Back to tags
Tag

Agent Skills with tag: Optimization

22 skills match this tag. Use tags to discover related Agent Skills and explore similar workflows.

fine-tuning-peft

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

[Fine-TuningPEFTLoRAQLoRA
Orchestra Research
Orchestra Research
33,1943,891

inference-serving-tensorrt-llm

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

[InferenceServingTensorRT-LLMNVIDIA
Orchestra Research
Orchestra Research
33,1943,891

infrastructure-skypilot

Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or batch jobs across multiple clouds, leverage spot instances with auto-recovery, or optimize GPU costs across providers.

[InfrastructureMulti-CloudOrchestrationGPU
Orchestra Research
Orchestra Research
33,1943,891

optimization-awq

Activation-aware weight quantization for 4-bit LLM compression with 3x speedup and minimal accuracy loss. Use when deploying large models (7B-70B) on limited GPU memory, when you need faster inference than GPTQ with better accuracy preservation, or for instruction-tuned and multimodal models. MLSys 2024 Best Paper Award winner.

[OptimizationAWQQuantization4-Bit
Orchestra Research
Orchestra Research
33,1943,891

optimization-bitsandbytes

Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limited, need to fit larger models, or want faster inference. Supports INT8, NF4, FP4 formats, QLoRA training, and 8-bit optimizers. Works with HuggingFace Transformers.

[OptimizationBitsandbytesQuantization8-Bit
Orchestra Research
Orchestra Research
33,1943,891

optimization-flash-attention

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

[OptimizationFlashAttentionAttention
Orchestra Research
Orchestra Research
33,1943,891

optimization-gptq

Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on consumer GPUs, when you need 4× memory reduction with <2% perplexity degradation, or for faster inference (3-4× speedup) vs FP16. Integrates with transformers and PEFT for QLoRA fine-tuning.

[OptimizationGPTQQuantization4-Bit
Orchestra Research
Orchestra Research
33,1943,891

optimization-hqq

Half-Quadratic Quantization for LLMs without calibration data. Use when quantizing models to 4/3/2-bit precision without needing calibration datasets, for fast quantization workflows, or when deploying with vLLM or HuggingFace Transformers.

[QuantizationHQQOptimizationMemory
Orchestra Research
Orchestra Research
33,1943,891

post-training-simpo

Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.

[Post-TrainingSimPOPreferenceOptimization
Orchestra Research
Orchestra Research
33,1943,891

prompt-engineering-dspy

Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and agents with DSPy - Stanford NLP's framework for systematic LM programming

[PromptEngineeringDSPyDeclarative
Orchestra Research
Orchestra Research
33,1943,891

tensorrt-llm

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

[InferenceServingTensorRT-LLMNVIDIA
poorrican
poorrican
0

dspy

DSPy: declarative LM programs, auto-optimize prompts, RAG.

[PromptEngineeringDSPyDeclarative
poorrican
poorrican
0

optimizing-attention-flash

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

[OptimizationFlashAttentionAttention
poorrican
poorrican
0

peft-fine-tuning

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

[Fine-TuningPEFTLoRAQLoRA
poorrican
poorrican
0

simpo-training

Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.

[Post-TrainingSimPOPreferenceOptimization
poorrican
poorrican
0

unsloth

Unsloth: 2-5x faster LoRA/QLoRA fine-tuning, less VRAM.

[Fine-TuningUnslothFastTraining
poorrican
poorrican
0

tensorrt-llm

Optimizes LLM inference with NVIDIA TensorRT for maximum throughput and lowest latency. Use for production deployment on NVIDIA GPUs (A100/H100), when you need 10-100x faster inference than PyTorch, or for serving models with quantization (FP8/INT4), in-flight batching, and multi-GPU scaling.

[InferenceServingTensorRT-LLMNVIDIA
PoorRican
PoorRican
0

dspy

DSPy: declarative LM programs, auto-optimize prompts, RAG.

[PromptEngineeringDSPyDeclarative
PoorRican
PoorRican
0

optimizing-attention-flash

Optimizes transformer attention with Flash Attention for 2-4x speedup and 10-20x memory reduction. Use when training/running transformers with long sequences (>512 tokens), encountering GPU memory issues with attention, or need faster inference. Supports PyTorch native SDPA, flash-attn library, H100 FP8, and sliding window attention.

[OptimizationFlashAttentionAttention
PoorRican
PoorRican
0

peft-fine-tuning

Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

[Fine-TuningPEFTLoRAQLoRA
PoorRican
PoorRican
0

simpo-training

Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.

[Post-TrainingSimPOPreferenceOptimization
PoorRican
PoorRican
0

unsloth

Unsloth: 2-5x faster LoRA/QLoRA fine-tuning, less VRAM.

[Fine-TuningUnslothFastTraining
PoorRican
PoorRican
0