discourse/discourse-ai
1
0
Fork 0
mirror of https://github.com/discourse/discourse-ai.git synced 2025-07-08 15:22:47 +00:00
Code Issues Packages Projects Releases Wiki Activity

FEATURE: PDF support for rag pipeline (#1118)

Browse Source 
This PR introduces several enhancements and refactorings to the AI Persona and RAG (Retrieval-Augmented Generation) functionalities within the discourse-ai plugin. Here's a breakdown of the changes:

**1. LLM Model Association for RAG and Personas:**

-   **New Database Columns:** Adds `rag_llm_model_id` to both `ai_personas` and `ai_tools` tables. This allows specifying a dedicated LLM for RAG indexing, separate from the persona's primary LLM.  Adds `default_llm_id` and `question_consolidator_llm_id` to `ai_personas`.
-   **Migration:**  Includes a migration (`20250210032345_migrate_persona_to_llm_model_id.rb`) to populate the new `default_llm_id` and `question_consolidator_llm_id` columns in `ai_personas` based on the existing `default_llm` and `question_consolidator_llm` string columns, and a post migration to remove the latter.
-   **Model Changes:**  The `AiPersona` and `AiTool` models now `belong_to` an `LlmModel` via `rag_llm_model_id`. The `LlmModel.proxy` method now accepts an `LlmModel` instance instead of just an identifier.  `AiPersona` now has `default_llm_id` and `question_consolidator_llm_id` attributes.
-   **UI Updates:**  The AI Persona and AI Tool editors in the admin panel now allow selecting an LLM for RAG indexing (if PDF/image support is enabled).  The RAG options component displays an LLM selector.
-   **Serialization:** The serializers (`AiCustomToolSerializer`, `AiCustomToolListSerializer`, `LocalizedAiPersonaSerializer`) have been updated to include the new `rag_llm_model_id`, `default_llm_id` and `question_consolidator_llm_id` attributes.

**2. PDF and Image Support for RAG:**

-   **Site Setting:** Introduces a new hidden site setting, `ai_rag_pdf_images_enabled`, to control whether PDF and image files can be indexed for RAG. This defaults to `false`.
-   **File Upload Validation:** The `RagDocumentFragmentsController` now checks the `ai_rag_pdf_images_enabled` setting and allows PDF, PNG, JPG, and JPEG files if enabled.  Error handling is included for cases where PDF/image indexing is attempted with the setting disabled.
-   **PDF Processing:** Adds a new utility class, `DiscourseAi::Utils::PdfToImages`, which uses ImageMagick (`magick`) to convert PDF pages into individual PNG images. A maximum PDF size and conversion timeout are enforced.
-   **Image Processing:** A new utility class, `DiscourseAi::Utils::ImageToText`, is included to handle OCR for the images and PDFs.
-   **RAG Digestion Job:** The `DigestRagUpload` job now handles PDF and image uploads. It uses `PdfToImages` and `ImageToText` to extract text and create document fragments.
-   **UI Updates:**  The RAG uploader component now accepts PDF and image file types if `ai_rag_pdf_images_enabled` is true. The UI text is adjusted to indicate supported file types.

**3. Refactoring and Improvements:**

-   **LLM Enumeration:** The `DiscourseAi::Configuration::LlmEnumerator` now provides a `values_for_serialization` method, which returns a simplified array of LLM data (id, name, vision_enabled) suitable for use in serializers. This avoids exposing unnecessary details to the frontend.
-   **AI Helper:** The `AiHelper::Assistant` now takes optional `helper_llm` and `image_caption_llm` parameters in its constructor, allowing for greater flexibility.
-   **Bot and Persona Updates:** Several updates were made across the codebase, changing the string based association to a LLM to the new model based.
-   **Audit Logs:** The `DiscourseAi::Completions::Endpoints::Base` now formats raw request payloads as pretty JSON for easier auditing.
- **Eval Script:** An evaluation script is included.

**4. Testing:**

-    The PR introduces a new eval system for LLMs, this allows us to test how functionality works across various LLM providers. This lives in `/evals`
This commit is contained in:
Sam 2025-02-14 12:15:07 +11:00 committed by GitHub
parent e2afbc26d3
commit 5e80f93e4c
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
54 changed files with 1329 additions and 141 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
.gitignore vendored
View File

@ -2,3 +2,5 @@ node_modules
/gems
/auto_generated
.env
evals/log
evals/cases

2
admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-tools-edit.js
View File

@ -14,5 +14,7 @@ export default class DiscourseAiToolsEditRoute extends DiscourseRoute {
controller.set("allTools", toolsModel);
controller.set("presets", toolsModel.resultSetMeta.presets);
controller.set("llms", toolsModel.resultSetMeta.llms);
controller.set("settings", toolsModel.resultSetMeta.settings);
}
}

2
admin/assets/javascripts/discourse/routes/admin-plugins-show-discourse-ai-tools-new.js
View File

@ -11,5 +11,7 @@ export default class DiscourseAiToolsNewRoute extends DiscourseRoute {
controller.set("allTools", toolsModel);
controller.set("presets", toolsModel.resultSetMeta.presets);
controller.set("llms", toolsModel.resultSetMeta.llms);
controller.set("settings", toolsModel.resultSetMeta.settings);
}
}

2
admin/assets/javascripts/discourse/templates/admin-plugins/show/discourse-ai-tools/edit.hbs
View File

@ -3,5 +3,7 @@
@tools={{this.allTools}}
@model={{this.model}}
@presets={{this.presets}}
@llms={{this.llms}}
@settings={{this.settings}}
/>
</section>

2
admin/assets/javascripts/discourse/templates/admin-plugins/show/discourse-ai-tools/new.hbs
View File

@ -3,5 +3,7 @@
@tools={{this.allTools}}
@model={{this.model}}
@presets={{this.presets}}
@llms={{this.llms}}
@settings={{this.settings}}
/>
</section>

22
app/controllers/discourse_ai/admin/ai_personas_controller.rb
View File

@ -32,10 +32,19 @@ module DiscourseAi
}
end
llms =
DiscourseAi::Configuration::LlmEnumerator
.values(allowed_seeded_llms: SiteSetting.ai_bot_allowed_seeded_models)
.map { |hash| { id: hash[:value], name: hash[:name] } }
render json: { ai_personas: ai_personas, meta: { tools: tools, llms: llms } }
DiscourseAi::Configuration::LlmEnumerator.values_for_serialization(
allowed_seeded_llm_ids: SiteSetting.ai_bot_allowed_seeded_models_map,
)
render json: {
ai_personas: ai_personas,
meta: {
tools: tools,
llms: llms,
settings: {
rag_pdf_images_enabled: SiteSetting.ai_rag_pdf_images_enabled,
},
},
}
end
def new
@ -187,7 +196,7 @@ module DiscourseAi
:priority,
:top_p,
:temperature,
:default_llm,
:default_llm_id,
:user_id,
:max_context_posts,
:vision_enabled,
@ -195,7 +204,8 @@ module DiscourseAi
:rag_chunk_tokens,
:rag_chunk_overlap_tokens,
:rag_conversation_chunks,
:question_consolidator_llm,
:rag_llm_model_id,
:question_consolidator_llm_id,
:allow_chat_channel_mentions,
:allow_chat_direct_messages,
:allow_topic_mentions,

1
app/controllers/discourse_ai/admin/ai_tools_controller.rb
View File

@ -90,6 +90,7 @@ module DiscourseAi
:summary,
:rag_chunk_tokens,
:rag_chunk_overlap_tokens,
:rag_llm_model_id,
rag_uploads: [:id],
parameters: [:name, :type, :description, :required, enum: []],
)

1
app/controllers/discourse_ai/admin/rag_document_fragments_controller.rb
View File

@ -49,6 +49,7 @@ module DiscourseAi
def validate_extension!(filename)
extension = File.extname(filename)[1..-1] || ""
authorized_extensions = %w[txt md]
authorized_extensions.concat(%w[pdf png jpg jpeg]) if SiteSetting.ai_rag_pdf_images_enabled
if !authorized_extensions.include?(extension)
raise Discourse::InvalidParameters.new(
I18n.t(

35
app/jobs/regular/digest_rag_upload.rb
View File

@ -28,7 +28,7 @@ module ::Jobs
# Check if this is the first time we process this upload.
if fragment_ids.empty?
document = get_uploaded_file(upload)
document = get_uploaded_file(upload: upload, target: target)
return if document.nil?
RagDocumentFragment.publish_status(upload, { total: 0, indexed: 0, left: 0 })
@ -163,7 +163,38 @@ module ::Jobs
[buffer, split_char]
end
def get_uploaded_file(upload)
def get_uploaded_file(upload:, target:)
if %w[pdf png jpg jpeg].include?(upload.extension) && !SiteSetting.ai_rag_pdf_images_enabled
raise Discourse::InvalidAccess.new(
"The setting ai_rag_pdf_images_enabled is false, can not index images and pdfs.",
)
end
if upload.extension == "pdf"
pages =
DiscourseAi::Utils::PdfToImages.new(
upload: upload,
user: Discourse.system_user,
).uploaded_pages
return(
DiscourseAi::Utils::ImageToText.as_fake_file(
uploads: pages,
llm_model: target.rag_llm_model,
user: Discourse.system_user,
)
)
end
if %w[png jpg jpeg].include?(upload.extension)
return(
DiscourseAi::Utils::ImageToText.as_fake_file(
uploads: [upload],
llm_model: target.rag_llm_model,
user: Discourse.system_user,
)
)
end
store = Discourse.store
@file ||=
if store.external?

32
app/models/ai_persona.rb
View File

@ -1,8 +1,8 @@
# frozen_string_literal: true
class AiPersona < ActiveRecord::Base
# TODO remove this line 01-1-2025
self.ignored_columns = %i[commands allow_chat mentionable]
# TODO remove this line 01-10-2025
self.ignored_columns = %i[default_llm question_consolidator_llm]
# places a hard limit, so per site we cache a maximum of 500 classes
MAX_PERSONAS_PER_SITE = 500
@ -12,7 +12,7 @@ class AiPersona < ActiveRecord::Base
validates :system_prompt, presence: true, length: { maximum: 10_000_000 }
validate :system_persona_unchangeable, on: :update, if: :system
validate :chat_preconditions
validate :allowed_seeded_model, if: :default_llm
validate :allowed_seeded_model, if: :default_llm_id
validates :max_context_posts, numericality: { greater_than: 0 }, allow_nil: true
# leaves some room for growth but sets a maximum to avoid memory issues
# we may want to revisit this in the future
@ -30,6 +30,10 @@ class AiPersona < ActiveRecord::Base
belongs_to :created_by, class_name: "User"
belongs_to :user
belongs_to :default_llm, class_name: "LlmModel"
belongs_to :question_consolidator_llm, class_name: "LlmModel"
belongs_to :rag_llm_model, class_name: "LlmModel"
has_many :upload_references, as: :target, dependent: :destroy
has_many :uploads, through: :upload_references
@ -62,7 +66,7 @@ class AiPersona < ActiveRecord::Base
user_id: persona.user_id,
username: persona.user.username_lower,
allowed_group_ids: persona.allowed_group_ids,
default_llm: persona.default_llm,
default_llm_id: persona.default_llm_id,
force_default_llm: persona.force_default_llm,
allow_chat_channel_mentions: persona.allow_chat_channel_mentions,
allow_chat_direct_messages: persona.allow_chat_direct_messages,
@ -157,12 +161,12 @@ class AiPersona < ActiveRecord::Base
user_id
system
mentionable
default_llm
default_llm_id
max_context_posts
vision_enabled
vision_max_pixels
rag_conversation_chunks
question_consolidator_llm
question_consolidator_llm_id
allow_chat_channel_mentions
allow_chat_direct_messages
allow_topic_mentions
@ -302,7 +306,7 @@ class AiPersona < ActiveRecord::Base
if (
allow_chat_channel_mentions || allow_chat_direct_messages || allow_topic_mentions ||
force_default_llm
) && !default_llm
) && !default_llm_id
errors.add(:default_llm, I18n.t("discourse_ai.ai_bot.personas.default_llm_required"))
end
end
@ -332,13 +336,12 @@ class AiPersona < ActiveRecord::Base
end
def allowed_seeded_model
return if default_llm.blank?
return if default_llm_id.blank?
llm = LlmModel.find_by(id: default_llm.split(":").last.to_i)
return if llm.nil?
return if !llm.seeded?
return if default_llm.nil?
return if !default_llm.seeded?
return if SiteSetting.ai_bot_allowed_seeded_models.include?(llm.id.to_s)
return if SiteSetting.ai_bot_allowed_seeded_models_map.include?(default_llm.id.to_s)
errors.add(:default_llm, I18n.t("discourse_ai.llm.configuration.invalid_seeded_model"))
end
@ -362,14 +365,12 @@ end
# temperature :float
# top_p :float
# user_id :integer
# default_llm :text
# max_context_posts :integer
# vision_enabled :boolean default(FALSE), not null
# vision_max_pixels :integer default(1048576), not null
# rag_chunk_tokens :integer default(374), not null
# rag_chunk_overlap_tokens :integer default(10), not null
# rag_conversation_chunks :integer default(10), not null
# question_consolidator_llm :text
# tool_details :boolean default(TRUE), not null
# tools :json not null
# forced_tool_count :integer default(-1), not null
@ -378,6 +379,9 @@ end
# allow_topic_mentions :boolean default(FALSE), not null
# allow_personal_messages :boolean default(TRUE), not null
# force_default_llm :boolean default(FALSE), not null
# rag_llm_model_id :bigint
# default_llm_id :bigint
# question_consolidator_llm_id :bigint
#
# Indexes
#

3
app/models/ai_tool.rb
View File

@ -8,6 +8,7 @@ class AiTool < ActiveRecord::Base
validates :script, presence: true, length: { maximum: 100_000 }
validates :created_by_id, presence: true
belongs_to :created_by, class_name: "User"
belongs_to :rag_llm_model, class_name: "LlmModel"
has_many :rag_document_fragments, dependent: :destroy, as: :target
has_many :upload_references, as: :target, dependent: :destroy
has_many :uploads, through: :upload_references
@ -371,4 +372,4 @@ end
# rag_chunk_tokens :integer default(374), not null
# rag_chunk_overlap_tokens :integer default(10), not null
# tool_name :string(100) default(""), not null
#
# rag_llm_model_id :bigint

2
app/models/llm_model.rb
View File

@ -70,7 +70,7 @@ class LlmModel < ActiveRecord::Base
end
def to_llm
DiscourseAi::Completions::Llm.proxy(identifier)
DiscourseAi::Completions::Llm.proxy(self)
end
def identifier

8
app/serializers/ai_custom_tool_list_serializer.rb
View File

@ -6,7 +6,13 @@ class AiCustomToolListSerializer < ApplicationSerializer
has_many :ai_tools, serializer: AiCustomToolSerializer, embed: :objects
def meta
{ presets: AiTool.presets }
{
presets: AiTool.presets,
llms: DiscourseAi::Configuration::LlmEnumerator.values_for_serialization,
settings: {
rag_pdf_images_enabled: SiteSetting.ai_rag_pdf_images_enabled,
},
}
end
def ai_tools

1
app/serializers/ai_custom_tool_serializer.rb
View File

@ -10,6 +10,7 @@ class AiCustomToolSerializer < ApplicationSerializer
:script,
:rag_chunk_tokens,
:rag_chunk_overlap_tokens,
:rag_llm_model_id,
:created_by_id,
:created_at,
:updated_at

5
app/serializers/localized_ai_persona_serializer.rb
View File

@ -14,7 +14,7 @@ class LocalizedAiPersonaSerializer < ApplicationSerializer
:allowed_group_ids,
:temperature,
:top_p,
:default_llm,
:default_llm_id,
:user_id,
:max_context_posts,
:vision_enabled,
@ -22,7 +22,8 @@ class LocalizedAiPersonaSerializer < ApplicationSerializer
:rag_chunk_tokens,
:rag_chunk_overlap_tokens,
:rag_conversation_chunks,
:question_consolidator_llm,
:rag_llm_model_id,
:question_consolidator_llm_id,
:tool_details,
:forced_tool_count,
:allow_chat_channel_mentions,

10
assets/javascripts/discourse/admin/models/ai-persona.js
View File

@ -15,7 +15,7 @@ const CREATE_ATTRIBUTES = [
"top_p",
"temperature",
"user_id",
"default_llm",
"default_llm_id",
"force_default_llm",
"user",
"max_context_posts",
@ -25,7 +25,8 @@ const CREATE_ATTRIBUTES = [
"rag_chunk_tokens",
"rag_chunk_overlap_tokens",
"rag_conversation_chunks",
"question_consolidator_llm",
"rag_llm_model_id",
"question_consolidator_llm_id",
"allow_chat",
"tool_details",
"forced_tool_count",
@ -43,7 +44,7 @@ const SYSTEM_ATTRIBUTES = [
"priority",
"tools",
"user_id",
"default_llm",
"default_llm_id",
"force_default_llm",
"user",
"max_context_posts",
@ -53,7 +54,8 @@ const SYSTEM_ATTRIBUTES = [
"rag_chunk_tokens",
"rag_chunk_overlap_tokens",
"rag_conversation_chunks",
"question_consolidator_llm",
"rag_llm_model_id",
"question_consolidator_llm_id",
"tool_details",
"allow_personal_messages",
"allow_topic_mentions",

1
assets/javascripts/discourse/admin/models/ai-tool.js
View File

@ -12,6 +12,7 @@ const CREATE_ATTRIBUTES = [
"rag_uploads",
"rag_chunk_tokens",
"rag_chunk_overlap_tokens",
"rag_llm_model_id",
"enabled",
];

19
assets/javascripts/discourse/components/ai-persona-editor.gjs
View File

@ -167,27 +167,27 @@ export default class PersonaEditor extends Component {
}
get mappedQuestionConsolidatorLlm() {
return this.editingModel?.question_consolidator_llm || "blank";
return this.editingModel?.question_consolidator_llm_id ?? "blank";
}
set mappedQuestionConsolidatorLlm(value) {
if (value === "blank") {
this.editingModel.question_consolidator_llm = null;
this.editingModel.question_consolidator_llm_id = null;
} else {
this.editingModel.question_consolidator_llm = value;
this.editingModel.question_consolidator_llm_id = value;
}
}
get mappedDefaultLlm() {
return this.editingModel?.default_llm || "blank";
return this.editingModel?.default_llm_id ?? "blank";
}
set mappedDefaultLlm(value) {
if (value === "blank") {
this.editingModel.default_llm = null;
this.editingModel.default_llm_id = null;
this.hasDefaultLlm = false;
} else {
this.editingModel.default_llm = value;
this.editingModel.default_llm_id = value;
this.hasDefaultLlm = true;
}
}
@ -596,9 +596,14 @@ export default class PersonaEditor extends Component {
@target={{this.editingModel}}
@updateUploads={{this.updateUploads}}
@onRemove={{this.removeUpload}}
@allowPdfsAndImages={{@personas.resultSetMeta.settings.rag_pdf_images_enabled}}
/>
</div>
<RagOptions @model={{this.editingModel}}>
<RagOptions
@model={{this.editingModel}}
@llms={{@personas.resultSetMeta.llms}}
@allowPdfsAndImages={{@personas.resultSetMeta.settings.rag_pdf_images_enabled}}
>
<div class="control-group">
<label>{{i18n
"discourse_ai.ai_persona.rag_conversation_chunks"

10
assets/javascripts/discourse/components/ai-tool-editor.gjs
View File

@ -90,7 +90,8 @@ export default class AiToolEditor extends Component {
"summary",
"rag_uploads",
"rag_chunk_tokens",
"rag_chunk_overlap_tokens"
"rag_chunk_overlap_tokens",
"rag_llm_model_id"
);
await this.args.model.save(data);
@ -244,9 +245,14 @@ export default class AiToolEditor extends Component {
@target={{this.editingModel}}
@updateUploads={{this.updateUploads}}
@onRemove={{this.removeUpload}}
@allowPdfsAndImages={{@settings.rag_pdf_images_enabled}}
/>
</div>
<RagOptions @model={{this.editingModel}} />
<RagOptions
@model={{this.editingModel}}
@llms={{@llms}}
@allowPdfsAndImages={{@settings.rag_pdf_images_enabled}}
/>
{{/if}}
<div class="control-group ai-tool-editor__action_panel">

31
assets/javascripts/discourse/components/rag-options.gjs
View File

@ -5,6 +5,7 @@ import { on } from "@ember/modifier";
import { action } from "@ember/object";
import DTooltip from "discourse/components/d-tooltip";
import { i18n } from "discourse-i18n";
import AiLlmSelector from "./ai-llm-selector";
export default class RagOptions extends Component {
@tracked showIndexingOptions = false;
@ -22,6 +23,22 @@ export default class RagOptions extends Component {
: i18n("discourse_ai.rag.options.show_indexing_options");
}
get visionLlms() {
return this.args.llms.filter((llm) => llm.vision_enabled);
}
get visionLlmId() {
return this.args.model.rag_llm_model_id ?? "blank";
}
set visionLlmId(value) {
if (value === "blank") {
this.args.model.rag_llm_model_id = null;
} else {
this.args.model.rag_llm_model_id = value;
}
}
<template>
{{#if @model.rag_uploads}}
<a
@ -64,6 +81,20 @@ export default class RagOptions extends Component {
}}
/>
</div>
{{#if @allowPdfsAndImages}}
<div class="control-group">
<label>{{i18n "discourse_ai.rag.options.rag_llm_model"}}</label>
<AiLlmSelector
class="ai-persona-editor__llms"
@value={{this.visionLlmId}}
@llms={{this.visionLlms}}
/>
<DTooltip
@icon="circle-question"
@content={{i18n "discourse_ai.rag.options.rag_llm_model_help"}}
/>
</div>
{{/if}}
{{yield}}
{{/if}}
</template>

14
assets/javascripts/discourse/components/rag-uploader.gjs
View File

@ -77,6 +77,14 @@ export default class RagUploader extends Component {
this.updateUploads(this.ragUploads);
}
get acceptedFileTypes() {
if (this.args?.allowPdfsAndImages) {
return ".txt,.md,.pdf,.png,.jpg,.jpeg";
} else {
return ".txt,.md";
}
}
@action
submitFiles() {
this.uppyUpload.openPicker();
@ -119,7 +127,11 @@ export default class RagUploader extends Component {
<template>
<div class="rag-uploader">
<h3>{{i18n "discourse_ai.rag.uploads.title"}}</h3>
{{#if @allowPdfsAndImages}}
<p>{{i18n "discourse_ai.rag.uploads.description_with_pdfs"}}</p>
{{else}}
<p>{{i18n "discourse_ai.rag.uploads.description"}}</p>
{{/if}}
{{#if this.ragUploads}}
<div class="rag-uploader__search-input-container">
@ -187,7 +199,7 @@ export default class RagUploader extends Component {
disabled={{this.uploading}}
type="file"
multiple="multiple"
accept=".txt,.md"
accept={{this.acceptedFileTypes}}
/>
<DButton
@label="discourse_ai.rag.uploads.button"

5
config/locales/client.en.yml
View File

@ -274,11 +274,14 @@ en:
rag_chunk_tokens_help: "The number of tokens to use for each chunk in the RAG model. Increase to increase the amount of context the AI can use. (changing will re-index all uploads)"
rag_chunk_overlap_tokens: "Upload chunk overlap tokens"
rag_chunk_overlap_tokens_help: "The number of tokens to overlap between chunks in the RAG model. (changing will re-index all uploads)"
rag_llm_model: "Indexing Language Model"
rag_llm_model_help: "The language model used for OCR during indexing of PDFs and images"
show_indexing_options: "Show upload options"
hide_indexing_options: "Hide upload options"
uploads:
title: "Uploads"
description: "Uploaded files should be formatted as plaintext (.txt) or markdown (.md)."
description: "Plaintext (.txt) or markdown (.md)"
description_with_pdfs: "Plaintext (.txt), markdown (.md), PDF (.pdf) or image (.png, .jpeg)"
button: "Add files"
filter: "Filter uploads"
indexed: "Indexed"

6
config/settings.yml
View File

@ -347,12 +347,14 @@ discourse_ai:
ai_spam_detection_enabled:
default: false
hidden: true
ai_spam_detection_user_id:
default: ""
hidden: true
ai_spam_detection_model_allowed_seeded_models:
default: ""
hidden: true
type: list
ai_rag_pdf_images_enabled:
default: false
hidden: true

7
db/migrate/20250210024600_add_rag_llm_model.rb Normal file
View File

@ -0,0 +1,7 @@
# frozen_string_literal: true
class AddRagLlmModel < ActiveRecord::Migration[7.2]
def change
add_column :ai_personas, :rag_llm_model_id, :bigint
add_column :ai_tools, :rag_llm_model_id, :bigint
end
end

19
db/migrate/20250210032345_migrate_persona_to_llm_model_id.rb Normal file
View File

@ -0,0 +1,19 @@
# frozen_string_literal: true
class MigratePersonaToLlmModelId < ActiveRecord::Migration[7.2]
def up
add_column :ai_personas, :default_llm_id, :bigint
add_column :ai_personas, :question_consolidator_llm_id, :bigint
# personas are seeded, we do not mark stuff as readonline
execute <<~SQL
UPDATE ai_personas
set
default_llm_id = (select id from llm_models where ('custom:' || id) = default_llm),
question_consolidator_llm_id = (select id from llm_models where ('custom:' || id) = question_consolidator_llm)
SQL
end
def down
raise ActiveRecord::IrreversibleMigration
end
end

11
db/post_migrate/20250210032351_post_migrate_persona_to_llm_model_id.rb Normal file
View File

@ -0,0 +1,11 @@
# frozen_string_literal: true
class PostMigratePersonaToLlmModelId < ActiveRecord::Migration[7.2]
def up
remove_column :ai_personas, :default_llm
remove_column :ai_personas, :question_consolidator_llm
end
def down
raise ActiveRecord::IrreversibleMigration
end
end

36
evals/lib/boot.rb Normal file
View File

@ -0,0 +1,36 @@
# frozen_string_literal: true
# got to ensure evals are here
# rubocop:disable Discourse/Plugins/NamespaceConstants
EVAL_PATH = File.join(__dir__, "../cases")
# rubocop:enable Discourse/Plugins/NamespaceConstants
#
if !Dir.exist?(EVAL_PATH)
puts "Evals are missing, cloning from discourse/discourse-ai-evals"
success =
system("git clone git@github.com:discourse/discourse-ai-evals.git '#{EVAL_PATH}' 2>/dev/null")
# Fall back to HTTPS if SSH fails
if !success
puts "SSH clone failed, falling back to HTTPS..."
success = system("git clone https://github.com/discourse/discourse-ai-evals.git '#{EVAL_PATH}'")
end
if success
puts "Successfully cloned evals repository"
else
abort "Failed to clone evals repository"
end
end
discourse_path = File.expand_path(File.join(__dir__, "../../../.."))
# rubocop:disable Discourse/NoChdir
Dir.chdir(discourse_path)
# rubocop:enable Discourse/NoChdir
require "/home/sam/Source/discourse/config/environment"
ENV["DISCOURSE_AI_NO_DEBUG"] = "1"
module DiscourseAi::Evals
end

47
evals/lib/cli.rb Normal file
View File

@ -0,0 +1,47 @@
# frozen_string_literal: true
require "optparse"
class DiscourseAi::Evals::Cli
class Options
attr_accessor :eval_name, :model, :list, :list_models
def initialize(eval_name: nil, model: nil, list: false, list_models: false)
@eval_name = eval_name
@model = model
@list = list
@list_models = list_models
end
end
def self.parse_options!
options = Options.new
parser =
OptionParser.new do |opts|
opts.banner = "Usage: evals/run [options]"
opts.on("-e", "--eval NAME", "Name of the evaluation to run") do |eval_name|
options.eval_name = eval_name
end
opts.on("--list-models", "List models") { |model| options.list_models = true }
opts.on(
"-m",
"--model NAME",
"Model to evaluate (will eval all models if not specified)",
) { |model| options.model = model }
opts.on("-l", "--list", "List evals") { |model| options.list = true }
end
show_help = ARGV.empty?
parser.parse!
if show_help
puts parser
exit 0
end
options
end
end

136
evals/lib/eval.rb Normal file
View File

@ -0,0 +1,136 @@
#frozen_string_literal: true
class DiscourseAi::Evals::Eval
attr_reader :type,
:path,
:name,
:description,
:id,
:args,
:vision,
:expected_output,
:expected_output_regex
def initialize(path:)
@yaml = YAML.load_file(path).symbolize_keys
@path = path
@name = @yaml[:name]
@id = @yaml[:id]
@description = @yaml[:description]
@vision = @yaml[:vision]
@args = @yaml[:args]&.symbolize_keys
@type = @yaml[:type]
@expected_output = @yaml[:expected_output]
@expected_output_regex = @yaml[:expected_output_regex]
@expected_output_regex =
Regexp.new(@expected_output_regex, Regexp::MULTILINE) if @expected_output_regex
@args[:path] = File.expand_path(File.join(File.dirname(path), @args[:path])) if @args&.key?(
:path,
)
end
def run(llm:)
result =
case type
when "helper"
helper(llm, **args)
when "pdf_to_text"
pdf_to_text(llm, **args)
when "image_to_text"
image_to_text(llm, **args)
end
if expected_output
if result == expected_output
{ result: :pass }
else
{ result: :fail, expected_output: expected_output, actual_output: result }
end
elsif expected_output_regex
if result.match?(expected_output_regex)
{ result: :pass }
else
{ result: :fail, expected_output: expected_output_regex, actual_output: result }
end
else
{ result: :unknown, actual_output: result }
end
end
def print
puts "#{id}: #{description}"
end
def to_json
{
type: @type,
path: @path,
name: @name,
description: @description,
id: @id,
args: @args,
vision: @vision,
expected_output: @expected_output,
expected_output_regex: @expected_output_regex,
}.compact
end
private
def helper(llm, input:, name:)
completion_prompt = CompletionPrompt.find_by(name: name)
helper = DiscourseAi::AiHelper::Assistant.new(helper_llm: llm.llm_proxy)
result =
helper.generate_and_send_prompt(
completion_prompt,
input,
current_user = Discourse.system_user,
_force_default_locale = false,
)
result[:suggestions].first
end
def image_to_text(llm, path:)
upload =
UploadCreator.new(File.open(path), File.basename(path)).create_for(Discourse.system_user.id)
text = +""
DiscourseAi::Utils::ImageToText
.new(upload: upload, llm_model: llm.llm_model, user: Discourse.system_user)
.extract_text do |chunk, error|
text << chunk if chunk
text << "\n\n" if chunk
end
text
ensure
upload.destroy if upload
end
def pdf_to_text(llm, path:)
upload =
UploadCreator.new(File.open(path), File.basename(path)).create_for(Discourse.system_user.id)
uploads =
DiscourseAi::Utils::PdfToImages.new(
upload: upload,
user: Discourse.system_user,
).uploaded_pages
text = +""
uploads.each do |page_upload|
DiscourseAi::Utils::ImageToText
.new(upload: page_upload, llm_model: llm.llm_model, user: Discourse.system_user)
.extract_text do |chunk, error|
text << chunk if chunk
text << "\n\n" if chunk
end
upload.destroy
end
text
ensure
upload.destroy if upload
end
end

129
evals/lib/llm.rb Normal file
View File

@ -0,0 +1,129 @@
# frozen_string_literal: true
class DiscourseAi::Evals::Llm
CONFIGS = {
"gpt-4o" => {
display_name: "GPT-4o",
name: "gpt-4o",
tokenizer: "DiscourseAi::Tokenizer::OpenAiTokenizer",
api_key_env: "OPENAI_API_KEY",
provider: "open_ai",
url: "https://api.openai.com/v1/chat/completions",
max_prompt_tokens: 131_072,
vision_enabled: true,
},
"gpt-4o-mini" => {
display_name: "GPT-4o-mini",
name: "gpt-4o-mini",
tokenizer: "DiscourseAi::Tokenizer::OpenAiTokenizer",
api_key_env: "OPENAI_API_KEY",
provider: "open_ai",
url: "https://api.openai.com/v1/chat/completions",
max_prompt_tokens: 131_072,
vision_enabled: true,
},
"claude-3.5-haiku" => {
display_name: "Claude 3.5 Haiku",
name: "claude-3-5-haiku-latest",
tokenizer: "DiscourseAi::Tokenizer::AnthropicTokenizer",
api_key_env: "ANTHROPIC_API_KEY",
provider: "anthropic",
url: "https://api.anthropic.com/v1/messages",
max_prompt_tokens: 200_000,
vision_enabled: false,
},
"claude-3.5-sonnet" => {
display_name: "Claude 3.5 Sonnet",
name: "claude-3-5-sonnet-latest",
tokenizer: "DiscourseAi::Tokenizer::AnthropicTokenizer",
api_key_env: "ANTHROPIC_API_KEY",
provider: "anthropic",
url: "https://api.anthropic.com/v1/messages",
max_prompt_tokens: 200_000,
vision_enabled: true,
},
"gemini-2.0-flash" => {
display_name: "Gemini 2.0 Flash",
name: "gemini-2-0-flash",
tokenizer: "DiscourseAi::Tokenizer::GeminiTokenizer",
api_key_env: "GEMINI_API_KEY",
provider: "google",
url: "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash",
max_prompt_tokens: 1_000_000,
vision_enabled: true,
},
"gemini-2.0-pro-exp" => {
display_name: "Gemini 2.0 pro",
name: "gemini-2-0-pro-exp",
tokenizer: "DiscourseAi::Tokenizer::GeminiTokenizer",
api_key_env: "GEMINI_API_KEY",
provider: "google",
url: "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-pro-exp",
max_prompt_tokens: 1_000_000,
vision_enabled: true,
},
}
def self.print
CONFIGS
.keys
.map do |config_name|
begin
new(config_name)
rescue StandardError
nil
end
end
.compact
.each { |llm| puts "#{llm.config_name}: #{llm.name} (#{llm.provider})" }
end
def self.choose(config_name)
if CONFIGS[config_name].nil?
CONFIGS
.keys
.map do |config_name|
begin
new(config_name)
rescue => e
puts "Error initializing #{config_name}: #{e}"
nil
end
end
.compact
elsif !CONFIGS.include?(config_name)
raise "Invalid llm"
else
[new(config_name)]
end
end
attr_reader :llm_model
attr_reader :llm_proxy
attr_reader :config_name
def initialize(config_name)
config = CONFIGS[config_name].dup
api_key_env = config.delete(:api_key_env)
if !ENV[api_key_env]
raise "Missing API key for #{config_name}, should be set via #{api_key_env}"
end
config[:api_key] = ENV[api_key_env]
@llm_model = LlmModel.new(config)
@llm_proxy = DiscourseAi::Completions::Llm.proxy(@llm_model)
@config_name = config_name
end
def provider
@llm_model.provider
end
def name
@llm_model.display_name
end
def vision?
@llm_model.vision_enabled
end
end

183
evals/lib/runner.rb Normal file
View File

@ -0,0 +1,183 @@
#frozen_string_literal: true
class DiscourseAi::Evals::Runner
class StructuredLogger
def initialize
@log = []
@current_step = @log
end
def log(name, args: nil, start_time: nil, end_time: nil)
start_time ||= Time.now.utc
end_time ||= Time.now.utc
args ||= {}
object = { name: name, args: args, start_time: start_time, end_time: end_time }
@current_step << object
end
def step(name, args: nil)
start_time = Time.now.utc
start_step = @current_step
new_step = { type: :step, name: name, args: args || {}, log: [], start_time: start_time }
@current_step << new_step
@current_step = new_step[:log]
yield new_step
@current_step = start_step
new_step[:end_time] = Time.now.utc
end
def to_trace_event_json
trace_events = []
process_id = 1
thread_id = 1
to_trace_event(@log, process_id, thread_id, trace_events)
JSON.pretty_generate({ traceEvents: trace_events })
end
private
def to_trace_event(log_items, pid, tid, trace_events, parent_start_time = nil)
log_items.each do |item|
if item.is_a?(Hash) && item[:type] == :step
trace_events << {
name: item[:name],
cat: "default",
ph: "B", # Begin event
pid: pid,
tid: tid,
args: item[:args],
ts: timestamp_in_microseconds(item[:start_time]),
}
to_trace_event(item[:log], pid, tid, trace_events, item[:start_time])
trace_events << {
name: item[:name],
cat: "default",
ph: "E", # End event
pid: pid,
tid: tid,
ts: timestamp_in_microseconds(item[:end_time]),
}
else
trace_events << {
name: item[:name],
cat: "default",
ph: "B",
pid: pid,
tid: tid,
args: item[:args],
ts: timestamp_in_microseconds(item[:start_time] || parent_start_time || Time.now.utc),
s: "p", # Scope: process
}
trace_events << {
name: item[:name],
cat: "default",
ph: "E",
pid: pid,
tid: tid,
ts: timestamp_in_microseconds(item[:end_time] || Time.now.utc),
s: "p",
}
end
end
end
def timestamp_in_microseconds(time)
(time.to_f * 1_000_000).to_i
end
end
attr_reader :llms, :cases
def self.evals_paths
@eval_paths ||= Dir.glob(File.join(File.join(__dir__, "../cases"), "*/*.yml"))
end
def self.evals
@evals ||= evals_paths.map { |path| DiscourseAi::Evals::Eval.new(path: path) }
end
def self.print
evals.each(&:print)
end
def initialize(eval_name:, llms:)
@llms = llms
@eval = self.class.evals.find { |c| c.id == eval_name }
if !@eval
puts "Error: Unknown evaluation '#{eval_name}'"
exit 1
end
if @llms.empty?
puts "Error: Unknown model 'model'"
exit 1
end
end
def run!
puts "Running evaluation '#{@eval.id}'"
structured_log_filename = "#{@eval.id}-#{Time.now.strftime("%Y%m%d-%H%M%S")}.json"
log_filename = "#{@eval.id}-#{Time.now.strftime("%Y%m%d-%H%M%S")}.log"
logs_dir = File.join(__dir__, "../log")
FileUtils.mkdir_p(logs_dir)
log_path = File.expand_path(File.join(logs_dir, log_filename))
structured_log_path = File.expand_path(File.join(logs_dir, structured_log_filename))
logger = Logger.new(File.open(log_path, "a"))
logger.info("Starting evaluation '#{@eval.id}'")
Thread.current[:llm_audit_log] = logger
structured_logger = Thread.current[:llm_audit_structured_log] = StructuredLogger.new
structured_logger.step("Evaluating #{@eval.id}", args: @eval.to_json) do
llms.each do |llm|
if @eval.vision && !llm.vision?
logger.info("Skipping LLM: #{llm.name} as it does not support vision")
next
end
structured_logger.step("Evaluating with LLM: #{llm.name}") do |step|
logger.info("Evaluating with LLM: #{llm.name}")
print "#{llm.name}: "
result = @eval.run(llm: llm)
step[:args] = result
step[:cname] = result[:result] == :pass ? :good : :bad
if result[:result] == :fail
puts "Failed 🔴"
puts "---- Expected ----\n#{result[:expected_output]}"
puts "---- Actual ----\n#{result[:actual_output]}"
logger.error("Evaluation failed with LLM: #{llm.name}")
elsif result[:result] == :pass
puts "Passed 🟢"
logger.info("Evaluation passed with LLM: #{llm.name}")
else
STDERR.puts "Error: Unknown result #{eval.inspect}"
logger.error("Unknown result: #{eval.inspect}")
end
end
end
end
#structured_logger.save(structured_log_path)
File.write("#{structured_log_path}", structured_logger.to_trace_event_json)
puts
puts "Log file: #{log_path}"
puts "Structured log file (ui.perfetto.dev): #{structured_log_path}"
# temp code
# puts File.read(structured_log_path)
end
end

25
evals/run Executable file
View File

@ -0,0 +1,25 @@
#!/usr/bin/env ruby
# frozen_string_literal: true
require_relative "lib/boot"
require_relative "lib/llm"
require_relative "lib/cli"
require_relative "lib/runner"
require_relative "lib/eval"
options = DiscourseAi::Evals::Cli.parse_options!
if options.list
DiscourseAi::Evals::Runner.print
exit 0
end
if options.list_models
DiscourseAi::Evals::Llm.print
exit 0
end
DiscourseAi::Evals::Runner.new(
eval_name: options.eval_name,
llms: DiscourseAi::Evals::Llm.choose(options.model),
).run!

5
lib/ai_bot/bot.rb
View File

@ -16,7 +16,8 @@ module DiscourseAi
def initialize(bot_user, persona, model = nil)
@bot_user = bot_user
@persona = persona
@model = model || self.class.guess_model(bot_user) || @persona.class.default_llm
@model =
model || self.class.guess_model(bot_user) || LlmModel.find(@persona.class.default_llm_id)
end
attr_reader :bot_user
@ -245,7 +246,7 @@ module DiscourseAi
return if associated_llm.nil? # Might be a persona user. Handled by constructor.
"custom:#{associated_llm.id}"
associated_llm
end
def build_placeholder(summary, details, custom_raw: nil)

10
lib/ai_bot/personas/persona.rb
View File

@ -17,7 +17,7 @@ module DiscourseAi
1_048_576
end
def question_consolidator_llm
def question_consolidator_llm_id
nil
end
@ -173,9 +173,11 @@ module DiscourseAi
TEXT
question_consolidator_llm = llm
if self.class.question_consolidator_llm.present?
question_consolidator_llm =
DiscourseAi::Completions::Llm.proxy(self.class.question_consolidator_llm)
if self.class.question_consolidator_llm_id.present?
question_consolidator_llm ||=
DiscourseAi::Completions::Llm.proxy(
LlmModel.find_by(id: self.class.question_consolidator_llm_id),
)
end
if context[:custom_instructions].present?

18
lib/ai_helper/assistant.rb
View File

@ -13,6 +13,20 @@ module DiscourseAi
prompt_cache.flush!
end
def initialize(helper_llm: nil, image_caption_llm: nil)
@helper_llm = helper_llm
@image_caption_llm = image_caption_llm
end
def helper_llm
@helper_llm || DiscourseAi::Completions::Llm.proxy(SiteSetting.ai_helper_model)
end
def image_caption_llm
@image_caption_llm ||
DiscourseAi::Completions::Llm.proxy(SiteSetting.ai_helper_image_caption_model)
end
def available_prompts(user)
key = "prompt_cache_#{I18n.locale}"
self
@ -115,7 +129,7 @@ module DiscourseAi
end
def generate_prompt(completion_prompt, input, user, force_default_locale = false, &block)
llm = DiscourseAi::Completions::Llm.proxy(SiteSetting.ai_helper_model)
llm = helper_llm
prompt = completion_prompt.messages_with_input(input)
localize_prompt!(prompt, user, force_default_locale)
@ -182,7 +196,7 @@ module DiscourseAi
)
raw_caption =
DiscourseAi::Completions::Llm.proxy(SiteSetting.ai_helper_image_caption_model).generate(
image_caption_llm.generate(
prompt,
user: user,
max_tokens: 1024,

50
lib/completions/endpoints/base.rb
View File

@ -156,7 +156,7 @@ module DiscourseAi
)
if !@streaming_mode
return(
response_data =
non_streaming_response(
response: response,
xml_tool_processor: xml_tool_processor,
@ -164,7 +164,7 @@ module DiscourseAi
partials_raw: partials_raw,
response_raw: response_raw,
)
)
return response_data
end
begin
@ -223,10 +223,46 @@ module DiscourseAi
log.duration_msecs = (Time.now - start_time) * 1000
log.save!
LlmQuota.log_usage(@llm_model, user, log.request_tokens, log.response_tokens)
if Rails.env.development?
if Rails.env.development? && !ENV["DISCOURSE_AI_NO_DEBUG"]
puts "#{self.class.name}: request_tokens #{log.request_tokens} response_tokens #{log.response_tokens}"
end
end
if log && (logger = Thread.current[:llm_audit_log])
call_data = <<~LOG
#{self.class.name}: request_tokens #{log.request_tokens} response_tokens #{log.response_tokens}
request:
#{format_possible_json_payload(log.raw_request_payload)}
response:
#{response_data}
LOG
logger.info(call_data)
end
if log && (structured_logger = Thread.current[:llm_audit_structured_log])
llm_request =
begin
JSON.parse(log.raw_request_payload)
rescue StandardError
log.raw_request_payload
end
# gemini puts passwords in query params
# we don't want to log that
structured_logger.log(
"llm_call",
args: {
class: self.class.name,
completion_url: request.uri.to_s.split("?")[0],
request: llm_request,
result: response_data,
request_tokens: log.request_tokens,
response_tokens: log.response_tokens,
duration: log.duration_msecs,
stream: @streaming_mode,
},
start_time: start_time.utc,
end_time: Time.now.utc,
)
end
end
end
end
@ -298,6 +334,14 @@ module DiscourseAi
private
def format_possible_json_payload(payload)
begin
JSON.pretty_generate(JSON.parse(payload))
rescue JSON::ParserError
payload
end
end
def start_log(
provider_id:,
request_body:,

27
lib/configuration/llm_enumerator.rb
View File

@ -16,11 +16,8 @@ module DiscourseAi
AiPersona
.where("force_default_llm = ?", true)
.pluck(:default_llm, :name, :id)
.each do |llm_name, name, id|
llm_id = llm_name.split(":").last.to_i
rval[llm_id] << { type: :ai_persona, name: name, id: id }
end
.pluck(:default_llm_id, :name, :id)
.each { |llm_id, name, id| rval[llm_id] << { type: :ai_persona, name: name, id: id } }
end
if SiteSetting.ai_helper_enabled
@ -50,6 +47,26 @@ module DiscourseAi
true
end
# returns an array of hashes (id: , name:, vision_enabled:)
def self.values_for_serialization(allowed_seeded_llm_ids: nil)
builder = DB.build(<<~SQL)
SELECT id, display_name AS name, vision_enabled
FROM llm_models
/*where*/
SQL
if allowed_seeded_llm_ids.is_a?(Array) && !allowed_seeded_llm_ids.empty?
builder.where(
"id > 0 OR id IN (:allowed_seeded_llm_ids)",
allowed_seeded_llm_ids: allowed_seeded_llm_ids,
)
else
builder.where("id > 0")
end
builder.query_hash.map(&:symbolize_keys)
end
def self.values(allowed_seeded_llms: nil)
values = DB.query_hash(<<~SQL).map(&:symbolize_keys)
SELECT display_name AS name, id AS value

2
lib/discord/bot/persona_replier.rb
View File

@ -13,7 +13,7 @@ module DiscourseAi
DiscourseAi::AiBot::Bot.as(
Discourse.system_user,
persona: @persona,
model: @persona.class.default_llm,
model: LlmModel.find(@persona.class.default_llm_id),
)
super(body)
end

213
lib/utils/image_to_text.rb Normal file
View File

@ -0,0 +1,213 @@
# frozen_string_literal: true
class DiscourseAi::Utils::ImageToText
BACKOFF_SECONDS = [5, 30, 60]
MAX_IMAGE_SIZE = 10.megabytes
class Reader
def initialize(uploads:, llm_model:, user:)
@uploads = uploads
@llm_model = llm_model
@user = user
@buffer = +""
@to_process = uploads.dup
end
# return nil if no more data
def read(length)
# for implementation simplicity we will process one image at a time
if !@buffer.empty?
part = @buffer.slice!(0, length)
return part
end
return nil if @to_process.empty?
upload = @to_process.shift
extractor =
DiscourseAi::Utils::ImageToText.new(upload: upload, llm_model: @llm_model, user: @user)
extractor.extract_text do |chunk, error|
if error
Discourse.warn_exception(
error,
message: "Discourse AI: Failed to extract text from image",
)
else
# this introduces chunk markers so discourse rag ingestion requires no overlaps
@buffer << "\n[[metadata ]]\n"
@buffer << chunk
end
end
read(length)
end
end
def self.as_fake_file(uploads:, llm_model:, user:)
# given our implementation for extracting text expect a file, return a simple object that can simulate read(size)
# and stream content
Reader.new(uploads: uploads, llm_model: llm_model, user: user)
end
attr_reader :upload, :llm_model, :user
def initialize(upload:, llm_model:, user:)
@upload = upload
@llm_model = llm_model
@user = user
end
def extract_text(retries: 3)
uploads ||= @uploaded_pages
raise "must specify a block" if !block_given?
extracted = nil
error = nil
backoff = BACKOFF_SECONDS.dup
retries.times do
seconds = nil
begin
extracted = extract_text_from_page(upload)
break
rescue => e
error = e
seconds = backoff.shift || seconds
sleep(seconds)
end
end
if extracted
extracted.each { |chunk| yield(chunk) }
else
yield(nil, error)
end
extracted || []
end
private
def system_message
<<~MSG
OCR the following page into Markdown. Tables should be formatted as Github flavored markdown.
Do not surround your output with triple backticks.
Chunk the document into sections of roughly 250 - 1000 words. Our goal is to identify parts of the page with same semantic theme. These chunks will be embedded and used in a RAG pipeline.
Always prefer returning text in Markdown vs HTML.
Describe all the images and graphs you encounter.
Only return text that will assist in the querying of data. Omit text such as "I had trouble recognizing images" and so on.
Surround the chunks with <chunk> </chunk> html tags.
MSG
end
def extract_text_from_page(page)
raw_text = extract_text_with_tesseract(page)
llm = llm_model.to_llm
if raw_text.present?
messages = [
{
type: :user,
content:
"The following text was extracted from an image using OCR. Please enhance, correct, and structure this content while maintaining the original meaning:\n\n#{raw_text}",
upload_ids: [page.id],
},
]
else
messages = [
{ type: :user, content: "Please OCR the content in the image.", upload_ids: [page.id] },
]
end
prompt = DiscourseAi::Completions::Prompt.new(system_message, messages: messages)
result = llm.generate(prompt, user: Discourse.system_user)
extract_chunks(result)
end
def extract_text_with_tesseract(page)
upload_path =
if page.local?
Discourse.store.path_for(page)
else
Discourse.store.download_safe(page, max_file_size_kb: MAX_IMAGE_SIZE)&.path
end
return "" if !upload_path || !File.exist?(upload_path)
tmp_output_file = Tempfile.new(%w[tesseract_output .txt])
tmp_output = tmp_output_file.path
tmp_output_file.unlink
command = [
"tesseract",
upload_path,
tmp_output.sub(/\.txt$/, ""), # Tesseract adds .txt automatically
]
success =
Discourse::Utils.execute_command(
*command,
timeout: 20.seconds,
failure_message: "Failed to OCR image with Tesseract",
)
if success && File.exist?("#{tmp_output}")
text = File.read("#{tmp_output}")
begin
File.delete("#{tmp_output}")
rescue StandardError
nil
end
text.strip
else
Rails.logger.error("Tesseract OCR failed for #{upload_path}")
""
end
rescue => e
Rails.logger.error("Error during OCR processing: #{e.message}")
""
end
def extract_chunks(text)
return [] if text.nil? || text.empty?
if text.include?("<chunk>") && text.include?("</chunk>")
chunks = []
remaining_text = text.dup
while remaining_text.length > 0
if remaining_text.start_with?("<chunk>")
# Extract chunk content
chunk_end = remaining_text.index("</chunk>")
if chunk_end
chunk = remaining_text[7..chunk_end - 1].strip
chunks << chunk unless chunk.empty?
remaining_text = remaining_text[chunk_end + 8..-1] || ""
else
# Malformed chunk - add remaining text and break
chunks << remaining_text[7..-1].strip
break
end
else
# Handle text before next chunk if it exists
next_chunk = remaining_text.index("<chunk>")
if next_chunk
text_before = remaining_text[0...next_chunk].strip
chunks << text_before unless text_before.empty?
remaining_text = remaining_text[next_chunk..-1]
else
# No more chunks - add remaining text and break
chunks << remaining_text.strip
break
end
end
end
return chunks.reject(&:empty?)
end
[text]
end
end

80
lib/utils/pdf_to_images.rb Normal file
View File

@ -0,0 +1,80 @@
# frozen_string_literal: true
class DiscourseAi::Utils::PdfToImages
MAX_PDF_SIZE = 100.megabytes
# this is long, mutool can be faster than magick, 10 minutes will be enough for quite large pdfs
MAX_CONVERT_SECONDS = 600
BACKOFF_SECONDS = [5, 30, 60]
attr_reader :upload, :user
def initialize(upload:, user:)
@upload = upload
@user = user
@uploaded_pages = UploadReference.where(target: upload).map(&:upload).presence
end
def uploaded_pages
@uploaded_pages ||= extract_pages
end
def extract_pages
Dir.mktmpdir("discourse-pdf-#{SecureRandom.hex(8)}")
begin
pdf_path =
if upload.local?
Discourse.store.path_for(upload)
else
Discourse.store.download_safe(upload, max_file_size_kb: MAX_PDF_SIZE)&.path
end
raise Discourse::InvalidParameters.new("Failed to download PDF") if pdf_path.nil?
temp_pdf = File.join(temp_dir, "source.pdf")
FileUtils.cp(pdf_path, temp_pdf)
# Convert PDF to individual page images
output_pattern = File.join(temp_dir, "page-%04d.png")
command = [
"magick",
"-density",
"300",
temp_pdf,
"-background",
"white",
"-auto-orient",
"-quality",
"85",
output_pattern,
]
Discourse::Utils.execute_command(
*command,
failure_message: "Failed to convert PDF to images",
timeout: MAX_CONVERT_SECONDS,
)
uploads = []
Dir
.glob(File.join(temp_dir, "page-*.png"))
.sort
.each do |page_path|
upload =
UploadCreator.new(File.open(page_path), "page-#{File.basename(page_path)}").create_for(
@user.id,
)
uploads << upload
end
# Create upload references
UploadReference.ensure_exist!(upload_ids: uploads.map(&:id), target: @upload)
@uploaded_pages = uploads
ensure
FileUtils.rm_rf(temp_dir)
end
end
end

33
spec/configuration/llm_enumerator_spec.rb
View File

@ -2,6 +2,39 @@
RSpec.describe DiscourseAi::Configuration::LlmEnumerator do
fab!(:fake_model)
fab!(:llm_model)
fab!(:seeded_model)
describe "#values_for_serialization" do
it "returns an array for that can be used for serialization" do
fake_model.destroy!
expect(described_class.values_for_serialization).to eq(
[
{
id: llm_model.id,
name: llm_model.display_name,
vision_enabled: llm_model.vision_enabled,
},
],
)
expect(
described_class.values_for_serialization(allowed_seeded_llm_ids: [seeded_model.id.to_s]),
).to contain_exactly(
{
id: seeded_model.id,
name: seeded_model.display_name,
vision_enabled: seeded_model.vision_enabled,
},
{
id: llm_model.id,
name: llm_model.display_name,
vision_enabled: llm_model.vision_enabled,
},
)
end
end
describe "#global_usage" do
before do

17
spec/jobs/regular/digest_rag_upload_spec.rb
View File

@ -2,8 +2,8 @@
RSpec.describe Jobs::DigestRagUpload do
fab!(:persona) { Fabricate(:ai_persona) }
fab!(:upload)
fab!(:upload) { Fabricate(:upload, extension: "txt") }
fab!(:pdf_upload) { Fabricate(:upload, extension: "pdf") }
let(:document_file) { StringIO.new("some text" * 200) }
fab!(:cloudflare_embedding_def)
@ -31,6 +31,19 @@ RSpec.describe Jobs::DigestRagUpload do
end
describe "#execute" do
context "when processing a PDF upload" do
it "will reject the indexing if the site setting is not enabled" do
SiteSetting.ai_rag_pdf_images_enabled = false
expect {
described_class.new.execute(
upload_id: pdf_upload.id,
target_id: persona.id,
target_type: persona.class.to_s,
)
}.to raise_error(Discourse::InvalidAccess)
end
end
context "when processing an upload containing metadata" do
it "correctly splits on metadata boundary" do
# be explicit here about chunking strategy

5
spec/jobs/regular/stream_discord_reply_spec.rb
View File

@ -13,9 +13,12 @@ RSpec.describe Jobs::StreamDiscordReply, type: :job do
}.to_json.to_s
end
fab!(:llm_model)
fab!(:persona) { Fabricate(:ai_persona, default_llm_id: llm_model.id) }
before do
SiteSetting.ai_discord_search_mode = "persona"
SiteSetting.ai_discord_search_persona = -1
SiteSetting.ai_discord_search_persona = persona.id
end
it "calls PersonaReplier when search mode is persona" do

5
spec/lib/discord/bot/persona_replier_spec.rb
View File

@ -8,8 +8,11 @@ RSpec.describe DiscourseAi::Discord::Bot::PersonaReplier do
end
let(:persona_replier) { described_class.new(interaction_body) }
fab!(:llm_model)
fab!(:persona) { Fabricate(:ai_persona, default_llm_id: llm_model.id) }
before do
SiteSetting.ai_discord_search_persona = "-1"
SiteSetting.ai_discord_search_persona = persona.id.to_s
allow_any_instance_of(DiscourseAi::AiBot::Bot).to receive(:reply).and_return(
"This is a reply from bot!",
)

2
spec/lib/modules/ai_bot/entry_point_spec.rb
View File

@ -43,7 +43,7 @@ RSpec.describe DiscourseAi::AiBot::EntryPoint do
:ai_persona,
enabled: true,
allowed_group_ids: [bot_allowed_group.id],
default_llm: "claude-2",
default_llm_id: claude_2.id,
force_default_llm: true,
)
persona.create_user!

3
spec/lib/modules/ai_bot/personas/persona_spec.rb
View File

@ -8,7 +8,6 @@ class TestPersona < DiscourseAi::AiBot::Personas::Persona
DiscourseAi::AiBot::Tools::Image,
]
end
def system_prompt
<<~PROMPT
{site_url}
@ -337,7 +336,7 @@ RSpec.describe DiscourseAi::AiBot::Personas::Persona do
name: "custom",
rag_conversation_chunks: 3,
allowed_group_ids: [Group::AUTO_GROUPS[:trust_level_0]],
question_consolidator_llm: "custom:#{llm_model.id}",
question_consolidator_llm_id: llm_model.id,
)
UploadReference.ensure_exist!(target: custom_ai_persona, upload_ids: [upload.id])

35
spec/lib/modules/ai_bot/playground_spec.rb
View File

@ -241,7 +241,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
system_prompt: "You are a helpful bot",
vision_enabled: true,
vision_max_pixels: 1_000,
default_llm: "custom:#{opus_model.id}",
default_llm_id: opus_model.id,
allow_topic_mentions: true,
)
end
@ -293,7 +293,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
persona.create_user!
persona.update!(
default_llm: "custom:#{claude_2.id}",
default_llm_id: claude_2.id,
allow_chat_channel_mentions: true,
allow_topic_mentions: true,
)
@ -313,7 +313,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
SiteSetting.ai_bot_enabled = true
SiteSetting.chat_allowed_groups = "#{Group::AUTO_GROUPS[:trust_level_0]}"
Group.refresh_automatic_groups!
persona.update!(allow_chat_channel_mentions: true, default_llm: "custom:#{opus_model.id}")
persona.update!(allow_chat_channel_mentions: true, default_llm_id: opus_model.id)
end
it "should behave in a sane way when threading is enabled" do
@ -428,7 +428,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
allow_chat_direct_messages: true,
allow_topic_mentions: false,
allow_chat_channel_mentions: false,
default_llm: "custom:#{opus_model.id}",
default_llm_id: opus_model.id,
)
SiteSetting.ai_bot_enabled = true
end
@ -629,7 +629,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
post = nil
DiscourseAi::Completions::Llm.with_prepared_responses(
["Yes I can", "Magic Title"],
llm: "custom:#{claude_2.id}",
llm: claude_2,
) do
post =
create_post(
@ -648,10 +648,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
llm2.toggle_companion_user
DiscourseAi::Completions::Llm.with_prepared_responses(
["Hi from bot two"],
llm: "custom:#{llm2.id}",
) do
DiscourseAi::Completions::Llm.with_prepared_responses(["Hi from bot two"], llm: llm2) do
create_post(
user: admin,
raw: "hi @#{llm2.user.username.capitalize} how are you",
@ -664,12 +661,9 @@ RSpec.describe DiscourseAi::AiBot::Playground do
expect(last_post.user_id).to eq(persona.user_id)
# tether llm, so it can no longer be switched
persona.update!(force_default_llm: true, default_llm: "custom:#{claude_2.id}")
persona.update!(force_default_llm: true, default_llm_id: claude_2.id)
DiscourseAi::Completions::Llm.with_prepared_responses(
["Hi from bot one"],
llm: "custom:#{claude_2.id}",
) do
DiscourseAi::Completions::Llm.with_prepared_responses(["Hi from bot one"], llm: claude_2) do
create_post(
user: admin,
raw: "hi @#{llm2.user.username.capitalize} how are you",
@ -689,7 +683,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
DiscourseAi::Completions::Llm.with_prepared_responses(
["Yes I can", "Magic Title"],
llm: "custom:#{claude_2.id}",
llm: claude_2,
) do
post =
create_post(
@ -731,11 +725,11 @@ RSpec.describe DiscourseAi::AiBot::Playground do
toggle_enabled_bots(bots: [gpt_35_turbo, claude_2])
post = nil
persona.update!(force_default_llm: true, default_llm: "custom:#{gpt_35_turbo.id}")
persona.update!(force_default_llm: true, default_llm_id: gpt_35_turbo.id)
DiscourseAi::Completions::Llm.with_prepared_responses(
["Yes I can", "Magic Title"],
llm: "custom:#{gpt_35_turbo.id}",
llm: gpt_35_turbo,
) do
post =
create_post(
@ -768,7 +762,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
DiscourseAi::Completions::Llm.with_prepared_responses(
["Yes I can", "Magic Title"],
llm: "custom:#{gpt_35_turbo.id}",
llm: gpt_35_turbo,
) do
messages =
MessageBus.track_publish do
@ -805,10 +799,7 @@ RSpec.describe DiscourseAi::AiBot::Playground do
)
# replies as correct persona if replying direct to persona
DiscourseAi::Completions::Llm.with_prepared_responses(
["Another reply"],
llm: "custom:#{gpt_35_turbo.id}",
) do
DiscourseAi::Completions::Llm.with_prepared_responses(["Another reply"], llm: gpt_35_turbo) do
create_post(
raw: "Please ignore this bot, I am replying to a user",
topic: post.topic,

14
spec/models/ai_persona_spec.rb
View File

@ -1,6 +1,9 @@
# frozen_string_literal: true
RSpec.describe AiPersona do
fab!(:llm_model)
fab!(:seeded_llm_model) { Fabricate(:llm_model, id: -1) }
it "validates context settings" do
persona =
AiPersona.new(
@ -118,7 +121,7 @@ RSpec.describe AiPersona do
forum_helper = AiPersona.find_by(name: "Forum Helper")
forum_helper.update!(
user_id: 1,
default_llm: "anthropic:claude-2",
default_llm_id: llm_model.id,
max_context_posts: 3,
allow_topic_mentions: true,
allow_personal_messages: true,
@ -133,7 +136,7 @@ RSpec.describe AiPersona do
# tl 0 by default
expect(klass.allowed_group_ids).to eq([10])
expect(klass.user_id).to eq(1)
expect(klass.default_llm).to eq("anthropic:claude-2")
expect(klass.default_llm_id).to eq(llm_model.id)
expect(klass.max_context_posts).to eq(3)
expect(klass.allow_topic_mentions).to eq(true)
expect(klass.allow_personal_messages).to eq(true)
@ -149,7 +152,7 @@ RSpec.describe AiPersona do
system_prompt: "test",
tools: [],
allowed_group_ids: [],
default_llm: "anthropic:claude-2",
default_llm_id: llm_model.id,
max_context_posts: 3,
allow_topic_mentions: true,
allow_personal_messages: true,
@ -164,7 +167,7 @@ RSpec.describe AiPersona do
expect(klass.system).to eq(false)
expect(klass.allowed_group_ids).to eq([])
expect(klass.user_id).to eq(1)
expect(klass.default_llm).to eq("anthropic:claude-2")
expect(klass.default_llm_id).to eq(llm_model.id)
expect(klass.max_context_posts).to eq(3)
expect(klass.allow_topic_mentions).to eq(true)
expect(klass.allow_personal_messages).to eq(true)
@ -227,10 +230,9 @@ RSpec.describe AiPersona do
system_prompt: "test",
tools: [],
allowed_group_ids: [],
default_llm: "seeded_model:-1",
default_llm_id: seeded_llm_model.id,
)
llm_model = Fabricate(:llm_model, id: -1)
SiteSetting.ai_bot_allowed_seeded_models = ""
expect(persona.valid?).to eq(false)

2
spec/requests/admin/ai_llms_controller_spec.rb
View File

@ -16,7 +16,7 @@ RSpec.describe DiscourseAi::Admin::AiLlmsController do
:ai_persona,
name: "Cool persona",
force_default_llm: true,
default_llm: "custom:#{llm_model2.id}",
default_llm_id: llm_model2.id,
)
end

34
spec/requests/admin/ai_personas_controller_spec.rb
View File

@ -4,6 +4,7 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
fab!(:admin)
fab!(:ai_persona)
fab!(:embedding_definition)
fab!(:llm_model)
before do
sign_in(admin)
@ -27,9 +28,13 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
expect(response).to be_successful
expect(response.parsed_body["meta"]["llms"]).to eq(
DiscourseAi::Configuration::LlmEnumerator.values.map do |hash|
{ "id" => hash[:value], "name" => hash[:name] }
end,
[
{
id: llm_model.id,
name: llm_model.display_name,
vision_enabled: llm_model.vision_enabled,
}.stringify_keys,
],
)
end
@ -44,7 +49,8 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
allow_personal_messages: true,
allow_chat_channel_mentions: true,
allow_chat_direct_messages: true,
default_llm: "anthropic:claude-2",
default_llm_id: llm_model.id,
question_consolidator_llm_id: llm_model.id,
forced_tool_count: 2,
)
persona2.create_user!
@ -60,7 +66,8 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
expect(serializer_persona2["allow_chat_channel_mentions"]).to eq(true)
expect(serializer_persona2["allow_chat_direct_messages"]).to eq(true)
expect(serializer_persona2["default_llm"]).to eq("anthropic:claude-2")
expect(serializer_persona2["default_llm_id"]).to eq(llm_model.id)
expect(serializer_persona2["question_consolidator_llm_id"]).to eq(llm_model.id)
expect(serializer_persona2["user_id"]).to eq(persona2.user_id)
expect(serializer_persona2["user"]["id"]).to eq(persona2.user_id)
expect(serializer_persona2["forced_tool_count"]).to eq(2)
@ -178,7 +185,8 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
allow_personal_messages: true,
allow_chat_channel_mentions: true,
allow_chat_direct_messages: true,
default_llm: "anthropic:claude-2",
default_llm_id: llm_model.id,
question_consolidator_llm_id: llm_model.id,
forced_tool_count: 2,
}
end
@ -190,18 +198,20 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
headers: {
"CONTENT_TYPE" => "application/json",
}
expect(response).to be_successful
persona_json = response.parsed_body["ai_persona"]
expect(persona_json["name"]).to eq("superbot")
expect(persona_json["top_p"]).to eq(0.1)
expect(persona_json["temperature"]).to eq(0.5)
expect(persona_json["default_llm"]).to eq("anthropic:claude-2")
expect(persona_json["default_llm_id"]).to eq(llm_model.id)
expect(persona_json["forced_tool_count"]).to eq(2)
expect(persona_json["allow_topic_mentions"]).to eq(true)
expect(persona_json["allow_personal_messages"]).to eq(true)
expect(persona_json["allow_chat_channel_mentions"]).to eq(true)
expect(persona_json["allow_chat_direct_messages"]).to eq(true)
expect(persona_json["question_consolidator_llm_id"]).to eq(llm_model.id)
persona = AiPersona.find(persona_json["id"])
@ -258,6 +268,8 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
rag_chunk_tokens: "102",
rag_chunk_overlap_tokens: "12",
rag_conversation_chunks: "13",
rag_llm_model_id: llm_model.id,
question_consolidator_llm_id: llm_model.id,
},
}
@ -267,6 +279,8 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
expect(persona.rag_chunk_tokens).to eq(102)
expect(persona.rag_chunk_overlap_tokens).to eq(12)
expect(persona.rag_conversation_chunks).to eq(13)
expect(persona.rag_llm_model_id).to eq(llm_model.id)
expect(persona.question_consolidator_llm_id).to eq(llm_model.id)
end
it "supports updating vision params" do
@ -424,7 +438,7 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
end
it "ensures question exists" do
ai_persona.update!(default_llm: "custom:#{llm.id}")
ai_persona.update!(default_llm_id: llm.id)
post "/admin/plugins/discourse-ai/ai-personas/stream-reply.json",
params: {
@ -436,7 +450,7 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
end
it "ensure persona has a user specified" do
ai_persona.update!(default_llm: "custom:#{llm.id}")
ai_persona.update!(default_llm_id: llm.id)
post "/admin/plugins/discourse-ai/ai-personas/stream-reply.json",
params: {
@ -498,7 +512,7 @@ RSpec.describe DiscourseAi::Admin::AiPersonasController do
ai_persona.create_user!
ai_persona.update!(
allowed_group_ids: [Group::AUTO_GROUPS[:trust_level_0]],
default_llm: "custom:#{llm.id}",
default_llm_id: llm.id,
allow_personal_messages: true,
system_prompt: "you are a helpful bot",
)

2
spec/requests/admin/ai_tools_controller_spec.rb
View File

@ -1,6 +1,7 @@
# frozen_string_literal: true
RSpec.describe DiscourseAi::Admin::AiToolsController do
fab!(:llm_model)
fab!(:admin)
fab!(:ai_tool) do
AiTool.create!(
@ -33,6 +34,7 @@ RSpec.describe DiscourseAi::Admin::AiToolsController do
expect(response).to be_successful
expect(response.parsed_body["ai_tools"].length).to eq(AiTool.count)
expect(response.parsed_body["meta"]["presets"].length).to be > 0
expect(response.parsed_body["meta"]["llms"].length).to be > 0
end
end

34
spec/requests/admin/rag_document_fragments_controller_spec.rb
View File

@ -12,6 +12,8 @@ RSpec.describe DiscourseAi::Admin::RagDocumentFragmentsController do
SiteSetting.ai_embeddings_enabled = true
end
after { @cleanup_files&.each(&:unlink) }
describe "GET #indexing_status_check" do
it "works for AiPersona" do
get "/admin/plugins/discourse-ai/rag-document-fragments/files/status.json?target_type=AiPersona&target_id=#{ai_persona.id}"
@ -22,6 +24,15 @@ RSpec.describe DiscourseAi::Admin::RagDocumentFragmentsController do
end
describe "POST #upload_file" do
let :fake_pdf do
@cleanup_files ||= []
tempfile = Tempfile.new(%w[test .pdf])
tempfile.write("fake pdf")
tempfile.rewind
@cleanup_files << tempfile
tempfile
end
it "works" do
post "/admin/plugins/discourse-ai/rag-document-fragments/files/upload.json",
params: {
@ -33,5 +44,28 @@ RSpec.describe DiscourseAi::Admin::RagDocumentFragmentsController do
upload = Upload.last
expect(upload.original_filename).to end_with("spec.txt")
end
it "rejects PDF files if site setting is not enabled" do
SiteSetting.ai_rag_pdf_images_enabled = false
post "/admin/plugins/discourse-ai/rag-document-fragments/files/upload.json",
params: {
file: Rack::Test::UploadedFile.new(fake_pdf),
}
expect(response.status).to eq(400)
end
it "allows PDF files if site setting is enabled" do
SiteSetting.ai_rag_pdf_images_enabled = true
post "/admin/plugins/discourse-ai/rag-document-fragments/files/upload.json",
params: {
file: Rack::Test::UploadedFile.new(fake_pdf),
}
upload = Upload.last
expect(upload.original_filename).to end_with(".pdf")
end
end
end

2
spec/system/ai_bot/ai_bot_helper_spec.rb
View File

@ -39,7 +39,7 @@ RSpec.describe "AI chat channel summarization", type: :system, js: true do
# lets disable bots but still allow 1 persona
allowed_persona.create_user!
allowed_persona.update!(default_llm: "custom:#{gpt_4.id}")
allowed_persona.update!(default_llm_id: gpt_4.id)
gpt_4.update!(enabled_chat_bot: false)
gpt_3_5_turbo.update!(enabled_chat_bot: false)

10
test/javascripts/unit/models/ai-persona-test.js
View File

@ -37,7 +37,7 @@ module("Discourse AI | Unit | Model | ai-persona", function () {
description: "Description",
top_p: 0.8,
temperature: 0.7,
default_llm: "Default LLM",
default_llm_id: 1,
force_default_llm: false,
user: null,
user_id: null,
@ -48,7 +48,8 @@ module("Discourse AI | Unit | Model | ai-persona", function () {
rag_chunk_tokens: 374,
rag_chunk_overlap_tokens: 10,
rag_conversation_chunks: 10,
question_consolidator_llm: "Question Consolidator LLM",
rag_llm_model_id: 1,
question_consolidator_llm_id: 2,
allow_chat: false,
tool_details: true,
forced_tool_count: -1,
@ -85,7 +86,7 @@ module("Discourse AI | Unit | Model | ai-persona", function () {
temperature: 0.7,
user: null,
user_id: null,
default_llm: "Default LLM",
default_llm_id: 1,
max_context_posts: 5,
vision_enabled: true,
vision_max_pixels: 100,
@ -93,7 +94,7 @@ module("Discourse AI | Unit | Model | ai-persona", function () {
rag_chunk_tokens: 374,
rag_chunk_overlap_tokens: 10,
rag_conversation_chunks: 10,
question_consolidator_llm: "Question Consolidator LLM",
question_consolidator_llm_id: 2,
allow_chat: false,
tool_details: true,
forced_tool_count: -1,
@ -102,6 +103,7 @@ module("Discourse AI | Unit | Model | ai-persona", function () {
allow_chat_channel_mentions: true,
allow_chat_direct_messages: true,
force_default_llm: false,
rag_llm_model_id: 1,
};
const aiPersona = AiPersona.create({ ...properties });