ruby-openai 5.0.0 → 5.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4f487bd64d6f7a7c4f0ccc3198d9bec592b199792763b6885b12b21f267ff80a
-  data.tar.gz: 566af61bb906edbb2315ee343aae62f82e6837aa318dfb505d67d73497fcb27a
+  metadata.gz: 6ea6e5d9149ffa94f53c0952491e827a5a082830cb5a4ecbdafe4e4d2523f54e
+  data.tar.gz: 1e8072b9fce1c48612b0120e1df4d6f45e422daec13a41e4b985f60d6cc07f6a
 SHA512:
-  metadata.gz: 7e500ce6b1cff92bdb78b4cf455aac251cb229e43e52c4c863a98f44a351ac1ef4abd9692dde13f822f458b7373ee063e34dc0670b72dc67e02827533e316b13
-  data.tar.gz: 5d90d4ae80e14da163655a29d598c09834cfd0cad728bebe18f09b015aff7fa0881e736408360b228aa4921497de3f41b6122ee0a5e7a6011001b32403f70b41
+  metadata.gz: 00b71588418d3c33fb2511147e9a500755cf864c1d4cd7c420599b7b7af7d10bd4bb0b1490ce5399efbf38ce2527461ad40c21f20685b6aba40db275d7c9c633
+  data.tar.gz: e2574855121d6ed5126aa809b32feab815b1bd8f668c9eff1d3f6c9e9a25ed83cbb45c19e1900695b814d79b708a9cedfaf911ea5112ba2ff6eadcc76332f980

data/CHANGELOG.md

@@ -5,6 +5,12 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [5.1.0] - 2023-08-20
+
+### Added
+
+- Added rough_token_count to estimate tokens in a string according to OpenAI's "rules of thumb". Thank you to [@jamiemccarthy](https://github.com/jamiemccarthy) for the idea and implementation!
+
 ## [5.0.0] - 2023-08-14
 
 ### Added

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    ruby-openai (5.0.0)
+    ruby-openai (5.1.0)
       faraday (>= 1)
       faraday-multipart (>= 1)
 

data/README.md

@@ -24,13 +24,17 @@ gem "ruby-openai"
 
 And then execute:
 
+```bash
 $ bundle install
+```
 
 ### Gem install
 
 Or install with:
 
+```bash
 $ gem install ruby-openai
+```
 
 and require with:
 
@@ -86,7 +90,8 @@ client = OpenAI::Client.new(
     extra_headers: {
       "X-Proxy-TTL" => "43200", # For https://github.com/6/openai-caching-proxy-worker#specifying-a-cache-ttl
       "X-Proxy-Refresh": "true", # For https://github.com/6/openai-caching-proxy-worker#refreshing-the-cache
-      "Helicone-Auth": "Bearer HELICONE_API_KEY" # For https://docs.helicone.ai/getting-started/integration-method/openai-proxy
+      "Helicone-Auth": "Bearer HELICONE_API_KEY", # For https://docs.helicone.ai/getting-started/integration-method/openai-proxy
+      "helicone-stream-force-format" => "true", # Use this with Helicone otherwise streaming drops chunks # https://github.com/alexrudall/ruby-openai/issues/251
     }
 )
 ```
@@ -122,6 +127,18 @@ To use the [Azure OpenAI Service](https://learn.microsoft.com/en-us/azure/cognit
 
 where `AZURE_OPENAI_URI` is e.g. `https://custom-domain.openai.azure.com/openai/deployments/gpt-35-turbo`
 
+### Counting Tokens
+
+OpenAI parses prompt text into [tokens](https://help.openai.com/en/articles/4936856-what-are-tokens-and-how-to-count-them), which are words or portions of words. (These tokens are unrelated to your API access_token.) Counting tokens can help you estimate your [costs](https://openai.com/pricing). It can also help you ensure your prompt text size is within the max-token limits of your model's context window, and choose an appropriate [`max_tokens`](https://platform.openai.com/docs/api-reference/chat/create#chat/create-max_tokens) completion parameter so your response will fit as well.
+
+To estimate the token-count of your text:
+
+```ruby
+OpenAI.rough_token_count("Your text")
+```
+
+If you need a more accurate count, try [tiktoken_ruby](https://github.com/IAPark/tiktoken_ruby).
+
 ### Models
 
 There are different models that can be used to generate text. For a full list and to retrieve information about a single model:
@@ -180,7 +197,7 @@ client.chat(
 # => "Anna is a young woman in her mid-twenties, with wavy chestnut hair that falls to her shoulders..."
 ```
 
-Note: the API docs state that token usage is included in the streamed chat chunk objects, but this doesn't currently appear to be the case. If you need to work out how many tokens are being used while streaming, try [tiktoken_ruby](https://github.com/IAPark/tiktoken_ruby).
+Note: the API docs state that token usage is included in the streamed chat chunk objects, but this doesn't currently appear to be the case. To count tokens while streaming, try `OpenAI.rough_token_count` or [tiktoken_ruby](https://github.com/IAPark/tiktoken_ruby).
 
 ### Functions
 

data/lib/openai/version.rb

@@ -1,3 +1,3 @@
 module OpenAI
-  VERSION = "5.0.0".freeze
+  VERSION = "5.1.0".freeze
 end

data/lib/openai.rb

@@ -52,4 +52,16 @@ module OpenAI
   def self.configure
     yield(configuration)
   end
+
+  # Estimate the number of tokens in a string, using the rules of thumb from OpenAI:
+  # https://help.openai.com/en/articles/4936856-what-are-tokens-and-how-to-count-them
+  def self.rough_token_count(content = "")
+    raise ArgumentError, "rough_token_count requires a string" unless content.is_a? String
+    return 0 if content.empty?
+
+    count_by_chars = content.size / 4.0
+    count_by_words = content.split.size * 4.0 / 3
+    estimate = ((count_by_chars + count_by_words) / 2.0).round
+    [1, estimate].max
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: ruby-openai
 version: !ruby/object:Gem::Version
-  version: 5.0.0
+  version: 5.1.0
 platform: ruby
 authors:
 - Alex
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2023-08-14 00:00:00.000000000 Z
+date: 2023-08-20 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: faraday