Awq Quantization Agent Skill

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.

6k tokens

context cost

the whole folder, loaded on every use

files

instructions only

copies elsewhere

how many repositories repackaged it

11362

stars on the repo

on the repository, not the skill itself

Install

one command, takes just this skill from the repository

npx skills add https://github.com/Orchestra-Research/AI-Research-SKILLs --skill awq-quantization

The instruction itself

26 sections, as written by the author

AWQ (Activation-aware Weight Quantization)

4-bit quantization that preserves salient weights based on activation patterns, achieving 3x speedup with minimal accuracy loss.

When to use AWQ

Use AWQ when:

Need 4-bit quantization with <5% accuracy loss
Deploying instruction-tuned or chat models (AWQ generalizes better)
Want ~2.5-3x inference speedup over FP16
Using vLLM for production serving
Have Ampere+ GPUs (A100, H100, RTX 40xx) for Marlin kernel support

Use GPTQ instead when:

Need maximum ecosystem compatibility (more tools support GPTQ)
Working with ExLlamaV2 backend specifically
Have older GPUs without Marlin support

Use bitsandbytes instead when:

Need zero calibration overhead (quantize on-the-fly)
Want to fine-tune with QLoRA
Prefer simpler integration

Quick start

Installation

# Default (Triton kernels)
pip install autoawq

# With optimized CUDA kernels + Flash Attention
pip install autoawq[kernels]

# Intel CPU/XPU optimization
pip install autoawq[cpu]

Requirements: Python 3.8+, CUDA 11.8+, Compute Capability 7.5+

Load pre-quantized model

from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer

model_name = "TheBloke/Mistral-7B-Instruct-v0.2-AWQ"

model = AutoAWQForCausalLM.from_quantized(
    model_name,
    fuse_layers=True  # Enable fused attention for speed
)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# Generate
inputs = tokenizer("Explain quantum computing", return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Quantize your own model

from awq import AutoAWQForCausalLM
from transformers import AutoTokenizer

model_path = "mistralai/Mistral-7B-Instruct-v0.2"

# Load model and tokenizer
model = AutoAWQForCausalLM.from_pretrained(model_path)
tokenizer = AutoTokenizer.from_pretrained(model_path)

# Quantization config
quant_config = {
    "zero_point": True,      # Use zero-point quantization
    "q_group_size": 128,     # Group size (128 recommended)
    "w_bit": 4,              # 4-bit weights
    "version": "GEMM"        # GEMM for batch, GEMV for single-token
}

# Quantize (uses pileval dataset by default)
model.quantize(tokenizer, quant_config=quant_config)

# Save
model.save_quantized("mistral-7b-awq")
tokenizer.save_pretrained("mistral-7b-awq")

Timing: ~10-15 min for 7B, ~1 hour for 70B models.

AWQ vs GPTQ vs bitsandbytes

|---------|-----|------|--------------|

| Speedup (4-bit) | ~2.5-3x | ~2x | ~1.5x |

| Accuracy loss | <5% | ~5-10% | ~5-15% |

| Overfitting risk | Low | Higher | N/A |

Key insight: AWQ assumes not all weights are equally important. It protects ~1% of salient weights identified by activation patterns, reducing quantization error without mixed-precision overhead.

Kernel backends

GEMM (default, batch inference)

quant_config = {
    "zero_point": True,
    "q_group_size": 128,
    "w_bit": 4,
    "version": "GEMM"  # Best for batch sizes > 1
}

GEMV (single-token generation)

quant_config = {
    "version": "GEMV"  # 20% faster for batch_size=1
}

Limitation: Only batch size 1, not good for large context.

Marlin (Ampere+ GPUs)

from transformers import AwqConfig, AutoModelForCausalLM

config = AwqConfig(
    bits=4,
    version="marlin"  # 2x faster on A100/H100
)

model = AutoModelForCausalLM.from_pretrained(
    "TheBloke/Mistral-7B-AWQ",
    quantization_config=config
)

Requirements: Compute Capability 8.0+ (A100, H100, RTX 40xx)

ExLlamaV2 (AMD compatible)

config = AwqConfig(
    bits=4,
    version="exllama"  # Faster prefill, AMD GPU support
)

HuggingFace Transformers integration

Direct loading

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained(
    "TheBloke/zephyr-7B-alpha-AWQ",
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("TheBloke/zephyr-7B-alpha-AWQ")

Fused modules (recommended)

from transformers import AwqConfig, AutoModelForCausalLM

config = AwqConfig(
    bits=4,
    fuse_max_seq_len=512,  # Max sequence length for fusing
    do_fuse=True           # Enable fused attention/MLP
)

model = AutoModelForCausalLM.from_pretrained(
    "TheBloke/Mistral-7B-OpenOrca-AWQ",
    quantization_config=config
)

Note: Fused modules cannot combine with FlashAttention2.

vLLM integration

from vllm import LLM, SamplingParams

# vLLM auto-detects AWQ models
llm = LLM(
    model="TheBloke/Llama-2-7B-AWQ",
    quantization="awq",
    dtype="half"
)

sampling = SamplingParams(temperature=0.7, max_tokens=200)
outputs = llm.generate(["Explain AI"], sampling)

Performance benchmarks

Memory reduction

|-------|------|-----------|-----------|

| Mistral 7B | 14 GB | 5.5 GB | 2.5x |

| Llama 2-13B | 26 GB | 10 GB | 2.6x |

| Llama 2-70B | 140 GB | 35 GB | 4x |

Inference speed (RTX 4090)

|-------|-----------------|----------------|--------|

| Mistral 7B GEMM | 3,897 | 114 | 5.55 GB |

| TinyLlama 1B GEMV | 5,179 | 431 | 2.10 GB |

| Llama 2-13B GEMM | 2,279 | 74 | 10.28 GB |

Accuracy (perplexity)

|-------|------|-----------|-------------|

| Llama 3 8B | 8.20 | 8.48 | +3.4% |

| Mistral 7B | 5.25 | 5.42 | +3.2% |

| Qwen2 72B | 4.85 | 4.95 | +2.1% |

Custom calibration data

# Use custom dataset for domain-specific models
model.quantize(
    tokenizer,
    quant_config=quant_config,
    calib_data="wikitext",       # Or custom list of strings
    max_calib_samples=256,       # More samples = better accuracy
    max_calib_seq_len=512        # Sequence length
)

# Or provide your own samples
calib_samples = [
    "Your domain-specific text here...",
    "More examples from your use case...",
]
model.quantize(tokenizer, quant_config=quant_config, calib_data=calib_samples)

Multi-GPU deployment

model = AutoAWQForCausalLM.from_quantized(
    "TheBloke/Llama-2-70B-AWQ",
    device_map="auto",  # Auto-split across GPUs
    max_memory={0: "40GB", 1: "40GB"}
)

Supported models

35+ architectures including:

Llama family: Llama 2/3, Code Llama, Mistral, Mixtral
Qwen: Qwen, Qwen2, Qwen2.5-VL
Others: Falcon, MPT, Phi, Yi, DeepSeek, Gemma
Multimodal: LLaVA, LLaVA-Next, Qwen2-VL

Common issues

CUDA OOM during quantization:

# Reduce batch size
model.quantize(tokenizer, quant_config=quant_config, max_calib_samples=64)

Slow inference:

# Enable fused layers
model = AutoAWQForCausalLM.from_quantized(model_name, fuse_layers=True)

AMD GPU support:

# Use ExLlama backend
config = AwqConfig(bits=4, version="exllama")

Deprecation notice

AutoAWQ is officially deprecated. For new projects, consider:

vLLM llm-compressor: https://github.com/vllm-project/llm-compressor
MLX-LM: For Mac devices with Apple Silicon

Existing quantized models remain usable.

References

Paper: AWQ: Activation-aware Weight Quantization (arXiv:2306.00978) - MLSys 2024 Best Paper
GitHub: https://github.com/casper-hansen/AutoAWQ
MIT Han Lab: https://github.com/mit-han-lab/llm-awq
Models: https://huggingface.co/models?library=awq

Other skills for the same job

different authors, same section of the catalogue

Skill Creator

by anthropics

vendor ×10

Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.

Access NCBI GEO for gene expression/genomics data. Search/download microarray and RNA-seq datasets (GSE, GSM, GPL), retrieve SOFT/Matrix files, for transcriptomics and expression analysis.

12k tokens

Pymc Bayesian Modeling

by christophacham

×4

Bayesian modeling with PyMC. Build hierarchical models, MCMC (NUTS), variational inference, LOO/WAIC comparison, posterior checks, for probabilistic programming and inference.

Multi-objective optimization framework. NSGA-II, NSGA-III, MOEA/D, Pareto fronts, constraint handling, benchmarks (ZDT, DTLZ), for engineering design and optimization problems.

Statistical modeling toolkit. OLS, GLM, logistic, ARIMA, time series, hypothesis tests, diagnostics, AIC/BIC, for rigorous statistical inference and econometric analysis.

Add unsigned integer (uint) type support to PyTorch operators by updating AT_DISPATCH macros. Use when adding support for uint16, uint32, uint64 types to operators, kernels, or when user mentions enabling unsigned types, barebones unsigned types, or uint support.

Convert PyTorch AT_DISPATCH macros to AT_DISPATCH_V2 format in ATen C++ code. Use when porting AT_DISPATCH_ALL_TYPES_AND*, AT_DISPATCH_FLOATING_TYPES*, or other dispatch macros to the new v2 API. For ATen kernel files, CUDA kernels, and native operator implementations.

Write docstrings for PyTorch functions and methods following PyTorch conventions. Use when writing or updating docstrings in PyTorch code.

3k tokens

How to use it

Copy the folder

Take orchestra-research/awq-quantization from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.

Install what it needs

The instructions reference pip. Without those the skill loads but fails at the first command.