@Orchestra-Research
Orchestra Research
- Skills
- 40
- GitHub stars
- 14k
- Installs
- 103k
Top skills by Orchestra Research
Orchestra Research
AI Research SKILLs
Orchestra Researchgrpo-rl-training
Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training
Orchestra Research
AI Research SKILLs
Orchestra Researchawq-quantization
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…
Orchestra Research
AI Research SKILLs
Orchestra Researchdeepspeed
Expert guidance for distributed training with DeepSpeed - ZeRO optimization stages, pipeline parallelism, FP16/BF16/FP8, 1-bit Adam, sparse attention
Orchestra Research
AI Research SKILLs
Orchestra Researchdistributed-llm-pretraining-torchtitan
Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8…
Orchestra Research
AI Research SKILLs
Orchestra Researchfine-tuning-with-trl
Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need…
Orchestra Research
AI Research SKILLs
Orchestra Researchhuggingface-accelerate
Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpeed/FSDP/Megatron/DDP. Automatic device placement, mixed precision…
Orchestra Research
AI Research SKILLs
Orchestra Researchhuggingface-tokenizers
Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram algorithms. Train custom vocabularies…
Orchestra Research
AI Research SKILLs
Orchestra Researchimplementing-llms-litgpt
Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model implementations, educational…
Orchestra Research
AI Research SKILLs
Orchestra Researchlambda-labs-gpu-cloud
Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances with simple SSH access, persistent filesystems, or high-performance…
Orchestra Research
AI Research SKILLs
Orchestra Researchmamba-architecture
State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware design. Mamba-1 (d_state=16) and…
Orchestra Research
AI Research SKILLs
Orchestra Researchmiles-rl-training
Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training large MoE models with FP8/INT4, needing train-inference alignment, or…
Orchestra Research
AI Research SKILLs
Orchestra Researchmodal-serverless-gpu
Serverless GPU cloud platform for running ML workloads. Use when you need on-demand GPU access without infrastructure management, deploying ML models as APIs, or running batch jobs with…
Orchestra Research
AI Research SKILLs
Orchestra Researchnanogpt
Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learning transformers. By Andrej Karpathy. Perfect for understanding GPT…
Orchestra Research
AI Research SKILLs
Orchestra Researchnemo-curator
GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII…
Orchestra Research
AI Research SKILLs
Orchestra Researchnnsight-remote-interpretability
Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on…
Orchestra Research
AI Research SKILLs
Orchestra Researchopenrlhf-training
High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with…
Orchestra Research
AI Research SKILLs
Orchestra Researchpeft-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…
Orchestra Research
AI Research SKILLs
Orchestra Researchpytorch-fsdp2
Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and distributed checkpointing. Use when models exceed single-GPU memory or…
Orchestra Research
AI Research SKILLs
Orchestra Researchpytorch-lightning
High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with…
Orchestra Research
AI Research SKILLs
Orchestra Researchpyvene-interventions
Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework. Use when conducting causal tracing, activation patching…
Orchestra Research
AI Research SKILLs
Orchestra Researchquantizing-models-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…
Orchestra Research
AI Research SKILLs
Orchestra Researchray-data
Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Integrates with Ray Train, PyTorch, TensorFlow. Scales from single machine…
Orchestra Research
AI Research SKILLs
Orchestra Researchray-train
Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance…
Orchestra Research
AI Research SKILLs
Orchestra Researchrwkv-architecture
RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), infer like RNN (sequential). Linux Foundation AI project. Production at…
Orchestra Research
AI Research SKILLs
Orchestra Researchskypilot-multi-cloud-orchestration
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…
Orchestra Research
AI Research SKILLs
Orchestra Researchslime-rl-training
Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM models, implementing custom data generation workflows, or needing tight…
Orchestra Research
AI Research SKILLs
Orchestra Researchsparse-autoencoder-training
Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable…
Orchestra Research
AI Research SKILLs
Orchestra Researchtorchforge-rl-training
Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use when you want clean RL abstractions, easy algorithm experimentation…
Orchestra Research
AI Research SKILLs
Orchestra Researchtraining-llms-megatron
Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Use when training models >1B parameters, need maximum GPU efficiency (47%…
Orchestra Research
AI Research SKILLs
Orchestra Researchtransformer-lens-interpretability
Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when…
Orchestra Research
AI Research SKILLs
Orchestra Researchverl-rl-training
Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale…
Orchestra Research
AI Research SKILLs
Orchestra Researchaxolotl
Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support
Orchestra Research
AI Research SKILLs
Orchestra Researchsentencepiece
Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences/sec), lightweight (6MB memory), deterministic vocabulary. Used by T5…
Orchestra Research
AI Research SKILLs
Orchestra Researchsimpo-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…
Orchestra Research
AI Research SKILLs
Orchestra Researchconstitutional-ai
Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then RLAIF (RL from AI Feedback). Use for safety…
Orchestra Research
AI Research SKILLs
Orchestra Researchllamaguard
Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual content, weapons, substances, self-harm, criminal planning. 94-95%…
Orchestra Research
AI Research SKILLs
Orchestra Researchnemo-guardrails
NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-checking, hallucination detection, PII filtering, toxicity detection…
Orchestra Research
AI Research SKILLs
Orchestra Researchprompt-guard
Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+ TPR, <1% FPR. Fast (<2ms GPU). Multilingual (8 languages). Deploy with…
Orchestra Research
AI Research SKILLs
Orchestra Researchllama-factory
Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support
Orchestra Research
AI Research SKILLs
Orchestra Researchunsloth
Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization