Closed nsvrana closed 3 months ago
dosubot[bot]
commented
5 months ago Hello @nsvrana! 👋 I'm Dosu, a bot designed to help you solve bugs, answer your questions, and guide you on contributing to the Dify project. While we wait for a human maintainer, I'm here to assist you. I'm taking a glance at the issue you've reported with the Preview View in Dify v0.6.3 and will provide you with a full answer shortly. Let's get this sorted out together!
crazywoola
commented
5 months ago Can you attach the workflow.yml so we can test it out if you don't mind.
nsvrana
commented
5 months ago I have stripped this of any complexity in the prompts, but it still has the same results.
app:
description: ''
icon: "\U0001F916"
icon_background: '#FFEAD5'
mode: workflow
name: Empty Test
workflow:
features:
file_upload:
image:
enabled: false
number_limits: 3
transfer_methods:
- remote_url
- local_file
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
configs: []
enabled: false
type: ''
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
sourceType: if-else
targetType: end
id: 1712592115351-end
selected: false
source: '1712592115351'
sourceHandle: 'true'
target: end
targetHandle: target
type: custom
- data:
sourceType: if-else
targetType: llm
id: 1712592115351-1712592181613
selected: false
source: '1712592115351'
sourceHandle: 'false'
target: '1712592181613'
targetHandle: target
type: custom
- data:
sourceType: llm
targetType: if-else
id: 1712591674103-1712592115351
selected: false
source: '1712591674103'
sourceHandle: source
target: '1712592115351'
targetHandle: target
type: custom
- data:
sourceType: start
targetType: llm
id: start-1712589019429
source: start
sourceHandle: source
target: '1712589019429'
targetHandle: target
type: custom
- data:
sourceType: llm
targetType: llm
id: 1712589019429-llm
source: '1712589019429'
sourceHandle: source
target: llm
targetHandle: target
type: custom
- data:
sourceType: llm
targetType: end
id: 1712592181613-1713285594878
selected: false
source: '1712592181613'
sourceHandle: source
target: '1713285594878'
targetHandle: target
type: custom
- data:
sourceType: llm
targetType: llm
id: llm-1712590128698
source: llm
sourceHandle: source
target: '1712590128698'
targetHandle: target
type: custom
- data:
sourceType: llm
targetType: llm
id: 1712590128698-1712591674103
source: '1712590128698'
sourceHandle: source
target: '1712591674103'
targetHandle: target
type: custom
nodes:
- data:
selected: true
title: START
type: start
variables:
- default: ''
description: null
hint: null
label: Who do you want to read this?
max_length: 48
options: null
required: true
type: text-input
variable: audience
- default: ''
description: null
hint: null
label: What would you like the reader to do?
max_length: 48
options: null
required: true
type: text-input
variable: desired_outcome
- default: ''
description: null
hint: null
label: How will the content be shared?
max_length: null
options:
- LinkedIn
- Twitter/X
- Homebase
- Email
- Deck Slides
- One Pager
- Internal Newsletter
required: true
type: select
variable: target_format
- default: ''
description: null
hint: null
label: What information have you already written?
max_length: null
options: null
required: true
type: paragraph
variable: existing_content
height: 168
id: start
position:
x: 80
y: 282
positionAbsolute:
x: 80
y: 282
selected: true
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: null
memory: null
model:
completion_params:
frequency_penalty: 0.5
max_tokens: 2048
presence_penalty: 0.5
stop: []
temperature: 0.2
top_p: 0.75
mode: chat
name: gpt-4
provider: azure_openai
prompt_template:
- role: user
text: '=== Input ===
Existing Content: {{#start.existing_content#}}
Audience: {{#start.audience#}}
Desired Outcome: {{#start.desired_outcome#}}
=== Input ===
Breakdown the content into the key ideas.
'
selected: false
title: Content Breakdown
type: llm
vision:
configs: null
enabled: false
variable_selector: null
height: 98
id: llm
position:
x: 778.1719349145415
y: 282
positionAbsolute:
x: 778.1719349145415
y: 282
selected: false
type: custom
width: 244
- data:
outputs:
- value_selector:
- '1712590128698'
- text
variable: text
selected: false
title: End
type: end
height: 90
id: end
position:
x: 1969.952986908449
y: 146.40342419966382
positionAbsolute:
x: 1969.952986908449
y: 146.40342419966382
selected: false
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.3
max_tokens: 512
presence_penalty: 0.2
temperature: 0.5
top_p: 0.85
mode: chat
name: gpt-4
provider: azure_openai
prompt_template:
- role: system
text: '=== Input ===
Audience: {{#start.audience#}}
Desired Outcome: {{#start.desired_outcome#}}
Target Format: {{#start.target_format#}}
Content Summary: {{#start.existing_content#}}
=== Input ===
Write an audience brief.'
selected: false
title: Audience Evaluation
type: llm
variables: []
vision:
enabled: false
height: 98
id: '1712589019429'
position:
x: 489.19861296555086
y: 282
positionAbsolute:
x: 489.19861296555086
y: 282
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.1
max_tokens: 4096
presence_penalty: 0.1
temperature: 0.8
top_p: 0.9
mode: chat
name: gpt-4
provider: azure_openai
prompt_template:
- role: system
text: '=== Brief ===
Audience Brief: {{#1712589019429.text#}}
Content Brief: {{#llm.text#}}
Desired Outcome: {{#start.desired_outcome#}}
Target Format: {{#start.target_format#}}
=== Brief ===
Please write the content based on the brief you have been given and the
brand guide.'
selected: false
title: Regular Writer
type: llm
variables: []
vision:
enabled: false
height: 98
id: '1712590128698'
position:
x: 1065.6627273915026
y: 282
positionAbsolute:
x: 1065.6627273915026
y: 282
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.3
max_tokens: 1024
presence_penalty: 0.2
temperature: 0.5
top_p: 0.85
mode: chat
name: gpt-4
provider: azure_openai
prompt_template:
- role: system
text: 'Analyze the original input and newly written content and determine
if the original meaning is maintained and if the content fulfills the
audience brief. If the content fits the brief respond with "Content complete."
=== Original Content ===
{{#start.existing_content#}}
=== Original Content ===
=== Audience Brief ===
{{#1712589019429.text#}}
=== Audience Brief ===
=== New Content ===
{{#1712590128698.text#}}
=== New Content ==='
selected: false
title: Evaluate
type: llm
variables: []
vision:
enabled: false
height: 98
id: '1712591674103'
position:
x: 1360.889268333141
y: 282
positionAbsolute:
x: 1360.889268333141
y: 282
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
conditions:
- comparison_operator: contains
id: '1712592116696'
value: Content complete
variable_selector:
- '1712591674103'
- text
desc: ''
logical_operator: and
selected: false
title: IF/ELSE
type: if-else
height: 126
id: '1712592115351'
position:
x: 1669.072006005269
y: 282
positionAbsolute:
x: 1669.072006005269
y: 282
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.3
max_tokens: 4096
presence_penalty: 0.2
temperature: 0.5
top_p: 0.85
mode: chat
name: gpt-4
provider: azure_openai
prompt_template:
- role: system
text: '=== Content ===
{{#1712590128698.text#}}
=== Content ===
=== Feedback ===
{{#1712591674103.text#}}
=== Feedback ===
Please update the content based on the feedback you have received.'
selected: false
title: Rewrite
type: llm
variables: []
vision:
enabled: false
height: 98
id: '1712592181613'
position:
x: 1969.952986908449
y: 362.84436451362143
positionAbsolute:
x: 1969.952986908449
y: 362.84436451362143
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
desc: ''
outputs:
- value_selector:
- '1712592181613'
- text
variable: text
selected: false
title: Rewritten End
type: end
height: 90
id: '1713285594878'
position:
x: 2291.626855495137
y: 362.84436451362143
positionAbsolute:
x: 2291.626855495137
y: 362.84436451362143
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
viewport:
x: -506.6831142273469
y: 266.6694411671007
zoom: 0.5It looks like the issue with the Preview View not displaying formatted markdown text and the workflow appearing to be in an unfinished state could be related to how markdown rendering and loading states are managed in Dify. Specifically, the rendering of markdown in the Preview View is handled by components designed for markdown processing, and the loading state is managed by components responsible for displaying loading animations.
To address the issue, you should investigate the following areas:
Markdown Rendering Components: Check the components responsible for markdown rendering, such as ReactMarkdown and its associated plugins. These components are crucial for converting markdown text into the formatted display you expect in the Preview View.
Loading State Management: Examine the Loading component to ensure it correctly transitions from a loading state to a loaded state once the workflow execution completes. A perpetual loading state might indicate an issue with how the loading state is managed or an error that prevents the workflow from completing.
Workflow Configuration: Since the workflow's configuration directly affects its execution, ensure that the nodes and edges are correctly set up to produce the expected output. Incorrect configurations could lead to execution issues, including the ones you're experiencing.
Error Handling and Logs: Look for any errors or exceptions that might occur during the workflow execution. Errors could prevent the workflow from completing successfully and might explain why the Preview View remains in a loading state.
If after checking these areas the issue persists, providing more specific details such as error messages from the console or logs, and steps that consistently reproduce the issue, could help further diagnose and resolve the problem.
Self Checks
Dify version
0.6.3
Cloud or Self Hosted
Self Hosted (Docker)
Steps to reproduce
Create a complex workflow which outputs text, then use that workflow via the workspace or the "preview" view. The steps to execute are shown and the "End" output is a JSON object. However, when I look in the logs of the workflow, the output is formatted text. Additionally, the workflow loading circle never stops spinning.
✔️ Expected Behavior
To see formatted markdown text.
❌ Actual Behavior
The workflow remains in an unfinished state and the product of the output is not displayed.