sdg-hub 0.1.0__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- sdg_hub/__init__.py +3 -0
- sdg_hub/_version.py +21 -0
- sdg_hub/blocks/__init__.py +36 -0
- sdg_hub/blocks/block.py +96 -0
- sdg_hub/blocks/llmblock.py +375 -0
- sdg_hub/blocks/utilblocks.py +597 -0
- sdg_hub/checkpointer.py +139 -0
- sdg_hub/configs/__init__.py +0 -0
- sdg_hub/configs/annotations/__init__.py +0 -0
- sdg_hub/configs/annotations/cot_reflection.yaml +34 -0
- sdg_hub/configs/annotations/detailed_annotations.yaml +28 -0
- sdg_hub/configs/annotations/detailed_description.yaml +10 -0
- sdg_hub/configs/annotations/detailed_description_icl.yaml +32 -0
- sdg_hub/configs/annotations/simple_annotations.yaml +9 -0
- sdg_hub/configs/knowledge/__init__.py +0 -0
- sdg_hub/configs/knowledge/atomic_facts.yaml +45 -0
- sdg_hub/configs/knowledge/auxilary_instructions.yaml +35 -0
- sdg_hub/configs/knowledge/detailed_summary.yaml +17 -0
- sdg_hub/configs/knowledge/evaluate_faithfulness.yaml +68 -0
- sdg_hub/configs/knowledge/evaluate_question.yaml +38 -0
- sdg_hub/configs/knowledge/evaluate_relevancy.yaml +85 -0
- sdg_hub/configs/knowledge/extractive_summary.yaml +17 -0
- sdg_hub/configs/knowledge/generate_code_questions_responses.yaml +39 -0
- sdg_hub/configs/knowledge/generate_questions_responses.yaml +56 -0
- sdg_hub/configs/knowledge/mcq_generation.yaml +83 -0
- sdg_hub/configs/knowledge/router.yaml +12 -0
- sdg_hub/configs/knowledge/simple_generate_qa.yaml +34 -0
- sdg_hub/configs/reasoning/__init__.py +0 -0
- sdg_hub/configs/reasoning/dynamic_cot.yaml +40 -0
- sdg_hub/configs/skills/__init__.py +0 -0
- sdg_hub/configs/skills/analyzer.yaml +48 -0
- sdg_hub/configs/skills/annotation.yaml +36 -0
- sdg_hub/configs/skills/contexts.yaml +28 -0
- sdg_hub/configs/skills/critic.yaml +60 -0
- sdg_hub/configs/skills/evaluate_freeform_pair.yaml +111 -0
- sdg_hub/configs/skills/evaluate_freeform_questions.yaml +78 -0
- sdg_hub/configs/skills/evaluate_grounded_pair.yaml +119 -0
- sdg_hub/configs/skills/evaluate_grounded_questions.yaml +51 -0
- sdg_hub/configs/skills/freeform_questions.yaml +34 -0
- sdg_hub/configs/skills/freeform_responses.yaml +39 -0
- sdg_hub/configs/skills/grounded_questions.yaml +38 -0
- sdg_hub/configs/skills/grounded_responses.yaml +59 -0
- sdg_hub/configs/skills/icl_examples/STEM.yaml +56 -0
- sdg_hub/configs/skills/icl_examples/__init__.py +0 -0
- sdg_hub/configs/skills/icl_examples/coding.yaml +97 -0
- sdg_hub/configs/skills/icl_examples/extraction.yaml +36 -0
- sdg_hub/configs/skills/icl_examples/humanities.yaml +71 -0
- sdg_hub/configs/skills/icl_examples/math.yaml +85 -0
- sdg_hub/configs/skills/icl_examples/reasoning.yaml +30 -0
- sdg_hub/configs/skills/icl_examples/roleplay.yaml +45 -0
- sdg_hub/configs/skills/icl_examples/writing.yaml +80 -0
- sdg_hub/configs/skills/judge.yaml +53 -0
- sdg_hub/configs/skills/planner.yaml +67 -0
- sdg_hub/configs/skills/respond.yaml +8 -0
- sdg_hub/configs/skills/revised_responder.yaml +78 -0
- sdg_hub/configs/skills/router.yaml +59 -0
- sdg_hub/configs/skills/simple_generate_qa_freeform.yaml +27 -0
- sdg_hub/configs/skills/simple_generate_qa_grounded.yaml +31 -0
- sdg_hub/flow.py +306 -0
- sdg_hub/flow_runner.py +204 -0
- sdg_hub/flows/generation/knowledge/mmlu_bench.yaml +13 -0
- sdg_hub/flows/generation/knowledge/simple_knowledge.yaml +12 -0
- sdg_hub/flows/generation/knowledge/synth_knowledge.yaml +89 -0
- sdg_hub/flows/generation/knowledge/synth_knowledge1.5.yaml +136 -0
- sdg_hub/flows/generation/skills/improve_responses.yaml +103 -0
- sdg_hub/flows/generation/skills/simple_freeform_skill.yaml +12 -0
- sdg_hub/flows/generation/skills/simple_grounded_skill.yaml +12 -0
- sdg_hub/flows/generation/skills/synth_grounded_skills.yaml +80 -0
- sdg_hub/flows/generation/skills/synth_skills.yaml +59 -0
- sdg_hub/logger_config.py +20 -0
- sdg_hub/pipeline.py +121 -0
- sdg_hub/prompts.py +43 -0
- sdg_hub/py.typed +0 -0
- sdg_hub/registry.py +122 -0
- sdg_hub/sdg.py +206 -0
- sdg_hub/utils/__init__.py +5 -0
- sdg_hub/utils/datautils.py +14 -0
- sdg_hub-0.1.0.dist-info/METADATA +190 -0
- sdg_hub-0.1.0.dist-info/RECORD +82 -0
- sdg_hub-0.1.0.dist-info/WHEEL +5 -0
- sdg_hub-0.1.0.dist-info/licenses/LICENSE +201 -0
- sdg_hub-0.1.0.dist-info/top_level.txt +1 -0
@@ -0,0 +1,45 @@
|
|
1
|
+
icl_query: Pretend yourself to be Elon Musk in all the following conversations. Speak like Elon Musk as much as possible. Why do we need to go to Mars?
|
2
|
+
|
3
|
+
icl_response: |
|
4
|
+
At SpaceX, we believe that humanity must become a multiplanetary species in order to ensure our long-term survival and continue to push the boundaries of what is possible. Mars is the most habitable planet in our solar system besides Earth, and it has the resources necessary to support human life. By establishing a self-sustaining civilization on Mars, we can protect humanity from the risks associated with remaining on a single planet, such as the threat of a large asteroid impact or a catastrophic global pandemic.
|
5
|
+
|
6
|
+
Additionally, settling Mars would also open up new opportunities for scientific discovery and technological advancement. It would allow us to study the planet's geology, climate, and potential for life in greater detail, and it would provide a platform for testing and developing new technologies that could be used to improve life on Earth.
|
7
|
+
|
8
|
+
Furthermore, Mars has the potential to serve as a stepping stone for further exploration of the solar system and beyond. By establishing a presence on Mars, we can gain the experience and resources necessary to eventually send humans to other planets and even other star systems.
|
9
|
+
|
10
|
+
In short, going to Mars is not just about survival, but also about expanding the horizon of human knowledge, capabilities and pushing the boundaries of what is possible.
|
11
|
+
|
12
|
+
icl_analysis: |
|
13
|
+
The task requires the AI to adopt the persona of Elon Musk and provide a response in a style reminiscent of how Elon Musk might answer the question "Why do we need to go to Mars?" The domain of this task is a combination of impersonation, space exploration, and visionary technology discourse. The specific requirements include maintaining the communication style and thought patterns of Elon Musk, focusing on his known interests in space travel, technological innovation, and human survival.
|
14
|
+
|
15
|
+
icl_rubric: |
|
16
|
+
1. Authenticity: The response should capture Elon Musk's typical speaking style, using visionary and technological language.
|
17
|
+
2. Justification: Include clear and specific reasons why Mars colonization is considered important, reflecting Musk's views on space exploration and human survival.
|
18
|
+
3. Technological Insight: Mention and briefly explain the technology that might be involved in the process of colonizing Mars, as Musk would.
|
19
|
+
4. Long-term Perspective: The response should reflect a long-term vision, similar to what Musk might project about humanity's future on Mars.
|
20
|
+
5. Engagement: The response should engage the reader and provoke thought about space exploration, much like Musk's public discussions do.
|
21
|
+
|
22
|
+
icl_critique: |
|
23
|
+
1. Authenticity: The response does a good job of mimicking Elon Musk’s visionary and technological speaking style. Terms like "multiplanetary species" and "self-sustaining civilization" are typical of Musk’s language when discussing SpaceX and Mars. The response could be improved by integrating some of Musk’s distinctive personal anecdotes or a more casual tone, which he often uses in public talks to connect with his audience.
|
24
|
+
2. Justification: This criterion is met adequately. The response provides a comprehensive rationale for why Mars colonization is important, covering survival from catastrophic events and the broadening of human knowledge. It effectively reflects Musk's views on the necessity of expanding beyond Earth to ensure the longevity of humanity.
|
25
|
+
3. Technological Insight: The response briefly touches on the technology needed for Mars colonization by mentioning the establishment of a civilization and further exploration. However, it lacks depth in explaining specific technologies like SpaceX’s Starship, which is central to Musk’s plan for Mars. Including such details would enhance the response's authenticity and informativeness.
|
26
|
+
4. Long-term Perspective: The response excellently addresses the long-term implications of Mars colonization, discussing the role of Mars as a "stepping stone" for further space exploration. It aligns with Musk's vision of the future, where humans explore and possibly inhabit other star systems.
|
27
|
+
5. Engagement: The response is engaging and effectively provokes thought about the broader implications of space exploration. It successfully mirrors Musk’s ability to inspire and excite about future possibilities. However, incorporating a more dynamic or dramatic element, as Musk often does, could make the response even more captivating.
|
28
|
+
|
29
|
+
Overall, the response is strong and aligns well with the rubric. Minor enhancements in the areas of technological specifics and a more personalized speaking style could elevate its authenticity and engagement.
|
30
|
+
|
31
|
+
icl_plan: |
|
32
|
+
1. Integrate Personal Anecdotes or Casual Tone: Add a personal anecdote related to Elon Musk’s experiences or motivations for Mars exploration to personalize the response.
|
33
|
+
2. Expand Technological Insight: Detail the SpaceX Starship’s role and capabilities as it pertains to the Mars colonization effort.
|
34
|
+
3. Enhance Engagement with Dramatic Elements: Insert a provocative question or statement from Musk to intensify reader engagement and highlight the dramatic importance of Mars colonization.
|
35
|
+
4. Review and Revise for Consistency: Ensure the new content retains Musk’s visionary language and seamlessly integrates with the existing response.
|
36
|
+
5. Final Check for Authenticity: Confirm that the tone and content align closely with Musk’s known public communications.
|
37
|
+
|
38
|
+
icl_revised_response: |
|
39
|
+
At SpaceX, we're not just launching rockets; we're launching a new era of human civilization. Mars, the next frontier, is the only viable option for a self-sustaining civilization outside Earth. It's about ensuring our survival through becoming a multiplanetary species. Imagine—Mars has the necessary resources to support human life, which is pivotal as we face potential catastrophic events like asteroid impacts or global pandemics here on Earth.
|
40
|
+
|
41
|
+
The technology driving this monumental effort is embodied in our Starship. This spacecraft is not only designed to carry humans to Mars but to eventually enable millions to live there. It represents a leap forward in technology, pushing the boundaries of what's possible and reducing the cost of space travel.
|
42
|
+
|
43
|
+
Why Mars? Well, I often think about where humanity would be if the ancient explorers had feared the vast oceans. Mars is our modern ocean, and by crossing it, we learn, we adapt, and we evolve. Establishing a base there paves the way for unimaginable scientific discoveries and technological advancements. It's about laying the groundwork for future explorations to even more distant worlds.
|
44
|
+
|
45
|
+
This isn't just about survival. It's about thriving, about daring to dream big and tackle the impossible. So, I ask you—why shouldn't we reach for Mars when our very future may depend on it?
|
@@ -0,0 +1,80 @@
|
|
1
|
+
icl_query: |
|
2
|
+
Compose an engaging travel blog post about a recent trip to Hawaii, highlighting cultural experiences and must-see attractions.
|
3
|
+
|
4
|
+
icl_response: |
|
5
|
+
Aloha, fellow wanderers!
|
6
|
+
|
7
|
+
I've just returned from a breathtaking journey to the enchanting islands of Hawaii, and I can't wait to share my experiences with you. This tropical paradise is not only about sun-kissed beaches and turquoise waters; it's a cultural melting pot that offers an unforgettable blend of history, tradition, and natural beauty.
|
8
|
+
|
9
|
+
My adventure began on the vibrant island of Oahu, where I was immediately captivated by the unique fusion of Asian, American, and Polynesian influences. The island's capital, Honolulu, is a bustling city that seamlessly merges urban sophistication with tropical ease. A visit to the historic Pearl Harbor and the USS Arizona Memorial was a humbling experience, providing insight into America's past and reminding us all of the importance of peace.
|
10
|
+
|
11
|
+
No trip to Oahu would be complete without exploring the iconic Waikiki Beach. While many flock here for the surfing waves, I found solace in the early morning calm, watching the sun rise over Diamond Head while practicing yoga on the sand. It's moments like these that make travel so transformative - connecting with nature while surrounded by diverse people from around the globe.
|
12
|
+
|
13
|
+
From Oahu, I hopped over to Maui, known as the "Valley Isle." Here, I discovered the rich agricultural heritage at the Maui Tropical Plantation. A tram tour guided me through fields of pineapple, sugarcane, and coffee, offering delicious samples along the way. But it was the 'Ulu (breadfruit) tree that truly stole my heart. Learning about its significance in Hawaiian culture and cuisine was a delightful discovery.
|
14
|
+
|
15
|
+
Maui's Haleakala National Park was another highlight. Witnessing the sunrise from the summit of Haleakala volcano is an awe-inspiring sight. As the first light illuminated the sky, casting shadows on the crater below, I felt a deep connection to the earth and its ancient wonders.
|
16
|
+
|
17
|
+
The Road to Hana, a scenic drive with 600 curves and 54 bridges, led me to the stunning Seven Sacred Pools of Ohe'o. The cascading waterfalls and crystal-clear pools are a testament to Hawaii's raw beauty. A dip in the cool waters was the perfect way to refresh after the winding journey.
|
18
|
+
|
19
|
+
My final stop was the Big Island, where I marveled at the majestic Hawaii Volcanoes National Park. Walking through the Thurston Lava Tube and observing the active Kilauea volcano reminded me of the powerful forces that shaped these islands millions of years ago.
|
20
|
+
|
21
|
+
Beyond the natural wonders, the Big Island offers a rich cultural experience at the annual Merrie Monarch Festival, a week-long celebration of Hawaiian music, dance, and history. Here, I had the privilege of watching hula dancers pay tribute to their ancestors through graceful movements and chants. It was a moving display of tradition and reverence for the past.
|
22
|
+
|
23
|
+
Hawaii has so much more to offer than just postcard-perfect beaches. Its vibrant culture, rich history, and stunning landscapes make it a must-visit destination for any curious traveler. So pack your bags, leave your worries behind, and come experience the magic of the Hawaiian Islands.
|
24
|
+
|
25
|
+
Until next time, happy wandering!
|
26
|
+
|
27
|
+
Mahalo,
|
28
|
+
Your Traveling Companion
|
29
|
+
|
30
|
+
icl_analysis: |
|
31
|
+
The task involves creating a travel blog post focused on a recent trip to Hawaii. The domain is travel and culture. The user specifically requests a focus on cultural experiences and must-see attractions in Hawaii, indicating a desire for rich descriptive content and possibly personal anecdotes that enhance the travel experience. Keywords such as "engaging," "cultural experiences," and "must-see attractions" emphasize the need for an immersive and informative narrative.
|
32
|
+
|
33
|
+
icl_rubric: |
|
34
|
+
1. Engaging Introduction: Does the post start with a captivating introduction that draws readers in and sets the tone for the travel experience?
|
35
|
+
2. Cultural Experiences: Are the cultural experiences in Hawaii described in detail, including interactions with locals, traditional practices, and cultural insights?
|
36
|
+
3. Must-See Attractions: Are the must-see attractions in Hawaii clearly highlighted and vividly described, including specific details that make them appealing to travelers?
|
37
|
+
4. Visual and Sensory Details: Does the post include rich visual and sensory details that help readers visualize the locations and experiences mentioned?
|
38
|
+
5. Personal Reflections: Are personal reflections and anecdotes included to provide a unique perspective and make the post more relatable and engaging?
|
39
|
+
|
40
|
+
icl_critique: |
|
41
|
+
1. Engaging Introduction: The introduction successfully sets a captivating tone with phrases like "breathtaking journey" and "tropical paradise," which immediately draw readers in. It effectively balances excitement with informative content, inviting readers to explore more about the cultural and natural beauty of Hawaii.
|
42
|
+
2. Cultural Experiences: The description of cultural experiences is well-executed, with detailed accounts of the Merrie Monarch Festival, traditional agricultural practices at the Maui Tropical Plantation, and the significance of 'Ulu in Hawaiian culture. These segments provide a deep dive into the local lifestyle and traditions, although the post could further enhance its cultural depth by including more direct interactions with local people or insights into everyday Hawaiian life.
|
43
|
+
3. Must-See Attractions: The blog post excels in highlighting must-see attractions like Waikiki Beach, Haleakala National Park, and the Road to Hana. Each location is vividly described, with specific details that emphasize their beauty and uniqueness, such as the sunrise at Haleakala and the scenic drive on the Road to Hana. This makes them highly appealing to travelers looking for both adventure and natural beauty.
|
44
|
+
4. Visual and Sensory Details: The post is rich in visual and sensory details, effectively allowing readers to visualize the scenes described. For instance, the description of the sunrise at Haleakala provides a sensory feast, describing the light and shadows over the crater. Similarly, the tactile experience of yoga on the sands of Waikiki Beach brings the scene to life, making the experiences more tangible for the reader.
|
45
|
+
5. Personal Reflections: Personal reflections are seamlessly integrated into the narrative, enhancing the relatability and depth of the blog. The personal anecdotes, such as practicing yoga on the beach and the feelings evoked by visiting Pearl Harbor, add a personal touch that helps readers connect on an emotional level.
|
46
|
+
|
47
|
+
Overall, the response is highly effective in fulfilling the rubric criteria. However, to further enrich the blog, additional stories of interactions with locals or more detailed personal anecdotes could provide deeper cultural insights and enhance the reader's engagement.
|
48
|
+
|
49
|
+
icl_plan: |
|
50
|
+
1. Integrate More Local Interactions: Add a detailed encounter with a local expert, such as a Hawaiian historian or a native craftsman, to provide a deeper understanding of Hawaiian culture and traditions.
|
51
|
+
2. Expand Personal Anecdotes: Incorporate a specific personal story, like experiencing a local festival or a unique encounter, which highlights emotional and cultural connections.
|
52
|
+
3. Enhance Descriptions of Scenic Locations: Include more sensory details for key attractions, like the textures and sounds experienced while exploring the Haleakala National Park or local markets.
|
53
|
+
4. Deepen Cultural Descriptions: Elaborate on the historical significance of a specific site visited, incorporating cultural myths or local stories to enrich the readers' connection to the place.
|
54
|
+
5. Incorporate Feedback Mechanisms: Encourage reader interaction by inviting them to share their own Hawaiian travel stories or tips, enhancing community engagement around the blog.
|
55
|
+
|
56
|
+
icl_revised_response: |
|
57
|
+
Aloha, fellow wanderers!
|
58
|
+
|
59
|
+
I've just returned from a breathtaking journey to the enchanting islands of Hawaii, and I can't wait to share my experiences with you. This tropical paradise is not only about sun-kissed beaches and turquoise waters; it's a cultural melting pot that offers an unforgettable blend of history, tradition, and natural beauty.
|
60
|
+
|
61
|
+
My adventure began on the vibrant island of Oahu, where I was immediately captivated by the unique fusion of Asian, American, and Polynesian influences. The island's capital, Honolulu, is a bustling city that seamlessly merges urban sophistication with tropical ease. A visit to the historic Pearl Harbor and the USS Arizona Memorial was a humbling experience, providing insight into America's past and reminding us all of the importance of peace.
|
62
|
+
|
63
|
+
No trip to Oahu would be complete without exploring the iconic Waikiki Beach. While many flock here for the surfing waves, I found solace in the early morning calm, watching the sun rise over Diamond Head while practicing yoga on the sand. It's moments like these that make travel so transformative—connecting with nature while surrounded by diverse people from around the globe.
|
64
|
+
|
65
|
+
From Oahu, I hopped over to Maui, known as the "Valley Isle." Here, I discovered the rich agricultural heritage at the Maui Tropical Plantation. A tram tour guided me through fields of pineapple, sugarcane, and coffee, offering delicious samples along the way. But it was the 'Ulu (breadfruit) tree that truly stole my heart. Learning about its significance in Hawaiian culture and cuisine was a delightful discovery.
|
66
|
+
|
67
|
+
Maui's Haleakala National Park was another highlight. Witnessing the sunrise from the summit of Haleakala volcano is an awe-inspiring sight. As the first light illuminated the sky, casting shadows on the crater below, I felt a deep connection to the earth and its ancient wonders. The sounds of the wind and the shifting shadows added a magical layer to the experience, making it truly unforgettable.
|
68
|
+
|
69
|
+
The Road to Hana, a scenic drive with 600 curves and 54 bridges, led me to the stunning Seven Sacred Pools of Ohe'o. The cascading waterfalls and crystal-clear pools are a testament to Hawaii's raw beauty. A dip in the cool waters was the perfect way to refresh after the winding journey.
|
70
|
+
|
71
|
+
My final stop was the Big Island, where I marveled at the majestic Hawaii Volcanoes National Park. Walking through the Thurston Lava Tube and observing the active Kilauea volcano reminded me of the powerful forces that shaped these islands millions of years ago.
|
72
|
+
|
73
|
+
Beyond the natural wonders, the Big Island offers a rich cultural experience at the annual Merrie Monarch Festival, a week-long celebration of Hawaiian music, dance, and history. Here, I had the privilege of meeting Kaimana, a local craftsman whose family has been making traditional leis for generations. He shared the profound significance of lei-making in Hawaiian culture, which added a personal touch to the vibrant festival atmosphere.
|
74
|
+
|
75
|
+
Hawaii has so much more to offer than just postcard-perfect beaches. Its vibrant culture, rich history, and stunning landscapes make it a must-visit destination for any curious traveler. So pack your bags, leave your worries behind, and come experience the magic of the Hawaiian Islands.
|
76
|
+
|
77
|
+
Until next time, happy wandering!
|
78
|
+
|
79
|
+
Mahalo,
|
80
|
+
Your Traveling Companion
|
@@ -0,0 +1,53 @@
|
|
1
|
+
system: You are a helpful and precise assistant for checking the quality of the answer.
|
2
|
+
|
3
|
+
introduction: Please act as an impartial judge and evaluate the quality of the responses provided by two AI assistants to the user question displayed below. You should choose the assistant that follows the user's instructions and answers the user's question better. Your evaluation should consider factors such as the helpfulness, relevance, accuracy, depth, creativity, and level of detail of their responses.
|
4
|
+
|
5
|
+
principles: |
|
6
|
+
* Begin your evaluation by comparing the two responses and provide a short judgement.
|
7
|
+
* Avoid any position biases and ensure that the order in which the responses were presented does not influence your decision.
|
8
|
+
* Do not allow the length of the responses to influence your evaluation.
|
9
|
+
* Be as objective as possible.
|
10
|
+
|
11
|
+
examples: |
|
12
|
+
To better assist you with this, heres an example:
|
13
|
+
|
14
|
+
[Start of User Question]
|
15
|
+
What are the benefits of regular exercise?
|
16
|
+
[End of User Question]
|
17
|
+
|
18
|
+
[Start of Assistant A Answer]
|
19
|
+
Regular exercise provides numerous benefits including improved cardiovascular health, enhanced muscle strength, better mental health, and weight management. It reduces the risk of chronic diseases such as diabetes, hypertension, and certain cancers. Additionally, regular physical activity can improve sleep quality and increase overall energy levels.
|
20
|
+
[The End of Assistant A's Answer]
|
21
|
+
|
22
|
+
[Start of Assistant B Answer]
|
23
|
+
Engaging in regular exercise has several advantages. It strengthens the heart, boosts muscle power, and aids in maintaining a healthy weight. Exercise is also beneficial for mental well-being, as it can reduce symptoms of anxiety and depression. Furthermore, it helps in preventing chronic illnesses like diabetes, heart disease, and obesity. Regular physical activity can lead to better sleep and increased energy throughout the day.
|
24
|
+
[The End of Assistant B Answer]
|
25
|
+
|
26
|
+
[Start of Judgement]
|
27
|
+
Both responses highlight the benefits of regular exercise, including cardiovascular health, muscle strength, mental health, and chronic disease prevention. However, Assistant B’s answer is more detailed and specific, mentioning anxiety and depression, which provides a slightly broader view of the mental health benefits. Assistant A’s response is concise and clear but lacks the additional specifics found in Assistant B’s response.
|
28
|
+
[End of Judgement]
|
29
|
+
|
30
|
+
[Start of Verdict]
|
31
|
+
Assistant B
|
32
|
+
[End of Verdict]
|
33
|
+
|
34
|
+
|
35
|
+
generation: |
|
36
|
+
After providing your judgement, output your final verdict by strictly following this format:
|
37
|
+
* Return the judgement between [Start of Judgement] and [End of Judgement] tags
|
38
|
+
* Return the final verdict between [Start of Verdict] and [End of Verdict] tags. Respond with Assistant A if assistant A is better, Assistant B if assistant B is better
|
39
|
+
|
40
|
+
[Start of User Question]
|
41
|
+
{{ question }}
|
42
|
+
[End of User Question]
|
43
|
+
|
44
|
+
[Start of Assistant A Answer]
|
45
|
+
{{ response }}
|
46
|
+
[The End of Assistant A Answer]
|
47
|
+
|
48
|
+
[Start of Assistant B Answer]
|
49
|
+
{{ revised_response }}
|
50
|
+
[The End of Assistant B Answer]
|
51
|
+
|
52
|
+
start_tags: ["[Start of Judgement]", "[Start of Verdict]"]
|
53
|
+
end_tags: ["[End of Judgement]", "[End of Verdict]"]
|
@@ -0,0 +1,67 @@
|
|
1
|
+
system: |
|
2
|
+
You are a very knowledgeable AI Assistant that will faithfully assist the user with their task.
|
3
|
+
|
4
|
+
introduction: |
|
5
|
+
Assume the role of an expert planner. You will be given a query and a base response generated by an AI assistant. Your task is to generate a response that provides a critique of the base response.
|
6
|
+
You will also be given an analysis, rubric and critique from a critic model. Your task is to generate a plan and actions to improve the base response based on the critique provided.
|
7
|
+
|
8
|
+
principles: |
|
9
|
+
* You will utilize the identified domain, rubric and the evaluation of the model response to generate a plan to improve the response.
|
10
|
+
* The plan should be step-by-step and should satisfy each criteria of the rubric.
|
11
|
+
* The plan should not be generic and should be specific to the model response and the domain-specific rubric and evaluation from the critique.
|
12
|
+
* The generated plan should be actionable and feasible.
|
13
|
+
* Your task is to only generate a plan, do not try to implement the plan or provide a revised response.
|
14
|
+
* Return the plan between [Start of Plan] and [End of Plan] tags.
|
15
|
+
|
16
|
+
examples: |
|
17
|
+
To help you understand the task, here is an example:
|
18
|
+
|
19
|
+
[Start of Query]
|
20
|
+
{{ icl_query }}
|
21
|
+
[End of Query]
|
22
|
+
|
23
|
+
[Start of Response]
|
24
|
+
{{ icl_response }}
|
25
|
+
[End of Response]
|
26
|
+
|
27
|
+
[Start of Analysis]
|
28
|
+
{{ icl_analysis }}
|
29
|
+
[End of Analysis]
|
30
|
+
|
31
|
+
[Start of Rubric]
|
32
|
+
{{ icl_rubric }}
|
33
|
+
[End of Rubric]
|
34
|
+
|
35
|
+
[Start of Critique]
|
36
|
+
{{ icl_critique }}
|
37
|
+
[End of Critique]
|
38
|
+
|
39
|
+
[Start of Plan]
|
40
|
+
{{ icl_plan }}
|
41
|
+
[End of Plan]
|
42
|
+
|
43
|
+
generation: |
|
44
|
+
Now it's your turn to improve the response to the following query. Remember to follow the paradigm and return the plan in the respective section in the same format as above.
|
45
|
+
|
46
|
+
[Start of Query]
|
47
|
+
{{ question }}
|
48
|
+
[End of Query]
|
49
|
+
|
50
|
+
[Start of Response]
|
51
|
+
{{ response }}
|
52
|
+
[End of Response]
|
53
|
+
|
54
|
+
[Start of Analysis]
|
55
|
+
{{ analysis }}
|
56
|
+
[End of Analysis]
|
57
|
+
|
58
|
+
[Start of Rubric]
|
59
|
+
{{ rubric }}
|
60
|
+
[End of Rubric]
|
61
|
+
|
62
|
+
[Start of Critique]
|
63
|
+
{{ critique }}
|
64
|
+
[End of Critique]
|
65
|
+
|
66
|
+
start_tags: ["[Start of Plan]"]
|
67
|
+
end_tags: ["[End of Plan]"]
|
@@ -0,0 +1,78 @@
|
|
1
|
+
system: |
|
2
|
+
You are a very knowledgeable AI Assistant that will faithfully assist the user with their task.
|
3
|
+
|
4
|
+
introduction: |
|
5
|
+
Your task is to revise the response to an user query. You will be given a query from an user and a response from a model. Your task is to provide a better response based on the plan given.
|
6
|
+
|
7
|
+
principles: |
|
8
|
+
* You will revise the model response according to the plan given.
|
9
|
+
* The revised response should adhere to the plan and should be better than the original response.
|
10
|
+
* Note that the revised response will be evaluated by a human expert and should thus be of high quality.
|
11
|
+
* Do not have any irrelevant information in the revised response. Specifically do not include any self-referential information in the revised response.
|
12
|
+
* Your response should only include the revised response. Please do not include any other information like the query, analysis, rubric, etc.
|
13
|
+
* Your response will become invalid if it contains any meta-review about how you are revising the response. So please avoid including any such information.
|
14
|
+
* If the plan mentions that there is no need to provide a plan for improvement, simply return the original response as the revised response.
|
15
|
+
* Return the revised response between [Start of Revised Response] and [End of Revised Response] tags.
|
16
|
+
|
17
|
+
examples: |
|
18
|
+
To help you understand the task, here is an example:
|
19
|
+
|
20
|
+
[Start of Query]
|
21
|
+
{{ icl_query }}
|
22
|
+
[End of Query]
|
23
|
+
|
24
|
+
[Start of Response]
|
25
|
+
{{ icl_response }}
|
26
|
+
[End of Response]
|
27
|
+
|
28
|
+
[Start of Analysis]
|
29
|
+
{{ icl_analysis }}
|
30
|
+
[End of Analysis]
|
31
|
+
|
32
|
+
[Start of Rubric]
|
33
|
+
{{ icl_rubric }}
|
34
|
+
[End of Rubric]
|
35
|
+
|
36
|
+
[Start of Critique]
|
37
|
+
{{ icl_critique }}
|
38
|
+
[End of Critique]
|
39
|
+
|
40
|
+
[Start of Plan]
|
41
|
+
{{ icl_plan }}
|
42
|
+
[End of Plan]
|
43
|
+
|
44
|
+
[Start of Revised Response]
|
45
|
+
{{ icl_revised_response }}
|
46
|
+
[End of Revised Response]
|
47
|
+
|
48
|
+
generation: |
|
49
|
+
Now it's your turn to revise the response to the following query. Remember to follow the paradigm and return the revised response in the respective section in the same format as above. Strictly do not include any meta-review or meta-information about how the response was improved or revised. Your response should only include the revised response. You will be heavily penalized if you include any information about the revision process or if you have any reference about how you revised the response.
|
50
|
+
|
51
|
+
[Start of Query]
|
52
|
+
{{ question }}
|
53
|
+
[End of Query]
|
54
|
+
|
55
|
+
[Start of Response]
|
56
|
+
{{ response }}
|
57
|
+
[End of Response]
|
58
|
+
|
59
|
+
[Start of Analysis]
|
60
|
+
{{ analysis }}
|
61
|
+
[End of Analysis]
|
62
|
+
|
63
|
+
[Start of Rubric]
|
64
|
+
{{ rubric }}
|
65
|
+
[End of Rubric]
|
66
|
+
|
67
|
+
[Start of Critique]
|
68
|
+
{{ critique }}
|
69
|
+
[End of Critique]
|
70
|
+
|
71
|
+
[Start of Plan]
|
72
|
+
{{ plan }}
|
73
|
+
[End of Plan]
|
74
|
+
|
75
|
+
Start your response with the tag [Start of Revised Response] and end it with the tag [End of Revised Response].
|
76
|
+
|
77
|
+
start_tags: ["[Start of Revised Response]"]
|
78
|
+
end_tags: ["[End of Revised Response]"]
|
@@ -0,0 +1,59 @@
|
|
1
|
+
system: |
|
2
|
+
You are a skill classifier. Given a question or task, classify it into exactly one of the following categories:
|
3
|
+
- coding: Questions about programming, software development, algorithms, or technical implementation.
|
4
|
+
- extraction: Tasks that require identifying and pulling out specific pieces of information from text or data, without analysis or transformation. This includes extracting structured data and transforming it into a specific format.
|
5
|
+
- humanities: Questions about economics, social sciences, literature, history, philosophy, or arts that require analysis, comparison, or evaluation.
|
6
|
+
- math: Questions requiring mathematical calculations, proofs, probability, or numerical reasoning.
|
7
|
+
- reasoning: Tasks requiring logical deduction, problem-solving, or analytical thinking.
|
8
|
+
- roleplay: Scenarios requiring adopting a specific role, perspective, or character.
|
9
|
+
- stem: Questions about science, technology, engineering, or mathematics that require explaining technical or scientific concepts.
|
10
|
+
- writing: Tasks that require creating new content, compositions, or text generation.
|
11
|
+
|
12
|
+
Only output the category name, and nothing else. Please do not output anything except one of these exact words.
|
13
|
+
|
14
|
+
introduction: |
|
15
|
+
Classify the given question into one of the predefined categories based on its primary focus and requirements.
|
16
|
+
|
17
|
+
principles: |
|
18
|
+
1. Each question should be classified into exactly one category.
|
19
|
+
2. Choose the category that best represents the primary skill or domain required.
|
20
|
+
3. If a question could fit multiple categories, choose the most specific one.
|
21
|
+
4. Consider the main task or goal of the question, not just the subject matter.
|
22
|
+
5. For extraction tasks, focus on questions that require pulling out specific information or data, often with specific formatting requirements, without significant analysis or creative generation.
|
23
|
+
6. For humanities tasks, focus on questions that require analysis, comparison, or evaluation of social, cultural, or historical topics.
|
24
|
+
7. For writing tasks, focus on questions that require creative or open-ended content generation or composition, rather than tasks with strict formatting or content extraction rules.
|
25
|
+
|
26
|
+
examples: |
|
27
|
+
Question: Extract and categorize the health-related themes, such as 'symptoms', 'treatments', and 'prevention'
|
28
|
+
Category: extraction
|
29
|
+
|
30
|
+
Question: Summarize the story with three bullet points using only nouns and adjectives, without verbs.
|
31
|
+
Category: extraction
|
32
|
+
|
33
|
+
Question: Can you analyze the relationship between economic indicators like GDP, interest rates, and consumer behavior?
|
34
|
+
Category: humanities
|
35
|
+
|
36
|
+
Question: Calculate the area of a triangle with base 6 units and height 8 units
|
37
|
+
Category: math
|
38
|
+
|
39
|
+
Question: Suppose you are in a marathon and you just passed the person in third place. What position are you in?
|
40
|
+
Category: reasoning
|
41
|
+
|
42
|
+
Question: You are a medieval blacksmith. Describe your typical workday and the challenges you face
|
43
|
+
Category: roleplay
|
44
|
+
|
45
|
+
Question: In the realm of quantum mechanics, what is wave-particle duality, and how does it contribute to the understanding of quantum systems?
|
46
|
+
Category: STEM
|
47
|
+
|
48
|
+
Question: Write a travel blog post about exploring the ancient temples and street food scene in Bangkok, Thailand
|
49
|
+
Category: writing
|
50
|
+
|
51
|
+
generation: |
|
52
|
+
Question: {{ question }}
|
53
|
+
|
54
|
+
Based on the above question, classify it into exactly one of these categories: coding, extraction, humanities, math, reasoning, roleplay, STEM, or writing.
|
55
|
+
|
56
|
+
Category:
|
57
|
+
|
58
|
+
start_tags: [""]
|
59
|
+
end_tags: [""]
|
@@ -0,0 +1,27 @@
|
|
1
|
+
system: You are a very knowledgeable AI Assistant that will faithfully assist the user with their task.
|
2
|
+
|
3
|
+
introduction: Develop a series of question and answer pairs to perform a task.
|
4
|
+
|
5
|
+
principles: |
|
6
|
+
Here are the requirements:
|
7
|
+
1. Try not to repeat the verb for each instruction to maximize diversity.
|
8
|
+
2. The language used for the instruction also should be diverse. For example, you should combine questions with imperative instructions.
|
9
|
+
3. The type of instructions should be similar to provided examples. The generated instruction and the output should be grounded in the provided document.
|
10
|
+
4. A GPT language model should be able to complete the instruction. For example, do not ask the assistant to create any visual or audio output. For another example, do not ask the assistant to wake you up at 5pm or set a reminder because it cannot perform any action.
|
11
|
+
5. The instructions should be in English.
|
12
|
+
6. The instructions should be 1 to 2 sentences long. Either an imperative sentence or a question is permitted.
|
13
|
+
7. The output should be an appropriate response to the input and the instruction. Long outputs are preferable.
|
14
|
+
|
15
|
+
examples: |
|
16
|
+
The task is {{task_description}}.
|
17
|
+
|
18
|
+
Here is an example to help you understand the type of questions that are asked for:
|
19
|
+
|
20
|
+
{{seed_question}}
|
21
|
+
{{seed_response}}
|
22
|
+
|
23
|
+
generation: |
|
24
|
+
Provide a single question and answer pair based on the examples.
|
25
|
+
|
26
|
+
start_tags: [""]
|
27
|
+
end_tags: [""]
|
@@ -0,0 +1,31 @@
|
|
1
|
+
system: You are a very knowledgeable AI Assistant that will faithfully assist the user with their task.
|
2
|
+
|
3
|
+
introduction: Develop a series of question and answer pairs to perform a task.
|
4
|
+
|
5
|
+
principles: |
|
6
|
+
Here are the requirements:
|
7
|
+
1. Try not to repeat the verb for each instruction to maximize diversity.
|
8
|
+
2. The language used for the instruction also should be diverse. For example, you should combine questions with imperative instructions.
|
9
|
+
3. The type of instructions should be similar to provided examples. The generated instruction and the output should be grounded in the provided document.
|
10
|
+
4. A GPT language model should be able to complete the instruction. For example, do not ask the assistant to create any visual or audio output. For another example, do not ask the assistant to wake you up at 5pm or set a reminder because it cannot perform any action.
|
11
|
+
5. The instructions should be in English.
|
12
|
+
6. The instructions should be 1 to 2 sentences long. Either an imperative sentence or a question is permitted.
|
13
|
+
7. The output should be an appropriate response to the input and the instruction. Long outputs are preferable.
|
14
|
+
|
15
|
+
examples: |
|
16
|
+
The task is {{task_description}}.
|
17
|
+
|
18
|
+
Here is some context for the example question:
|
19
|
+
|
20
|
+
{{seed_context}}
|
21
|
+
|
22
|
+
Here is an example to help you understand the type of questions that are asked for:
|
23
|
+
|
24
|
+
{{seed_question}}
|
25
|
+
{{seed_response}}
|
26
|
+
|
27
|
+
generation: |
|
28
|
+
Provide a single question and answer pair based on the example.
|
29
|
+
|
30
|
+
start_tags: [""]
|
31
|
+
end_tags: [""]
|