FIX: Truncate OP for gists to help the model focus on the latest posts (

#883)
discourse · Oct 31, 2024 · e8eed71 · e8eed71
1 parent 32fb023
commit e8eed71
Show file tree

Hide file tree

Showing 5 changed files with 22 additions and 11 deletions.
diff --git a/lib/summarization/fold_content.rb b/lib/summarization/fold_content.rb
@@ -121,9 +121,9 @@ def fold(items, summary, cursor, user, &on_partial_blk)
         prompt =
           (
             if summary.blank?
-              strategy.first_summary_prompt(iteration_content)
+              strategy.first_summary_prompt(iteration_content, tokenizer)
             else
-              strategy.summary_extension_prompt(summary, iteration_content)
+              strategy.summary_extension_prompt(summary, iteration_content, tokenizer)
             end
           )
 

diff --git a/lib/summarization/strategies/base.rb b/lib/summarization/strategies/base.rb
@@ -34,12 +34,12 @@ def targets_data
         end
 
         # @returns { DiscourseAi::Completions::Prompt } - Prompt passed to the LLM when extending an existing summary.
-        def summary_extension_prompt(_summary, _texts_to_summarize)
+        def summary_extension_prompt(_summary, _texts_to_summarize, _tokenizer)
           raise NotImplementedError
         end
 
         # @returns { DiscourseAi::Completions::Prompt } - Prompt passed to the LLM for summarizing a single chunk of content.
-        def first_summary_prompt(_input)
+        def first_summary_prompt(_input, _tokenizer)
           raise NotImplementedError
         end
 

diff --git a/lib/summarization/strategies/chat_messages.rb b/lib/summarization/strategies/chat_messages.rb
@@ -23,7 +23,7 @@ def targets_data
             .map { { id: _1, poster: _2, text: _3 } }
         end
 
-        def summary_extension_prompt(summary, contents)
+        def summary_extension_prompt(summary, contents, _tokenizer)
           input =
             contents
               .map { |item| "(#{item[:id]} #{item[:poster]} said: #{item[:text]} " }
@@ -63,7 +63,7 @@ def summary_extension_prompt(summary, contents)
           prompt
         end
 
-        def first_summary_prompt(contents)
+        def first_summary_prompt(contents, _tokenizer)
           content_title = target.name
           input =
             contents.map { |item| "(#{item[:id]} #{item[:poster]} said: #{item[:text]} " }.join

diff --git a/lib/summarization/strategies/hot_topic_gists.rb b/lib/summarization/strategies/hot_topic_gists.rb
@@ -57,7 +57,7 @@ def targets_data
           end
         end
 
-        def summary_extension_prompt(summary, contents)
+        def summary_extension_prompt(summary, contents, _tokenizer)
           statements =
             contents
               .to_a
@@ -98,11 +98,22 @@ def summary_extension_prompt(summary, contents)
           prompt
         end
 
-        def first_summary_prompt(contents)
+        def first_summary_prompt(contents, tokenizer)
           content_title = target.title
           statements =
             contents.to_a.map { |item| "(#{item[:id]} #{item[:poster]} said: #{item[:text]} " }
 
+          op_statement = statements.shift.to_s
+          split_1, split_2 =
+            [op_statement[0, op_statement.size / 2], op_statement[(op_statement.size / 2)..-1]]
+
+          truncation_length = 500
+
+          op_statement = [
+            tokenizer.truncate(split_1, truncation_length),
+            tokenizer.truncate(split_2.reverse, truncation_length).reverse,
+          ].join(" ")
+
           prompt = DiscourseAi::Completions::Prompt.new(<<~TEXT.strip)
             You are an advanced summarization bot. Analyze a given conversation and produce a concise,
             single-sentence summary that conveys the main topic and current developments to someone with no prior context.
@@ -127,7 +138,7 @@ def first_summary_prompt(contents)
             
             The conversation began with the following statement:
         
-            #{statements.shift}\n
+            #{op_statement}\n
           TEXT
 
           if statements.present?

diff --git a/lib/summarization/strategies/topic_summary.rb b/lib/summarization/strategies/topic_summary.rb
@@ -27,7 +27,7 @@ def targets_data
           end
         end
 
-        def summary_extension_prompt(summary, contents)
+        def summary_extension_prompt(summary, contents, _tokenizer)
           resource_path = "#{Discourse.base_path}/t/-/#{target.id}"
           content_title = target.title
           input =
@@ -70,7 +70,7 @@ def summary_extension_prompt(summary, contents)
           prompt
         end
 
-        def first_summary_prompt(contents)
+        def first_summary_prompt(contents, _tokenizer)
           resource_path = "#{Discourse.base_path}/t/-/#{target.id}"
           content_title = target.title
           input =