FullMatch
 
182
  - step: &edge_cases
183
  type: ExactReplace
184
  description: "Remove a few edge cases."
185
+ pack: [ 'resources/edge_cases.csv' ]
186
  config:
187
+ packedReplacements: [ 'resources/edge_cases.csv' ]
188
 
189
  - step: &stats_output
190
  type: StatsOutput
191
  description: "Output statistics"
192
  sync: main
193
  config:
194
+ path: 'out/data/full/train/stats/stats.json'
195
+ vocabPath: 'out/data/full/train/stats/vocabulary.json.gz'
196
  gZipVocabulary: true
197
+ conversationsPath: 'out/data/full/train/stats/conversations.json.gz'
198
  gZipConversations: true
199
  _passes:
200
  - pass: &clean
 
211
  *bluemoon_fastchat, *stats_output
212
  ]
213
  artifacts:
214
+ - file: 'out/data/full/train/bluemoon.json'
215
  sha512: 70ddecc11f8b280a5d641ce502e91a2392aa9dc12e0704617edb1599e4c235b6e8015987f109a09c83746f717cec715fcad04587cc066897583eaf2bd439616b
216
 
217
  passes: [ *clean ]
debug.yaml CHANGED
@@ -56,10 +56,10 @@ _debug_steps:
56
 
57
  - step: &strip_usernames
58
  type: ExactReplace
59
- pack: [ 'pack/gen/usernames.csv' ]
60
  config:
61
  caseSensitive: true
62
- packedReplacements: [ 'pack/gen/usernames.csv' ]
63
 
64
 
65
  passes:
@@ -79,19 +79,19 @@ passes:
79
  - type: FileConcatenate
80
  sync: main
81
  config:
82
- path: 'pack/gen/usernames_all.csv'
83
  files: [ 'out/usernames_*.csv' ]
84
  ignoreDuplicateLines: true
85
  eol: "\r\n"
86
  delete: true
87
  artifacts:
88
- - file: 'pack/gen/usernames_all.csv'
89
  sha512: ffcc8aa980f4d8b0cace9092280d07215337cc316cb3fa45742a5f5a735d21a31f581d27884a414df21ac99616d14c7ebec1ab68a5766c10faa4b1564ceb7c20
90
 
91
  # Count the occurrences of usernames in messages in the source data
92
  - pass: &count_usernames
93
  required:
94
- - path: "pack/gen/usernames_all.csv"
95
  sha512: ffcc8aa980f4d8b0cace9092280d07215337cc316cb3fa45742a5f5a735d21a31f581d27884a414df21ac99616d14c7ebec1ab68a5766c10faa4b1564ceb7c20
96
  - path: "data/rentry/bluemoon/bluemoon-fanbased-roleplays-1x1.csv"
97
  sha512: f31d6bd278bc4211736d6aace3917cd0c1d0143bec9bf9d07054f5f9b32060e17399b6ea0935774b2271ac45a88309a60ef8eec8c4ac5283d1b353a255529cc5
@@ -99,10 +99,10 @@ passes:
99
  - *bluemoon
100
  - type: StatsCountOccurrences
101
  description: "Count occurrences of usernames."
102
- pack: [ 'pack/gen/usernames_all.csv' ]
103
  config:
104
  path: 'out/usernames_counted_%worker%.csv'
105
- packedPatterns: [ 'pack/gen/usernames_all.csv' ]
106
  countCol: 2
107
  regex: false
108
  min: 1
@@ -110,9 +110,9 @@ passes:
110
  description: "Merge username count csv files"
111
  sync: main
112
  config:
113
- path: 'pack/gen/usernames.csv'
114
  files: [ 'out/usernames_counted_*.csv' ]
115
  cols: [ 2 ]
116
  min: 1
117
  artifacts:
118
- - file: 'pack/gen/usernames.csv'
 
56
 
57
  - step: &strip_usernames
58
  type: ExactReplace
59
+ pack: [ 'resources/gen/usernames.csv' ]
60
  config:
61
  caseSensitive: true
62
+ packedReplacements: [ 'resources/gen/usernames.csv' ]
63
 
64
 
65
  passes:
 
79
  - type: FileConcatenate
80
  sync: main
81
  config:
82
+ path: 'resources/gen/usernames_all.csv'
83
  files: [ 'out/usernames_*.csv' ]
84
  ignoreDuplicateLines: true
85
  eol: "\r\n"
86
  delete: true
87
  artifacts:
88
+ - file: 'resources/gen/usernames_all.csv'
89
  sha512: ffcc8aa980f4d8b0cace9092280d07215337cc316cb3fa45742a5f5a735d21a31f581d27884a414df21ac99616d14c7ebec1ab68a5766c10faa4b1564ceb7c20
90
 
91
  # Count the occurrences of usernames in messages in the source data
92
  - pass: &count_usernames
93
  required:
94
+ - path: "resources/gen/usernames_all.csv"
95
  sha512: ffcc8aa980f4d8b0cace9092280d07215337cc316cb3fa45742a5f5a735d21a31f581d27884a414df21ac99616d14c7ebec1ab68a5766c10faa4b1564ceb7c20
96
  - path: "data/rentry/bluemoon/bluemoon-fanbased-roleplays-1x1.csv"
97
  sha512: f31d6bd278bc4211736d6aace3917cd0c1d0143bec9bf9d07054f5f9b32060e17399b6ea0935774b2271ac45a88309a60ef8eec8c4ac5283d1b353a255529cc5
 
99
  - *bluemoon
100
  - type: StatsCountOccurrences
101
  description: "Count occurrences of usernames."
102
+ pack: [ 'resources/gen/usernames_all.csv' ]
103
  config:
104
  path: 'out/usernames_counted_%worker%.csv'
105
+ packedPatterns: [ 'resources/gen/usernames_all.csv' ]
106
  countCol: 2
107
  regex: false
108
  min: 1
 
110
  description: "Merge username count csv files"
111
  sync: main
112
  config:
113
+ path: 'resources/gen/usernames.csv'
114
  files: [ 'out/usernames_counted_*.csv' ]
115
  cols: [ 2 ]
116
  min: 1
117
  artifacts:
118
+ - file: 'resources/gen/usernames.csv'
{pack β†’ resources}/edge_cases.csv RENAMED
File without changes
{pack β†’ resources}/spelling.csv RENAMED
File without changes