SkillsAdd

Source

firecrawl/ai-research-skills

83 skills · 1.9K combined installs

Skills from this source

#
Skill
Source
8W Activity
Installs
1
gptq Post-training 4-bit quantization for LLMs with minimal accuracy loss. Use for deploying large models (70B, 405B) on c…
firecrawl/ai-research-skills
22
2
grpo-rl-training Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training
firecrawl/ai-research-skills
22
3
huggingface-accelerate Simplest distributed training API. 4 lines to add distributed support to any PyTorch script. Unified API for DeepSpee…
firecrawl/ai-research-skills
22
4
implementing-llms-litgpt Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mi…
firecrawl/ai-research-skills
22
5
instructor Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse co…
firecrawl/ai-research-skills
22
6
lambda-labs-gpu-cloud Reserved and on-demand GPU cloud instances for ML training and inference. Use when you need dedicated GPU instances w…
firecrawl/ai-research-skills
22
7
llama-factory Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimod…
firecrawl/ai-research-skills
22
8
llamaguard Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual …
firecrawl/ai-research-skills
22
9
mamba-architecture State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV ca…
firecrawl/ai-research-skills
22
10
miles-rl-training Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training l…
firecrawl/ai-research-skills
22
11
mlflow Track ML experiments, manage model registry with versioning, deploy models to production, and reproduce experiments w…
firecrawl/ai-research-skills
22
12
model-pruning Reduce LLM size and accelerate inference using pruning techniques like Wanda and SparseGPT. Use when compressing mode…
firecrawl/ai-research-skills
22
13
moe-training Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limit…
firecrawl/ai-research-skills
22
14
nanogpt Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learni…
firecrawl/ai-research-skills
22
15
nemo-evaluator-sdk Evaluates LLMs across 100+ benchmarks from 18+ harnesses (MMLU, HumanEval, GSM8K, safety, VLM) with multi-backend exe…
firecrawl/ai-research-skills
22
16
nemo-guardrails NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-c…
firecrawl/ai-research-skills
22
17
nnsight-remote-interpretability Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote …
firecrawl/ai-research-skills
22
18
openrlhf-training High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B…
firecrawl/ai-research-skills
22
19
peft-fine-tuning Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70…
firecrawl/ai-research-skills
22
20
phoenix-observability Open-source AI observability platform for LLM tracing, evaluation, and monitoring. Use when debugging LLM application…
firecrawl/ai-research-skills
22
21
pytorch-fsdp2 Adds PyTorch FSDP2 (fully_shard) to training scripts with correct init, sharding, mixed precision/offload config, and…
firecrawl/ai-research-skills
22
22
pytorch-lightning High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks syste…
firecrawl/ai-research-skills
22
23
qdrant-vector-search High-performance vector similarity search engine for RAG and semantic search. Use when building production RAG system…
firecrawl/ai-research-skills
22
24
quantizing-models-bitsandbytes Quantizes LLMs to 8-bit or 4-bit for 50-75% memory reduction with minimal accuracy loss. Use when GPU memory is limit…
firecrawl/ai-research-skills
22
25
ray-data Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Inte…
firecrawl/ai-research-skills
22
26
rwkv-architecture RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), in…
firecrawl/ai-research-skills
22
27
simpo-training Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 poi…
firecrawl/ai-research-skills
22
28
skypilot-multi-cloud-orchestration Multi-cloud orchestration for ML workloads with automatic cost optimization. Use when you need to run training or bat…
firecrawl/ai-research-skills
22
29
slime-rl-training Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM model…
firecrawl/ai-research-skills
22
30
sparse-autoencoder-training Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network act…
firecrawl/ai-research-skills
22
31
training-llms-megatron Trains large language models (2B-462B parameters) using NVIDIA Megatron-Core with advanced parallelism strategies. Us…
firecrawl/ai-research-skills
22
32
verl-rl-training Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing…
firecrawl/ai-research-skills
22
33
weights-and-biases Track ML experiments with automatic logging, visualize training in real-time, optimize hyperparameters with sweeps, a…
firecrawl/ai-research-skills
22
Page 2 · 83 total Previous