Class: Rubino::Context::TokenBudget

Inherits:
Object
  • Object
show all
Defined in:
lib/rubino/context/token_budget.rb

Overview

Manages token budget calculations and determines when compaction is needed.

Constant Summary collapse

CHARS_PER_TOKEN =

Rough approximation

4
DEFAULT_CONTEXT_WINDOW =

Fallback when the user hasn’t pinned ‘model.context_length` in config. Generous-but-safe; truncation kicks in via `needs_compaction?` long before the real provider limit would be hit.

128_000

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(model_id:, config:) ⇒ TokenBudget

Returns a new instance of TokenBudget.



13
14
15
16
17
 
# File 'lib/rubino/context/token_budget.rb', line 13

def initialize(model_id:, config:)
  @model_id = model_id
  @config = config
  @context_window = determine_context_window
end

Instance Attribute Details

#context_window ⇒ Object (readonly)

Returns the value of attribute context_window.



19
20
21
 
# File 'lib/rubino/context/token_budget.rb', line 19

def context_window
  @context_window
end

Instance Method Details

#available_tokens ⇒ Object

Returns the max tokens available for conversation



22
23
24
25
 
# File 'lib/rubino/context/token_budget.rb', line 22

def available_tokens
  override = @config.dig("context", "max_tokens")
  override || @context_window
end

#compaction_target ⇒ Object

Returns the target token count after compaction



52
53
54
 
# File 'lib/rubino/context/token_budget.rb', line 52

def compaction_target
  (available_tokens * @config.compression_target_ratio).to_i
end

#critical?(messages) ⇒ Boolean

Returns true if critically close to context limit

Returns:

  • (Boolean) 


43
44
45
46
47
48
49
 
# File 'lib/rubino/context/token_budget.rb', line 43

def critical?(messages)
  return false unless @config.compression_enabled?

  estimated = estimate_tokens(messages)
  gateway = (available_tokens * @config.compression_gateway_threshold).to_i
  estimated > gateway
end

#estimate_tokens(messages) ⇒ Object

Estimates token count for a set of messages



28
29
30
31
 
# File 'lib/rubino/context/token_budget.rb', line 28

def estimate_tokens(messages)
  total_chars = messages.sum { |m| (m[:content] || "").length }
  (total_chars.to_f / CHARS_PER_TOKEN).ceil
end

#needs_compaction?(messages) ⇒ Boolean

Returns true if the messages exceed the compaction threshold

Returns:

  • (Boolean) 


34
35
36
37
38
39
40
 
# File 'lib/rubino/context/token_budget.rb', line 34

def needs_compaction?(messages)
  return false unless @config.compression_enabled?

  estimated = estimate_tokens(messages)
  threshold = (available_tokens * @config.compression_threshold).to_i
  estimated > threshold
end