---
library_name: transformers
tags:
- torchao
- code
- math
- chat
license: apache-2.0
language:
- multilingual
base_model:
- Qwen/Qwen3-32B
pipeline_tag: text-generation
---
[Qwen3-32B](https://huggingface.co/Qwen/Qwen3-32B) model quantized with [torchao](https://huggingface.co/docs/transformers/main/en/quantization/torchao) float8 dynamic activation and float8 weight quantization (per row granularity), by PyTorch team. Use it directly, or serve using [vLLM](https://docs.vllm.ai/en/latest/) with 47% VRAM reduction, around 1.5x speedup and little to no accuracy impact on H100.
# Inference with vLLM
```Shell
# Server
VLLM_DISABLE_COMPILE_CACHE=1 vllm serve pytorch/Qwen3-32B-float8dq --tokenizer Qwen/Qwen3-32B -O3
```
```Shell
# Client
curl http://localhost:8000/v1/chat/completions -H "Content-Type: application/json" -d '{
"model": "pytorch/Qwen3-32B-float8dq",
"messages": [
{"role": "user", "content": "Give me a short introduction to large language models."}
],
"temperature": 0.6,
"top_p": 0.95,
"top_k": 20,
"max_tokens": 32768
}'
```
# Inference with transformers
```Py
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "pytorch/Qwen3-32B-float8dq"
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 ()
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
```
# Quantization Recipe
Install the required packages:
```Shell
pip install git+https://github.com/huggingface/transformers@main
pip install torchao
pip install torch
pip install accelerate
```
Use the following code to get the float8 model using torchao library:
```Py
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TorchAoConfig
model_id = "Qwen/Qwen3-32B"
from torchao.quantization import Float8DynamicActivationFloat8WeightConfig, PerRow
quant_config = Float8DynamicActivationFloat8WeightConfig(granularity=PerRow())
quantization_config = TorchAoConfig(quant_type=quant_config)
quantized_model = AutoModelForCausalLM.from_pretrained(
model_id,
device_map="auto",
torch_dtype=torch.bfloat16,
quantization_config=quantization_config,
)
tokenizer = AutoTokenizer.from_pretrained(model_id)
```
Optionally, upload to your HF hub
```Py
USER_ID = "YOUR_USER_ID"
MODEL_NAME = model_id.split("/")[-1]
save_to = f"{USER_ID}/{MODEL_NAME}-float8dq"
quantized_model.push_to_hub(save_to, safe_serialization=False)
tokenizer.push_to_hub(save_to)
```
# Model Quality
We rely on [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness) to evaluate the quality of the quantized model.
| Benchmark | | |
|----------------------------------|----------------|---------------------------|
| | Qwen3-32B | Qwen3-32B-float8dq |
| **General** | | |
| mmlu | 80.71 | 80.67 |
| bbh | 37.49 | 38.01 |
| **Multilingual** | | |
| mgsm_en_cot_es | 58.4 | 52.0 |
| **Math** | | |
| gpqa_main_zeroshot | 41.96 | 42.63 |
| **Overall** | 54.64 | 53.33 |
Reproduce Model Quality Results
Need to install lm-eval from source:
https://github.com/EleutherAI/lm-evaluation-harness#install
## baseline
```Shell
lm_eval --model hf --model_args pretrained=Qwen/Qwen3-32B --tasks mmlu --device cuda:0 --batch_size 8
```
## float8 dynamic quantization (float8dq)
```Shell
export MODEL=pytorch/Qwen3-32B-float8dq
# or
# export MODEL=Qwen/Qwen3-32B
lm_eval --model hf --model_args pretrained=$MODEL --tasks mmlu --device cuda:0 --batch_size 8
```
# Memory Usage
| Memory (tested on H100) | | |
|----------------------------------|----------------|-------------------------------|
| | Qwen3-32B | Qwen3-32B-float8dq |
| Peak Memory | 65.72 GB | 34.54 GB (47.44% reduction) |
Reproduce Peak Memory Usage Results
Code
```Py
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "Qwen/Qwen3-32B" # pytorch/Qwen3-32B-float8dq
# load the tokenizer and the model
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
torch.cuda.reset_peak_memory_stats()
# prepare the model input
prompt = "Give me a short introduction to large language model."
messages = [
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True,
enable_thinking=True # Switches between thinking and non-thinking modes. Default is True.
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
# conduct text completion
generated_ids = model.generate(
**model_inputs,
max_new_tokens=32768
)
output_ids = generated_ids[0][len(model_inputs.input_ids[0]):].tolist()
# parsing thinking content
try:
# rindex finding 151668 ()
index = len(output_ids) - output_ids[::-1].index(151668)
except ValueError:
index = 0
thinking_content = tokenizer.decode(output_ids[:index], skip_special_tokens=True).strip("\n")
content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")
print("thinking content:", thinking_content)
print("content:", content)
mem = torch.cuda.max_memory_reserved() / 1e9
print(f"Peak Memory Usage: {mem:.02f} GB")
```
# Model Performance
| Benchmark (Tested on H100) | | |
|----------------------------------|----------------|-------------------------------|
| | Qwen3-32B | Qwen3-32B-float8dq |
| latency (batch_size=1) | 9.1s | 5.77s (1.58x speedup) |
| latency (batch_size=128) | 12.45s | 8.40s (1.48x speedup) |
Reproduce latency benchmarks
**1. Setup**
```Shell
git clone git@github.com:vllm-project/vllm.git
cd vllm
VLLM_USE_PRECOMPILED=1 pip install --editable .
```
**2. Latency benchmarking**
```Shell
export MODEL=Qwen/Qwen3-32B # or pytorch/Qwen3-32B-float8dq
VLLM_DISABLE_COMPILE_CACHE=1 python benchmarks/benchmark_latency.py --input-len 256 --output-len 256 --model $MODEL --batch-size 1
```
# Disclaimer
PyTorch has not performed safety evaluations or red teamed the quantized models. Performance characteristics, outputs, and behaviors may differ from the original models. Users are solely responsible for selecting appropriate use cases, evaluating and mitigating for accuracy, safety, and fairness, ensuring security, and complying with all applicable laws and regulations.
Nothing contained in this Model Card should be interpreted as or deemed a restriction or modification to the licenses the models are released under, including any limitations of liability or disclaimers of warranties provided therein.