AWQ flags

English | 简体中文

This document presents step-by-step instructions for auto-round llm quantization. You can refer to vlms user guide for vlms quantization and diffusions user guide for diffusions quantization.

• 1 Prerequisite
• 2 Prepare Calibration Dataset
  • Default Dataset
  • Customized Dataset
  • Dataset operations
• 3 Quantization
  • Supported Quantization Schemes
  • Supported Export Formats
  • Hardware Compatibility
  • Environment Configuration
  • Command Line Usage
  • API usage
    • AutoRound API Usage
    • Mixed bits Usage
    • AutoRoundBest recipe
    • AutoRoundLight recipe
    • Recipe recommendation
  • AutoScheme
    • CLI Usage
    • API Usage
    • Hyperparameters in AutoScheme
  • OPT RTN mode
  • AWQ Algorithm-Experimental
  • Model-Free Mode
  • GGUF format
  • Quantization Costs
  • Device/Multi-GPU setting in Quantization
    • Enable multiple gpus calibration in lm_head quantization
  • Adjust Hyperparameters
  • Rotation (Experimental)
• 4 Inference
  • CPU
  • Intel GPU
  • CUDA
  • HPU
  • Specify Inference Backend
  • Convert GPTQ/AWQ to AutoRound
• 5 Evaluation
  • Single GPU Evaluation
  • Multi-GPU Evaluation
  • Important Notes
• 6 Known Issues

1 Prerequisite

Install auto-round or install from source

pip install auto-round

2 Prepare Calibration Dataset

Default Dataset

The NeelNanda/pile-10k in huggingface is adopted as the default calibration data and will be downloaded automatically from the datasets Hub. Other available datasets include:

• swift/pile-val-backup from modelscope for addressing HF network issue
• BAAI/CCI3-HQ for Chinese
• codeparrot/github-code-clean for code
• HuggingFaceH4/ultrachat_200k for chat data
• madao33/new-title-chinese for Chinese
• mbpp for code
• openbmb/Ultra-FineWeb

Customized Dataset

• Option 1: Pass a local json file path to dataset argument

• Option 2: Register your dataset following the code and pass the new dataset and split args to initialize AutoRound object, e.g. autoround=Autoround(dataset="NeelNanda/pile-10k:train", ...)

• Option 3: pass list of string or list of input_ids to dataset.

  def customized_data():
      # Important Notice!!! AutoRound will drop data < args.seqlen and truncate data to args.seqlen
      data = ["AutoRound is an advanced quantization algorithm for low-bits LLM inference" * 240]
      return data
  
  
  def customized_data_with_tokenizer(tokenizer, seqlen=2048):
      # Import notice!!! AutoRound will drop data < args.seqlen
      data = ["AutoRound is an advanced quantization algorithm for low-bits LLM inference" * 240]
      tokens = []
      for d in data:
          token = tokenizer(d, truncation=True, max_length=seqlen, return_tensors="pt").data
          tokens.append(token)
      return tokens
  

Dataset operations

Dataset combination:We support combination of different datasets and parametrization of calibration datasets by using --dataset ./tmp.json,NeelNanda/pile-10k:num=256,mbpp:num=128. Both local calibration file and huggingface dataset are supported. You could specify splits of a dataset by setting split=split1+split2.

Samples concatenation: An optional setting allows users to concatenate calibration samples using --dataset NeelNanda/pile-10k:concat=True. All samples will be concatenated first, then split into chunks of seqlen length.

Apply chat template: Using --dataset NeelNanda/pile-10k:apply_chat_template enables application of a chat template to the calibration data before tokenization. This is commonly used for instruct-style models during generation. To customize the system prompt, use:--dataset 'NeelNanda/pile-10k:apply_chat_template:system_prompt="You are a helpful assistant."'

Note: If the concatenation option is not enabled, samples shorter than args.seqlen will be dropped.

Please use ',' to split datasets, ':' to split parameters of a dataset and '+' to add values for one targeted parameter.

3 Quantization

Supported Quantization Schemes

AutoRound supports several Schemes:

• W4A16(bits:4,group_size:128,sym:True,act_bits:16)
• W8A16(bits:8,group_size:128,sym:True,act_bits:16)
• W6A16(bits:6,group_size:128,sym:True,act_bits:16) — mlx format only
• W5A16(bits:5,group_size:128,sym:True,act_bits:16) — mlx format only
• W3A16(bits:3,group_size:128,sym:True,act_bits:16)
• W2A16(bits:2,group_size:128,sym:True,act_bits:16)
• GGUF:Q4_K_M(all Q_K,Q_0,Q*_1 provided by llamacpp are supported)
• Mixed Bits Weight only
• NVFP4(Experimental feature, recommend exporting to llm_compressor format.data_type nvfp4,act_data_type nvfp4,static_global_scale,group_size 16)
• MXFP4(Research feature, no real kernel, Standard MXFP4, data_type mxfp,act_data_type mxfp,bits 4, act_bits 4, group_size 32)
• MXINT4(Research feature, no real kernel, Standard MXINT4, data_type mxint,act_data_type mxint,bits 4, act_bits 4, group_size 32)
• MXFP4_RCEIL(Research feature,no real kernel, NVIDIA's variant, data_type mxfp,act_data_type mxfp_rceil,bits 4, act_bits 4, group_size 32)
• MXFP8(Research feature, no real kernel, data_type mxfp,act_data_type mxfp_rceil,group_size 32)
• FPW8A16(Research feature, no real kernel, data_type fp8,group_size 0->per tensor )
• FP8_STATIC(Research feature, no real kernel, data_type:fp8,act_data_type:fp8,group_size -1 ->per channel, act_group_size=0->per tensor)

Besides, you could modify the group_size, bits, sym and many other configs you want, though there are maybe no real kernels.

Supported Export Formats

You can use command auto_round list format to show all supported formats with support scheme.

AutoRound Format: This format is well-suited for CPU, Intel GPU, CUDA and HPU devices, 2 bits, as well as mixed-precision inference. [2,3,4,8] bits are supported. Please set --format auto_round

GGUF Format: Experimental feature. This format is well-suited for CPU devices and is widely adopted by the community. q*_k,q*_0,q*_1 are supported. Please set --format gguf:q4_k_m, --format gguf:q2_k_s, etc

AutoGPTQ Format: This format is well-suited for symmetric quantization on CUDA devices and is widely adopted by the community, [2,3,4,8] bits are supported. However, the asymmetric kernel has issues that can cause considerable accuracy drops, particularly at 2-bit quantization and small models. Besides, recently 3 bits may have some accuracy issues in Transformers. Please set --format auto_gptq

AutoAWQ Format: This format is well-suited for asymmetric 4-bit quantization on CUDA devices and is widely adopted within the community, only 4-bits quantization is supported. Please set --format auto_awq

LLM-Compressor Format: NVFP4, MXFP4(kernel in WIP), MXFP8 are supported. Please set --format llm_compressor

MLX Format[Experimental Feature]: This format targets Apple Silicon (M1/M2/M3/...) and is loaded directly by mlx-lm (text-only LLM) or mlx-vlm (vision/audio + language).

• Supports 2, 3, 4, 5, 6, 8 bits (5/6 bits are MLX-exclusive — GPTQ/AWQ have no standard packing for them).
• Native mixed-bit / mixed-group_size via layer_config or AutoScheme (--target_bits 3.5 --options "...");
• Use --format mlx for a native MLX checkpoint; use --format auto_round:mlx if you want HuggingFace transformers + AutoRound to load it (post-init repacks each layer into MLX QuantLinear on Darwin).
• Limitation: embedding layer quantization has not supported

Format and scheme support matrix

Gray indicates the absence of a kernel or the presence of only an inefficient/reference kernel. BF16 is mainly for AutoScheme

Hardware Compatibility

CPU, Intel GPU, HPU and CUDA for both quantization and inference. The MLX format is exclusive to Apple Silicon (macOS / Darwin) at inference time; quantization (export) itself can be done on any platform.

Environment Configuration

Before starting quantization, you may want to configure AutoRound's environment variables for optimal performance. For detailed information about available environment variables (logging levels, ModelScope integration, workspace settings, etc.), please refer to the Environment Variables Guide.

Command Line Usage

• AutoRound recipe:

  This setting offers a better trade-off between accuracy and tuning cost, and is recommended in all scenarios.

  auto-round --model Qwen/Qwen3-0.6B  --scheme "W4A16"  --format "auto_gptq,auto_awq,auto_round"
  

• AutoRoundBest recipe:

  This setting provides the best accuracy in most scenarios but is 4–5× slower than the standard AutoRound recipe. It is especially recommended for 2-bit quantization and is a good choice if sufficient resources are available.

  auto-round-best --model Qwen/Qwen3-0.6B  --scheme "W4A16"  --format "auto_gptq,auto_awq,auto_round"
  

• AutoRoundLight Settings:

  This setting offers the best speed (2-3X faster than AutoRound), but it may cause a significant accuracy drop for small models and 2-bit quantization. It is recommended for 4-bit settings and models larger than 3B

  auto-round-light --model Qwen/Qwen3-0.6B  --scheme "W4A16"  --format "auto_gptq,auto_awq,auto_round"
  

• AutoRoundOptRTN recipe (optimized RTN, without gradient computation):

  This setting runs the optimized RTN (Round-To-Nearest) path (iters=0 with disable_opt_rtn=False). It is calibration-free and several times faster than the default AutoRound recipe, while still applying AutoRound's RTN-side optimizations (e.g. improved scale/zero-point search and llamacpp-style refinements for GGUF). Recommended as a fast baseline when calibration data or tuning time is limited. See the OPT RTN Mode section for details.

  auto-round-opt-rtn --model Qwen/Qwen3-0.6B  --scheme "W4A16"  --format "auto_round"
  

• AutoRoundRTN recipe (pure RTN, calibration-free, no optimization):

  This setting runs pure RTN (iters=0 with disable_opt_rtn=True), without any AutoRound optimization. It is the fastest path and uses the least memory, but typically yields lower accuracy than auto-round-opt-rtn. When combined with a supported INT WOQ scheme, it is automatically routed through Model-Free Mode for minimal memory usage. Use this as a quick sanity-check or when you want a calibration-free baseline equivalent to traditional RTN.

  auto-round-rtn --model Qwen/Qwen3-0.6B  --scheme "W4A16"  --format "auto_round"
  

API usage

AutoRound API Usage

This setting offers a better trade-off between accuracy and tuning cost, and is recommended in all scenarios.

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"
ar = AutoRound(
    model_name_or_path,
    scheme="W4A16",
    # enable_torch_compile=True,
)

output_dir = "./tmp_autoround"
# format= 'auto_round'(default), 'auto_gptq', 'auto_awq'
ar.quantize_and_save(output_dir, format="auto_gptq,auto_awq,auto_round")

Mixed Bits Usage

AutoRound(>0.8) offers auto-scheme to generate mixed bits recipe autocially, please refer to AutoScheme section for more details.

Auto-GPTQ and Auto-AWQ only support a limited set of mixed-bit configurations. If you're unsure about the details, we recommend using the AutoRound format.

vLLM and SGLang fuse MoE and QKV layers, so it's recommended not to assign different bit widths to these layers.

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"

layer_config = {  #  Supports both full layer names and fuzzy (partial) matching
    "model.decoder.layers.6.self_attn.out_proj": {"bits": 8, "group_size": 32},
    "model.decoder.layers.*k_proj": {"bits": 2, "group_size": 32},
}
ar = AutoRound(
    model_name_or_path,
    layer_config=layer_config,
)

output_dir = "./tmp_autoround"
ar.quantize_and_save(output_dir, format="auto_round")

AutoRoundBest recipe

This setting provides the best accuracy in most scenarios but is 4–5× slower than the standard AutoRound recipe. It is especially recommended for 2-bit quantization and is a good choice if sufficient resources are available.

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"
ar = AutoRound(model=model_name_or_path, scheme="W4A16", nsamples=512, iters=1000, low_gpu_mem_usage=True)

output_dir = "./tmp_autoround"
ar.quantize_and_save(output_dir, format="auto_round")

AutoRoundLight recipe

This setting offers the best speed (2 - 3X faster than AutoRound), but it may cause a significant accuracy drop for small models and 2-bit quantization. It is recommended for 4-bit settings and models larger than 3B.

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"

ar = AutoRound(
    model=model_name_or_path,
    scheme="W4A16",
    iters=50,
    lr=5e-3,
)

output_dir = "./tmp_autoround"
ar.quantize_and_save(output_dir, format="auto_round")

Recipe recommendation

In conclusion, we recommend using auto-round for W4A16 and auto-round-best for W2A16. However, you may adjust the configuration to suit your specific requirements and available resources.

W4G128 Average Accuracy of 13 tasks and Time Cost Results(Testing was conducted on the Nvidia A100 80G using the version of PyTorch 2.6.0 with enable_torch_compile):

AWQ Algorithm

Experimental feature: our current implementation does not apply weight clipping yet, so accuracy may drop compared to the original AWQ algorithm.

AWQ (Activation-Aware Weight Quantization) is available as an alternative quantization algorithm. AWQ protects salient weight channels by analyzing activation patterns and applying channel-wise scaling before standard RTN quantization.

The canonical AWQ deployment path is W4A16 served by vLLM's AWQ/Marlin CUDA kernels. W8A8 with AWQ smoothing can also be served via vLLM's compressed_tensors backend (cutlass INT8 GEMM).

CLI Usage

auto-round --model Qwen/Qwen3-0.6B --scheme "W4A16" --algorithm awq --format "auto_round"

AWQ-specific options:

• --duo_scaling: Use both activations and weights for scaling. Options: true, false, or both (searches both modes and picks the best). (default: True).
• --n_grid: Number of grid points for scaling ratio search (default: 20).

API Usage

from auto_round import AutoRound

ar = AutoRound(
    "Qwen/Qwen3-0.6B",
    scheme="INT8",
    algorithm="awq",
)

output_dir = "./tmp_awq"
ar.quantize_and_save(output_dir, format="auto_round:llm_compressor")

AutoScheme

AutoScheme automatically generates adaptive mixed-bit and mixed-data-type quantization recipes. For accuracy results, see AutoScheme Accuracy Report.

Note: Mixed-data-types are supported during tuning, but cannot be exported to real models at this time.

CLI Usage

• --iters 0: RTN. Fast (seconds to minutes).
• --iters 200: Tuning-aware scheme selection. More accurate but much slower.

auto_round \
  --model_name  $model_name \
  --avg_bits 6 \
  --options "mxfp4,mxfp8" \
  --ignore_scale_zp_bits \
  --iters 0 \
  --format fake 

API Usage

avg_bits= 3.0
scheme = AutoScheme(avg_bits=avg_bits, options=("W2A16G64“, "W4A16","W8A16"))
ar = AutoRound(model=model_name, scheme=scheme, iters=0, nsamples=1)
ar.quantize_and_save()

Hyperparameters in AutoScheme

avg_bits(float) Target average bits for the whole model; only layers to be quantized will be counted in the average bits calculation.

options(Union[str, list[Union[QuantizationScheme, str]]) the options of quantization schemes to choose from. It could be a string like "W4A16", or a list of strings or QuantizationScheme objects.

ignore_scale_zp_bits(bool) Whether to ignore the bits of scale and zero point in average bits calculation. Default is False.

device_map (Optional[str,dict,torch.device]) only supported in API now, as auto-scheme used more VRAM than auto-round tuning, so you could set a different device_map for it.

shared_layers (Optional[Iterable[Iterable[str]]]) only supported in API now

batch_size (Optional[int]) could be set to 1 to reduce VRAM but increase time cost

low_gpu_mem_usage(bool=True) whether to reduce gpu memory usage at the cost of more time cost

In some serving frameworks, certain layers (e.g., QKV or MoE) are fused to accelerate inference. These fused layers may require the same data type and bit configuration. The shared_layers option simplifies this setup by supporting both regex and full-name matching. Note that regex matching is applied in a block-wise manner.

MoE expert layers are automatically grouped per block — all expert projections (gate/up/down across all experts) within the same transformer block are treated as a single entity during DP optimization. This means they share the same quantization scheme and their losses/numel are summed together. No manual shared_layers configuration is needed for expert layers.

from auto_round import AutoRound, AutoScheme

shared_layers = [
    ["*.self_attn.k_proj", "v_proj", "q_proj", "out_proj"],
    ("model.decoder.layers.6.fc1", "model.decoder.layers.6.fc2"),
    ("fc1", "fc2"),
]
target_bits = 5.0
model_name = "Qwen/Qwen3-0.6B"
scheme = AutoScheme(avg_bits=target_bits, options=("W4A16", "MXFP8"), shared_layers=shared_layers)
ar = AutoRound(model=model_name, scheme=scheme, iters=0, nsamples=1)
model, layer_config = ar.quantize()

Besides, if you want to fix the scheme for some layers, you could set it via layer_config in AutoRound API.

from auto_round import AutoRound, AutoScheme

model_name = "Qwen/Qwen3-8B"
avg_bits = 3.0
scheme = AutoScheme(avg_bits=avg_bits, options=("GGUF:Q2_K_S", "GGUF:Q4_K_S"), ignore_scale_zp_bits=True)
layer_config = {"lm_head": "GGUF:Q6_K"}

ar = AutoRound(model=model_name, scheme=scheme, layer_config=layer_config, iters=0)
ar.quantize_and_save()

AutoScheme Cost

We tested it on Nvidia A100 80G using torch v2.8.

We will try to optimize the RAM usage in the future. The RAM usage is about 1.1-1.5x of the model's BF16 size

Limitations

Embedding layer is not supported in AutoScheme, it will use the best scheme in options.

When using AutoScheme with model_free=True, only INT (W2A16/W4A16/W8A16) and MXFP (MXFP4/MXFP8) option families are supported. Options like W3A16, GGUF:*, and NVFP4 will raise a ValueError. INT and MXFP families cannot be mixed in the same AutoScheme.

AWQ Quantization Algorithm

AWQ (algorithm="awq") is a pre-processing quantization algorithm that analyzes activation patterns and applies channel-wise scaling to protect salient weights. It runs BEFORE the actual quantization (RTN by default, or auto_round/SignRound).

CLI Usage

# AWQ + default RTN (iters=0 auto-selected)
auto-round --model Qwen/Qwen3-0.6B --algorithm awq --scheme W4A16

# AWQ + AutoRound optimization
auto-round --model Qwen/Qwen3-0.6B --algorithm awq,auto_round --scheme W4A16

# AWQ flags
--duo-scaling true|false|both  (default: true)
--n-grid 20                    (default: 20)

API Usage

from auto_round import AutoRound
from auto_round.algorithms.quantization.awq.config import AWQConfig
from auto_round.algorithms.quantization.sign_round.config import SignRoundConfig

# AWQ + default RTN (simplest)
ar = AutoRound(model, tokenizer, algorithm="awq", scheme="W4A16")

# AWQ + AutoRound via alg_configs (explicit pipeline)
ar = AutoRound(model, tokenizer, alg_configs=[AWQConfig(), SignRoundConfig(iters=200)], scheme="W4A16")
ar.quantize_and_save(output_dir="./qmodel")

Important Note: algorithm="awq" (quantization algorithm) and format="auto_awq" (export format) are independent. You can use:

• algorithm="awq" + format="auto_round": AWQ smoothing + AutoRound packing
• algorithm="auto_round" + format="auto_awq": No AWQ smoothing + AutoAWQ packing

OPT RTN Mode

AutoRound also supports Optimized RTN (Round-To-Nearest) mode for fast, calibration-free baseline quantization. Setting iters=0 tp enable it and we recommend using group_size=32 for better results. Check accuracy comparison between RTN and OPT RTN mode

For the GGUF format, we have optimized the RTN algorithm inspired by llamacpp. To use the original (pure) RTN algorithm instead, enable the --disable_opt_rtn option.

CLI Usage

Two dedicated CLI entry points are provided as shortcuts:

• auto-round-opt-rtn — equivalent to auto-round --iters 0 --enable_opt_rtn (optimized RTN, recommended).
• auto-round-rtn — equivalent to auto-round --iters 0 --disable_opt_rtn (pure RTN, no optimization; auto-routes to Model-Free Mode for supported INT WOQ schemes).

# Optimized RTN (recommended fast baseline)
auto-round-opt-rtn --model Qwen/Qwen3-0.6B --scheme "W4A16" --format "auto_round"

# Pure RTN (fastest, lowest memory; baseline quality)
auto-round-rtn --model Qwen/Qwen3-0.6B --scheme "W4A16" --format "auto_round"

API Usage

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"
ar = AutoRound(
    model=model_name_or_path,
    scheme="W4A16",
    iters=0,
)

output_dir = "./tmp_autoround"
ar.quantize_and_save(output_dir, format="auto_round")

Model-Free Mode

Model-free mode performs RTN WOQ quantization without loading the full model into memory. It downloads safetensors files directly, quantizes each Linear weight tensor shard-by-shard, and saves the packed result. This is useful when you want fast, no-calibration quantization with minimal resource requirements.

Auto-enabled by default. As of v0.13, when you pass --iters 0 --disable_opt_rtn together with a supported INT WOQ or MXFP scheme, the CLI automatically takes the model-free path. This is bit-exactly equivalent to the regular --iters 0 --disable_opt_rtn flow but uses far less memory. Use --disable_model_free to opt out and force the original flow.

Key features:

• No model object required – only config.json and safetensors files are needed
• Low disk memory required (If no local model files) – downloads and quantizes one shard at a time, deleting the source shard after processing
• Per-layer configuration – supports --layer_config for per-layer bit-width overrides and --ignore_layers to keep specific layers in full precision
• Predefined ignore layers – automatically skips model-specific layers (e.g., MoE gates, MTP layers) based on config detection
• Bit-exact parity with the standard --iters 0 --disable_opt_rtn flow for all supported schemes
• AutoScheme integration – pass an AutoScheme object as scheme to get automatic mixed-bit selection followed by shard-by-shard packing (two-phase: score with model briefly loaded, then free and pack)

# Easiest: --iters 0 --disable_opt_rtn auto-routes to model-free
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --scheme W4A16 \
  --iters 0 --disable_opt_rtn \
  --output_dir ./int4-llama

# Equivalent explicit invocation
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --model_free \
  --scheme W4A16 \
  --output_dir ./int4-llama

# Opt out of auto-routing and use the regular flow instead
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --scheme W4A16 \
  --iters 0 --disable_opt_rtn --disable_model_free \
  --output_dir ./int4-llama

# With per-layer configuration and ignored layers
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --model_free \
  --scheme W4A16 \
  --group_size 32 \
  --asym \
  --layer_config "{k_proj:{bits:8},v_proj:{bits:8}}" \
  --ignore_layers "mlp" \
  --output_dir ./int4-llama

# MXFP4 quantization
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --model_free \
  --scheme MXFP4 \
  --output_dir ./mxfp4-llama

# MXFP8 quantization
auto_round meta-llama/Llama-3.2-1B-Instruct \
  --model_free \
  --scheme MXFP8 \
  --output_dir ./mxfp8-llama

from auto_round import AutoRound

AutoRound(
    model="meta-llama/Llama-3.2-1B-Instruct",
    scheme="W4A16",  # Or a QuantizationScheme instance for custom group_size / sym.
    layer_config={
        ".*k_proj": {"bits": 8, "group_size": 32},
        ".*v_proj": {"bits": 8, "group_size": 32},
    },
    ignore_layers="mlp",
    model_free=True,
).quantize_and_save("./int4-llama")

GGUF format

Experimental feature. This format is well-suited for CPU devices and is widely adopted by the community.

The optimized RTN mode is suggested (--iters 0) for all bits other than 3 bits.

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"
ar = AutoRound(
    model=model_name_or_path,
)
output_dir = "./tmp_autoround"
ar.quantize_and_save(output_dir, format="gguf:q4_k_m")  #  gguf:q*_k_s,gguf:q*_k_0,gguf:q*_k_1,

Quantization Costs

Testing was conducted on the Nvidia A100 80G using the nightly version of PyTorch 2.6.0.dev20241029+cu124. Please note that data loading and packing costs have been excluded from the evaluation. We recommend enabling torch.compile for PyTorch versions 2.6 and above.

To optimize GPU memory usage, in addition to activating low_gpu_mem_usage, you can set gradient_accumulate_steps=8 and a batch_size=1, though this may increase tuning time.

The 3B and 14B models were evaluated on Qwen 2.5, the 8X7B model is Mixtral, while the remaining models utilized LLaMA 3.1.

W4G128 Quantization Time and Memory Usage (Intel GPU B60 24G) Testing was conducted on the Intel GPU B60 24G using the release version of PyTorch 2.11.0+xpu. Please note that data loading and packing costs have been excluded from the evaluation. Time and memory usage were measured using Qwen2.5-series models.

Device/Multi-GPU setting in Quantization

The tuning device is specified using the device_map argument in AutoRound API, not through the device_map parameter used by Transformers.from_pretrained.

AutoRound tunes the model in a block-by-block manner. Although the block size is much smaller than the model size, it still requires a significant amount of GPU memory for tuning—typically 10 times the block size. This can lead to out-of-memory (OOM) errors when working with extremely large models.

For strategies to reduce GPU memory usage, please refer to the [Reduced GPU Memory Usage](###Adjust Hyperparameters) section below, where you can adjust hyperparameters to optimize memory consumption.

If adjusting hyperparameters does not resolve the issue a, a simple solution is just adding more devices in device_map, for example,

from auto_round import AutoRound

model_name_or_path = "Qwen/Qwen3-0.6B"
ar = AutoRound(
    model=model_name_or_path,
    device_map="0,1,2,3"
)

or

CUDA_VISIBLE_DEVICES=0,1,2,3 auto-round --model "Qwen/Qwen3-0.6B" --scheme "W4A16" --device_map "auto"

There are typically two scenarios that require multi-GPU tuning: one is the calibration phase mainly for lm-head quantization, and the other is quantizing extremely large models (e.g., models larger than 100 GB).

Enable multiple gpus calibration in lm_head quantization

For LM head tuning, AutoRound needs to cache the inputs to the lm-head, which requires the entire model to reside on the GPU for efficient calibration. If there is no enough VRAM, some layers will fallback to RTN mode

Manually set the device_map

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "opensourcerelease/DeepSeek-R1-bf16"

tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True, torch_dtype="auto")

block = model.model.layers
device_map = {}

for n, m in block.named_modules():
    if type(m) == torch.nn.Linear:
        if "experts" in n and ("shared_experts" not in n) and int(n.split(".")[-2]) < 63:
            device = "cuda:1"
        elif (
            "experts" in n
            and ("shared_experts" not in n)
            and int(n.split(".")[-2]) >= 63
            and int(n.split(".")[-2]) < 128
        ):
            device = "cuda:2"
        elif (
            "experts" in n
            and ("shared_experts" not in n)
            and int(n.split(".")[-2]) >= 128
            and int(n.split(".")[-2]) < 192
        ):
            device = "cuda:3"
        elif "experts" in n and ("shared_experts" not in n) and int(n.split(".")[-2]) >= 192:
            device = "cuda:4"
        else:
            device = "cuda:0"
        n = n[2:]

        device_map.update({n: device})

from auto_round import AutoRound

autoround = AutoRound(
    model=model,
    tokenizer=tokenizer,
    device_map=device_map,
    nsamples=512,
    batch_size=4,
    low_gpu_mem_usage=True,
    seqlen=2048,
)
autoround.quantize()
autoround.save_quantized(format="auto_awq", output_dir="tmp_autoround")

Adjust Hyperparameters

• Reduced GPU Memory Usage:

  • set enable_torch_compile to True

  • enable low_gpu_mem_usage(more tuning cost)

  • set --bs 1 --gradient_accumulate_steps 8 (more tuning cost)

  • reduce the bs to 4(potential accuracy drop)

  • reduce the seqlen to 512 (potential accuracy drop)

  • or combine them

• Reduced CPU Memory Usage :

  • Enable low_cpu_mem_usage (experimental): Only one export format is supported. The quantized model is saved immediately after each block is packed, reducing peak CPU memory usage.

  • Trigger immediate packing: Packing will be triggered immediately when using the command-line interface or the quantize_and_save API, as long as only one export format is specified.

• Speedup the tuning:

  • set enable_torch_compile to True

  • use auto-round-light configuration

  • reduce the seqlen to 512(potential large accuracy drop for some scenarios)

  • reduce the train bs to 4(little accuracy drop. )

  • or combine them

• Enable quantized lm-head:

  Currently only support in AutoRound format inference for this config

  auto-round --model_name Qwen/Qwen3-0.6B  --scheme "W4A16" --quant_lm_head --format "auto_round"
  

• Utilize the AdamW Optimizer:

  Include the flag --adam. Note that AdamW is less effective than sign gradient descent in many scenarios we tested.

Rotation (Experimental)

⚠️ Experimental feature: Rotation transform is still experimental. Inference relies on forward hooks, which are currently only supported by the Hugging Face Transformers backend, so rotated models may run slower than native (non-rotated) models.

Rotation redistributes outliers in weights and activations before quantization, making the distribution more uniform and quantization-friendly. It is most useful for aggressive low-bit schemes such as MXFP4, NVFP4 and W4A4.

AutoRound applies rotation through the rotation_config argument. The "quarot" preset — deterministic Hadamard rotation (QuaRot / SpinQuant), no training and no calibration data — is recommended for most use cases.

API Usage

from auto_round import AutoRound

model_name = "Qwen/Qwen3-0.6B"

# QuaRot preset: deterministic Hadamard, no training
ar = AutoRound(model_name, scheme="MXFP4", rotation_config="quarot")
ar.quantize_and_save(output_dir="./Qwen3-0.6B-mxfp4-quarot", format="auto_round")

Quantized models with rotation are saved and loaded transparently — rotation matrices and hooks are restored automatically on load, so inference needs no extra steps.

For rotation positions (R1–R4), full configuration options, deterministic vs random Hadamard, trainable SpinQuant, the per-linear block rotation variant, and save/load internals, see Rotation Details.

4 Inference

AutoRound automatically selects the best available backend based on the installed libraries and prompts the user to install additional libraries when a better backend is found.

Please avoid manually moving the quantized model to a different device (e.g., model.to('cpu')) during inference, as this may cause unexpected exceptions.

CPU

Supports 2, 4, and 8 bits. We recommend using auto-round-lib (ark) for inference. When using the ark backend, ensure that your PyTorch version is >= 2.8.0.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "OPEA/Qwen2.5-1.5B-Instruct-int4-sym-inc"
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="cpu", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

Intel GPU

Supports 4,8 bits. When using the ark backend, PyTorch (torch) >= 2.8.0 is required. We recommend using auto-round-lib (ark) for inference.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "OPEA/Qwen2.5-1.5B-Instruct-int4-sym-inc"
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="xpu", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

CUDA

Supports 2, 3, 4, and 8 bits. We recommend using GPTQModel for 4 and 8 bits inference.

from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "OPEA/Qwen2.5-1.5B-Instruct-int4-sym-inc"
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="cuda", torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

HPU

docker image with Gaudi Software Stack is recommended. More details can be found in Gaudi Guide.

import habana_frameworks.torch.core as htcore
import habana_frameworks.torch.hpu as hthpu
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_name = "Intel/Qwen2-7B-int4-inc"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name).to("hpu").to(torch.bfloat16)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

Specify Inference Backend

AutoRound automatically selects the backend for each layer based on compatibility. In general, the priority order is Marlin > ExLLaMAV2 > Triton, but the final choice depends on factors such as group size, bit width, packing format, hardware device, and other implementation details.

The backend may not always be the most suitable for certain devices. You can specify your preferred backend such as "ark" for CPU and Intel GPU, "marlin/exllamav2/triton" for CUDA, according to your needs or hardware compatibility. Please note that additional corresponding libraries may be required.

from transformers import AutoModelForCausalLM, AutoTokenizer, AutoRoundConfig

model_name = "OPEA/Qwen2.5-1.5B-Instruct-int4-sym-inc"
quantization_config = AutoRoundConfig(backend="ark")
model = AutoModelForCausalLM.from_pretrained(
    model_name, device_map="cpu", quantization_config=quantization_config, torch_dtype="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

Convert GPTQ/AWQ to AutoRound

Most GPTQ/AWQ models can be converted to the AutoRound format for better compatibility and support with Intel devices. Please note that the quantization config will be changed if the model is serialized.

from transformers import AutoModelForCausalLM, AutoTokenizer, AutoRoundConfig

model_name = "ybelkada/opt-125m-gptq-4bit"
quantization_config = AutoRoundConfig()
model = AutoModelForCausalLM.from_pretrained(
    model_name, device_map="cpu", quantization_config=quantization_config, torch_dtype="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
text = "There is a girl who likes adventure,"
inputs = tokenizer(text, return_tensors="pt").to(model.device)
print(tokenizer.decode(model.generate(**inputs, max_new_tokens=50, do_sample=False)[0]))

5 Evaluation

AutoRound leverages lm-eval-harness for evaluation. If --tasks is not specified, a set of default tasks (typically 10+ common benchmarks) will be automatically used.

Single GPU Evaluation

HF Backend (default):

auto-round --model Qwen/Qwen3-0.6B --bits 4 --format "auto_round,auto_gptq" --tasks mmlu

vLLM Backend:

auto-round --model Qwen/Qwen3-0.6B --bits 4 --format "auto_round,auto_gptq" --tasks mmlu --eval_backend vllm

Multi-GPU Evaluation

HF Backend:

auto-round --model="your_model_path" --eval --device_map 0,1 --tasks lambada_openai --eval_bs 16

vLLM Backend (Option 1 - using --device_map):

auto-round "your_model_path" --eval --device_map 0,1 --tasks lambada_openai --eval_backend vllm

vLLM Backend (Option 2 - manual configuration):

CUDA_VISIBLE_DEVICES=0,1 auto-round "your_model_path" --eval --tasks lambada_openai --eval_backend vllm --vllm_args="tensor_parallel_size=2,gpu_memory_utilization=0.8"

Important Notes

• Use the --eval flag to evaluate models directly. This supports both original and quantized models.
• The --eval_task_by_task option helps handle task failures by evaluating tasks sequentially. This only applies to the HF backend.
• When multiple formats are exported, the last format in the list will be used for evaluation.
• For vLLM backend, you can use --device 0,1,2 to specify GPU devices. This will automatically set CUDA_VISIBLE_DEVICES and configure tensor_parallel_size based on the number of devices. Alternatively, you can manually set these via environment variables and --vllm_args.

6 Known Issues

Randomness in quantization may affect tuning results for some models, set enable_deterministic_algorithms=True to ensure reproducibility.

Some VLMs require manual support.

Mamba is not supported.