AI/ML Prompts

Debug LLM applications with LangSmith. Features: 1. Trace every LLM call. 2. View chain execution steps. 3. Latency and token analysis. 4. Error tracking and debugging. 5. Dataset creation from logs. 6. Evaluation and testing. 7. Feedback collection. 8. Cost monitoring. Essential for production LLM apps. Use to identify bottlenecks and optimize prompts.

Build RAG systems with LlamaIndex. Workflow: 1. Load documents (PDF, DOCX, web). 2. Node parser for chunking. 3. Create embeddings with LLM. 4. Build index (Vector, Tree, Keyword). 5. Query engine for retrieval. 6. Response synthesizer. 7. Sub-question query engine. 8. Chat engine for conversations. Use ServiceContext for configuration and implement hybrid retrieval.

Street photography shot of a futuristic cyberpunk Tokyo alleyway at night, neon rain reflecting on wet pavement, a cyborg street vendor cooking noodles, steam rising, vibrant teal and magenta lighting, cinematic lighting, shot on 35mm lens, f/1.8, high contrast, photorealistic, 8k --ar 16:9 --v 6.0 --style raw

Generate 100,000+ high-quality training examples using LLMs. Features: 1. 'Seed data' input. 2. Variation logic (Rewrite, Summarize, Expand). 3. Self-correcting loop to remove bad samples. 4. Progress bar and cost estimator. 5. Download in JSONL/CSV format. Optimized for scale and diversity.

Reveal the hidden 'thinking' process of an LLM. UI shows: 1. Original User Query. 2. Internal Thinking steps (hidden from usual output). 3. Final Conclusion. Highlight keywords that triggered transitions between thoughts. Use a clean, educational layout to help debug prompt logic.

Inspect what an AI agent 'remembers'. Sections: 1. Short-term Memory (Chat History). 2. Long-term Memory (Vector retrieval). 3. Entity Memory (Facts about the user). 4. Importance/Weighting adjustment knobs. 5. Visual graph of related memories. Useful for building multi-turn complex agents.

A leaderboard-style comparison of different fine-tuned models. Compare: 1. Llama 3 (LoRA) vs GPT-4v (RLHF) vs Mistral (Base). 2. Benchmarks: MMLU, GSM8k, HumanEval. 3. Column to show 'Inference Cost' vs 'Accuracy'. 4. Radar chart for multi-dimensional performance analysis.

Visualize how similarity search works. Features: 1. Input text field. 2. Results list with 'Similarity Score' (0-1.0). 3. 3D t-SNE scatter plot showing vector clusters. 4. Filter by namespace/metadata. 5. Performance info (Query time in ms). Compatible with Pinecone, Chroma, and Weaviate APIs.

Monitor fine-tuning of Low-Rank Adaptation models. UI elements: 1. Real-time loss graph. 2. Epoch/Step counters. 3. Predicted remaining time. 4. Samples generated mid-training (checkpoints). 5. Hardware metrics: VRAM usage, GPU Temp. Use a dark, developer-focused aesthetic with neon accents.

Measure which prompt performs better. Features: 1. Two versions of a prompt (Variant A vs B). 2. Key Metrics: Relevancy Score, Accuracy, Response Speed, Token Usage. 3. Statistically significant 'Winner' badge. 4. User feedback collection tool for manual evaluation. 5. Chart comparing costs over 1,000 runs.

Professional diagram following Retrieval Augmented Generation architecture. Components: 1. Document Loader -> Splitting -> Embeddings. 2. Vector DB Storage. 3. Query Rewrite -> Retrieval -> Re-ranking. 4. Contextual Prompt -> LLM Generation. Use blue/violet gradients and high-quality technical icons.

Dynamic visualization of Microsoft AutoGen agent chats. Use a messaging interface style where each bubble indicates which agent (Assistant, Critic, User) spoke. Include a 'Context Window' sidebar that shows the tokens used and cost per message. Highlight when the 'TERMINATE' command is triggered.

Visualize a team of AI agents working together. The 'Crew' includes: 1. 'Researcher' (fetches facts). 2. 'Writer' (drafts blog). 3. 'Manager' (approves/feedback). Show a timeline of tasks being handed off between agents. Use a clean, modern dashboard with status badges (Ongoing, Completed, Failed).

Compare visual outputs from multiple SD models. Layout: 1. One prompt input that sends to SDXL, SD1.5, and Playground v2. 2. 3-column grid showing generated images. 3. Metadata overlay showing seed, sampler, and CFG scale. 4. 'Download All' button. 5. History sidebar of previous generations.

A tool to auto-generate Hugging Face model cards. Sections to include: 1. Model Description (Architecture, Parameters). 2. Training Data (Datasets used). 3. Evaluation Results (MMLU, HumanEval scores). 4. Intended Use and Biases. 5. Citation info. Minimalist layout with badges for 'Transformers', 'PyTorch', 'Safetensors'.

A UI for inspecting JSONL datasets for fine-tuning Llama 3. Features: 1. Raw JSON vs 'Chat View' toggle. 2. Token counter per example. 3. Quality score badge (AI-evaluated). 4. Search and filter by 'instruction' or 'response' keywords. 5. Export filtered view to CSV/Parquet.

Designed to take a messy user prompt and 'optimize' it for GPT-4o. The optimizer should add: 1. Specific Persona (Expert Scientist, Creative Writer). 2. Constraints (No jargon, under 200 words). 3. Step-by-step reasoning instructions. 4. Expected output format (JSON, Markdown). 5. Few-shot examples to guide the model.

Visualize a complex LangChain agent flow. Flow components: 1. User Input -> Embedding Model. 2. Vector DB (Pinecone) retrieval. 3. LLM (GPT-4) reasoning step. 4. Tool execution (Google Search, Python Repl). 5. Final Output. Use a node-based diagram style with directed arrows and color-coded component boxes.