huginn_ollama_agent 0.1.0 → 0.1.11

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: e3fa742e907e01a2d71241761e082ab51bc8488192278df8693523927eef9a7e
-  data.tar.gz: 618ef39a32e3ccc4ca3da53cc2d8f1876cc3706ad53d2cb66e1e928a19f3bc9f
+  metadata.gz: f2528d73ee7c66815ff599fea1ca57382036a367670849a0c9d1180c94dd4035
+  data.tar.gz: 73f49b9773db8c5015da9d42c01764023c61d76bfa9badac00b289446c20ca8f
 SHA512:
-  metadata.gz: 5398a357bdd6c5a22ef0ce4778f3d77a1f4bed97a50de6d4da54cd86e1a4093d1df3dd141b4656b57887d44c97bd45d51056aa665d805c8a84c0de8b6b4b71f4
-  data.tar.gz: aa2abc6c4a51a51d2274507159bb1c75dd9d37a1ca88946ef1a142234b769b17806853146b97cf09de2c9fc59292db91e2d0ab323a88d5c7cc8a90597ed4f924
+  metadata.gz: 5ad7f7f1881cb615762106c7e0f1d25aefe031515c61d1fc3c1e498ac130e8c84b3d449cb12444823c0ba0fb7154208c1255c563c09606a73e4d0bc167c931b1
+  data.tar.gz: 853c508e1a36aa8c0b1f96d73703c05fd73a220097f374080bd87bf2c99aa2cd004602ffea18c6cc485c7c3c7a3517079a0beec49f8ae98c800421e4597a9543

data/lib/huginn_ollama_agent/ollama_agent.rb

@@ -17,6 +17,8 @@ module Agents
 
       `prompt` the prompt to generate a response.
 
+      `image` is a base64-encoded image or URL (for multimodal models such as llava).
+
       `context` the context parameter returned from a previous request to /generate, this can be used to keep a short conversational memory.
 
       `stream` if false the response will be returned as a single response object, rather than a stream of objects.
@@ -140,6 +142,7 @@ module Agents
     form_configurable :model, type: :string
     form_configurable :url, type: :string
     form_configurable :prompt, type: :string
+    form_configurable :image, type: :string
     form_configurable :context, type: :string
     form_configurable :stream, type: :boolean
     form_configurable :raw, type: :boolean
@@ -246,6 +249,38 @@ module Agents
       false
     end
 
+    def encode_to_base64(data)
+      return Base64.strict_encode64(data)
+    end
+
+    def download_and_convert_to_base64(url)
+      begin
+        uri = URI(url)
+        response = Net::HTTP.get_response(uri)
+
+        log_curl_output(response.code,response.body)
+
+        if response.is_a?(Net::HTTPSuccess)
+          file_data = response.body
+          base64_data = encode_to_base64(file_data)
+          return base64_data
+        end
+      end
+    end
+
+    def detect_image_source(input)
+      if input =~ URI::DEFAULT_PARSER.regexp[:ABS_URI]
+        download_and_convert_to_base64(interpolated['image'])
+      elsif input.match?(/\A(?:[A-Za-z0-9+\/]{4})*(?:[A-Za-z0-9+\/]{2}==|[A-Za-z0-9+\/]{3}=)?\z/)
+        begin
+          decoded = Base64.strict_decode64(input)
+          if Base64.strict_encode64(decoded) == input
+            return input
+          end
+        end
+      end
+    end
+
     def generate_completion()
 
       request_payload = {}
@@ -254,6 +289,7 @@ module Agents
       request_payload['stream'] = boolify(interpolated['stream'])
       request_payload['raw'] = boolify(interpolated['raw'])
       request_payload['context'] = context if !interpolated['context'].empty?
+      request_payload['images'] = ["#{detect_image_source(interpolated['image'])}"] if !interpolated['image'].empty?
 
       if check_remote_model()
         if interpolated['debug'] == 'true'
@@ -272,6 +308,7 @@ module Agents
 
       req_options = {
         use_ssl: uri.scheme == "https",
+        read_timeout: 120
       }
 
       response = Net::HTTP.start(uri.hostname, uri.port, req_options) do |http|

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: huginn_ollama_agent
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.11
 platform: ruby
 authors:
 - Nicolas Germain
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2024-03-25 00:00:00.000000000 Z
+date: 2024-03-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler