kaba 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 17f06505d496fef06c186773df25fa37d7a2e465fac06682f51a300933e286d5
-  data.tar.gz: caefca3868f25c1cd15e6bf441ae3e595a2be4b5fb7574431c1a57284ca00327
+  metadata.gz: 0add8ecf9ac4613decb084b678c044809896db95136b59e24602ae47a426d448
+  data.tar.gz: 2dda3c5fa0152b4928c4a2fadc11b18ed429cc724e87c5b38c794083cb731c69
 SHA512:
-  metadata.gz: '09d533d49ac00fd23230ae8ea81f2c189d81b285738142e323b07bb5ea9cbac3cd91ae8f41daa56c37ea1c5297d99573e8dac2c411ec5add299644e8d888ea1e'
-  data.tar.gz: 15751ae8ef5e041b5ac8638626bac0e938fe989a5647af4c59c8a9cdc803e0796889f4353adf0a5fefe0af0430187f82ca21ccdd9dcd0a53c88a3edd8829c45d
+  metadata.gz: 622be2e55740919481236e525f2c2045f9f633eccfa36363a55d959c8d25cd859cfc349e6668df429571052d9501077f8068c35e1a72bac25ced650188c3407e
+  data.tar.gz: 56c1a740604cb48c6bfb14f5c93854597fa53922c2befd03e58cf192d29e2b2b1ac87d9480255ae6c6ef175d2652299b67b0cba8cb1e0deccf405adce9efa22b

data/README.md

@@ -30,4 +30,15 @@ alias kaba='docker run -it --rm -v "${PWD}:/workdir" ghcr.io/mjason/kaba:latest'
 ## 关联项目
 - [lisa_typechat_server](https://github.com/mjason/lisa_typechat_server)
 
-如果要修改服务地址你有两个方式，一个通过 `.env` 来处理，还有就是自己设置环境变量，变量名 `LISA_TYPECHAT_ENDPOINT`
+如果要修改服务地址你有两个方式，一个通过 `.env` 来处理，还有就是自己设置环境变量，变量名 `LISA_TYPECHAT_ENDPOINT`
+
+## changelog
+
+.env 需要更新
+```
+; LISA_TYPECHAT_ENDPOINT=https://lisa-typechat.listenai.com
+LISA_ACCESS_TOKEN=聆思平台的KEY
+
+JUDGE_ACCCESS_TOKEN=可以和LISA_ACCESS_TOKEN
+JUDGE_LLM_URI_BASE=如果需要其他提供商可以填，默认不填
+```

data/kaba.gemspec

@@ -12,6 +12,7 @@ Gem::Specification.new do |spec|
   spec.description = "用来做数据集的工具"
   spec.homepage = "https://github.com/mjason/kaba.git"
   spec.required_ruby_version = ">= 3.3.0"
+  spec.license = "MIT"
 
   # spec.metadata["allowed_push_host"] = "TODO: Set to your gem server 'https://example.com'"
 

data/lib/kaba/application.rb

@@ -10,11 +10,23 @@ class Application
 
     def llm_client
       @llm_client ||= OpenAI::Client.new(
+        log_errors: true,
         access_token: env!("LISA_ACCESS_TOKEN"),
         request_timeout: ENV.fetch("LISA_LLM_REQUEST_TIMEOUT", 120).to_i,
         uri_base: ENV.fetch("LISA_LLM_URI_BASE", "https://api.listenai.com")
       ) do |faraday|
-        faraday.adapter :async_http, clients: Async::HTTP::Faraday::PersistentClients
+        faraday.adapter Faraday.default_adapter, clients: Async::HTTP::Faraday::PersistentClients
+      end
+    end
+
+    def judge_llm_client
+      @judge_llm_client ||= OpenAI::Client.new(
+        log_errors: true,
+        access_token: env!("JUDGE_ACCCESS_TOKEN"),
+        request_timeout: ENV.fetch("LISA_LLM_REQUEST_TIMEOUT", 120).to_i,
+        uri_base: ENV.fetch("JUDGE_LLM_URI_BASE", "https://api.listenai.com")
+      ) do |faraday|
+        faraday.adapter Faraday.default_adapter, clients: Async::HTTP::Faraday::PersistentClients
       end
     end
 
@@ -28,4 +40,4 @@ class Application
       ENV[name] or raise "missing environment variable: #{name}"
     end
   end
-end
+end

data/lib/kaba/test_runner.rb

@@ -26,7 +26,8 @@ class TestRunner
     model: 'spark-general-4.0', 
     judge_model: 'spark-general-4.0',  
     judge_temperature: 0.1, 
-    temperature: 0.1
+    temperature: 0.1,
+    semaphore_limit: 5
     )
 
     progressbar = TTY::ProgressBar.new(
@@ -37,8 +38,9 @@ class TestRunner
     progressbar.start
 
     Async do
+      semaphore = Async::Semaphore.new(semaphore_limit)
       _each(limit: limit) do |row|
-        Async do |task|
+        semaphore.async do |task|
           input = @prompt.render(File.read row.input_file)
 
           target = <<~Markdown
@@ -46,14 +48,17 @@ class TestRunner
           #{JSON.pretty_generate(JSON.parse(File.read(row.target_path)))}
           ```
           Markdown
-          output = Application.llm_client.chat(
+          output = ""
+          Application.llm_client.chat(
             parameters: {
               model: model,
               messages: [ { role: 'user', content: input } ],
               temperature: temperature,
+              stream: proc do |chunk, _bytesize|
+                output += chunk.dig("choices", 0, "delta", "content")
+              end
             }
-          ).dig("choices", 0, "message", "content")
-          
+          )
 
           output_json = JSON.parse_llm_response output
 
@@ -61,7 +66,7 @@ class TestRunner
           @type_right_total += 1 if type_check_response["success"]
           
           judge_input = Judge.new(input: input, output: output, target: target).render
-          judge_response = Application.llm_client.chat(
+          judge_response = Application.judge_llm_client.chat(
             parameters: {
               model: judge_model,
               messages: [ { role: 'user', content: judge_input } ],
@@ -106,4 +111,4 @@ class TestRunner
     end
   end
 
-end
+end

data/lib/kaba/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Kaba
-  VERSION = "0.3.0"
+  VERSION = "0.4.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: kaba
 version: !ruby/object:Gem::Version
-  version: 0.3.0
+  version: 0.4.0
 platform: ruby
 authors:
 - MJ
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-11-15 00:00:00.000000000 Z
+date: 2024-11-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: async
@@ -153,7 +153,8 @@ files:
 - lib/kaba/version.rb
 - sig/kaba.rbs
 homepage: https://github.com/mjason/kaba.git
-licenses: []
+licenses:
+- MIT
 metadata:
   homepage_uri: https://github.com/mjason/kaba.git
   source_code_uri: https://github.com/mjason/kaba.git