Add 'pruned' configuration.

README.md

@@ -24,4 +24,5 @@ dataset.
 
 This dataset has been cleaned and formatted for use with fastchat.
 
-![Plot](assets/full-train.png)
+![Plot](assets/full-train.png)
+![Plot](assets/pruned-train.png)

data/pruned/bluemoon.test.raw

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9625715300aa954262f8721458ecdb890a2468a4fb7269d85169297afe28d89d
+size 824676

data/pruned/bluemoon.train.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d839732408bca8f0210d718b6ccd33b3750cc8cc6da64116615acfa28e3208b
+size 59947031

scripts/generate_plot.py

@@ -4,7 +4,8 @@ import matplotlib.pyplot as plt
 
 import dsbuild
 
-stats_path: list[str] = ['../stats/full/test/', '../stats/full/train']
+stats_path: list[str] = ['../stats/full/train/', '../stats/full/test',
+                         '../stats/pruned/train', '../stats/pruned/test']
 
 
 def __plot_mean_word_counts(path: list[str], show: bool = False, save: bool = False):
@@ -24,7 +25,7 @@ def __plot_mean_word_counts(path: list[str], show: bool = False, save: bool = Fa
 
 
 def main():
-    __plot_mean_word_counts(stats_path, show=True, save=True)
+    __plot_mean_word_counts(stats_path, show=False, save=True)
 
 
 if __name__ == '__main__':

split.yaml

@@ -0,0 +1,74 @@
+name: Bluemoon-Split
+description: "Splits"
+build:
+  conversationBatch: 100
+  cleanDirectory: [ 'out/data/pruned', 'out/stats/pruned' ]
+  verifyRequirements: false
+  verifyArtifacts: false
+  sendPackedFiles: true
+  gzipRemotePackedFiles: true
+  gzipLocalPackedFiles: false
+
+_input_steps:
+  - step: &bluemoon_cleaned
+      type: FastChatInput
+      sync: main
+      config:
+        path: "out/data/full/bluemoon.train.json"
+        preserveIds: true
+
+_passes:
+  - pass: &cfg--pruned
+      required:
+        - path: "out/data/full/bluemoon.train.json"
+      steps:
+        # Pre-cleaned source
+        - *bluemoon_cleaned
+        # Filter
+        - type: StatsConversationPrune
+          description: "Prune conversations for deviation or maximum message length"
+          pack: [ 'out/stats/full/train/stats.json', 'out/stats/full/train/conversations.json.gz' ]
+          config:
+            packedStatsPath: 'out/stats/full/train'
+            wordsStdDevRatioMin: 0.5
+            #wordsStdDevRatioMax: 10.0
+            #wordsClampMin: 120
+            wordsClampMax: 2048
+            messagesCountMin: 4
+          # Train split
+        - type: FastChatOutput
+          sync: main
+          config:
+            path: "out/data/pruned/bluemoon.train.json"
+            indent: 2
+        - type: StatsOutput
+          description: "Output statistics"
+          sync: main
+          config:
+            path: 'out/stats/pruned/train/stats.json'
+            vocabPath: 'out/stats/pruned/train/vocabulary.json.gz'
+            gZipVocabulary: true
+            conversationsPath: 'out/stats/pruned/train/conversations.json.gz'
+            gZipConversations: true
+        # Test split
+        - type: RawSampleOutput
+          sync: main
+          config:
+            ratio: 0.01
+            forwardFilter: true
+            path: "out/data/pruned/bluemoon.test.raw"
+        - type: StatsOutput
+          description: "Output statistics"
+          sync: main
+          config:
+            path: 'out/stats/pruned/test/stats.json'
+            vocabPath: 'out/stats/pruned/test/vocabulary.json.gz'
+            gZipVocabulary: true
+            conversationsPath: 'out/stats/pruned/test/conversations.json.gz'
+            gZipConversations: true
+      artifacts:
+        - file: "out/data/pruned/bluemoon.train.json"
+        - file: "out/data/pruned/bluemoon.test.raw"
+# Use the above pass as a template to output for new configurations and splits.
+
+passes: [ *cfg--pruned ]

stats/pruned/test/conversations.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a402f828e9550aa8cd3627d96905dcc00f21f1ae1c063dd12f46fe5af86c65a1
+size 4490

stats/pruned/test/stats.json

@@ -0,0 +1,20 @@
+{
+ "conversationsTotal": 10,
+ "messagesTotal": 229,
+ "tokensTotal": 168636,
+ "tokensMean": 736.4017467248908,
+ "tokensStdDev": 269.5242698707312,
+ "wordsTotal": 148109,
+ "wordsMean": 646.764192139738,
+ "wordsStdDev": 229.4064989806647,
+ "messagesLenMean": 3600.2052401746723,
+ "messagesLenStdDev": 1289.6597353510115,
+ "messagesCountMean": 22.9,
+ "wordsMin": 189,
+ "wordsMax": 1589,
+ "tokensMin": 204,
+ "tokensMax": 1858,
+ "messagesLenTotal": 824447,
+ "messagesLenMin": 1045,
+ "messagesLenMax": 8852
+}

stats/pruned/test/vocabulary.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e2f062298bc946046cd8b72f1f49c740076d0e9492d42d7bafc09be60a1b3606
+size 72150

stats/pruned/train/conversations.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0045429f76d68affa5caa3cb0c864051ebb253e0ebdd90a595ad1e34f7fefe8f
+size 407582

stats/pruned/train/stats.json

@@ -0,0 +1,20 @@
+{
+ "conversationsTotal": 1041,
+ "messagesTotal": 22770,
+ "tokensTotal": 11966704,
+ "tokensMean": 525.5469477382521,
+ "tokensStdDev": 267.0269096949294,
+ "wordsTotal": 10633309,
+ "wordsMean": 466.9876592007027,
+ "wordsStdDev": 236.8723244513958,
+ "messagesLenMean": 2565.7306543697846,
+ "messagesLenStdDev": 1319.9514485339039,
+ "messagesCountMean": 21.87319884726225,
+ "wordsMin": 142,
+ "wordsMax": 2009,
+ "tokensMin": 149,
+ "tokensMax": 2345,
+ "messagesLenTotal": 58421687,
+ "messagesLenMin": 695,
+ "messagesLenMax": 11164
+}

stats/pruned/train/vocabulary.json.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1b7d2f7bcaee45af4a848f8ec295905aece772e577691207c7b5f620a9b93dc
+size 659540