Instructions to use Qwen/Qwen3-235B-A22B-GPTQ-Int4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Qwen/Qwen3-235B-A22B-GPTQ-Int4 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Qwen/Qwen3-235B-A22B-GPTQ-Int4")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-235B-A22B-GPTQ-Int4")
model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen3-235B-A22B-GPTQ-Int4", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Qwen/Qwen3-235B-A22B-GPTQ-Int4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Qwen/Qwen3-235B-A22B-GPTQ-Int4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen3-235B-A22B-GPTQ-Int4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Qwen/Qwen3-235B-A22B-GPTQ-Int4

SGLang

How to use Qwen/Qwen3-235B-A22B-GPTQ-Int4 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Qwen/Qwen3-235B-A22B-GPTQ-Int4" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen3-235B-A22B-GPTQ-Int4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Qwen/Qwen3-235B-A22B-GPTQ-Int4" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Qwen/Qwen3-235B-A22B-GPTQ-Int4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Qwen/Qwen3-235B-A22B-GPTQ-Int4 with Docker Model Runner:
```
docker model run hf.co/Qwen/Qwen3-235B-A22B-GPTQ-Int4
```

feihu.hf commited on May 11, 2025

Commit

1d5ada7

1 Parent(s): 44823bd

update README

Browse files

Files changed (4) hide show

LICENSE +0 -0
README.md +5 -5
generation_config.json +12 -5
tokenizer_config.json +0 -1

LICENSE CHANGED Viewed

File without changes

README.md CHANGED Viewed

@@ -322,10 +322,10 @@ YaRN is currently supported by several inference frameworks, e.g., `transformers
 | Mode | QUANTIZATION TYPE | LiveBench 2024-11-25 | GPQA | MMLU-Redux | AIME24 |
 | --- | --- | --- | --- | --- | --- |
-| Thinking | bf16 | 74.3 | 65.8 | 89.5 | 80.4 |
-| Thinking | GPTQ-int4 | 71.5 | 60.1 | 88.8 | 79.5 |
-| Non-Thinking | bf16 | 59.4 | 54.8 | 84.1 | - |
-| Non-Thinking | GPTQ-int4 | 57.2 | 50.4 | 83.5 | - |
 ## Best Practices
@@ -356,4 +356,4 @@ If you find our work helpful, feel free to give us a cite.
     month  = {April},
     year   = {2025}
 }
-```

 | Mode | QUANTIZATION TYPE | LiveBench 2024-11-25 | GPQA | MMLU-Redux | AIME24 |
 | --- | --- | --- | --- | --- | --- |
+| Thinking | bf16 | 77.1 | 71.1 | 92.7 | - |
+| Thinking | GPTQ-int4 | 75.1 | 71.9 | 92.0 | - |
+| Non-Thinking | bf16 | 62.5 | 62.9 | 89.2 | - |
+| Non-Thinking | GPTQ-int4 | 61.1 | 62.8 | 89.0 | - |
 ## Best Practices
     month  = {April},
     year   = {2025}
 }
+```

generation_config.json CHANGED Viewed

@@ -1,6 +1,13 @@
 {
-  "_from_model_config": true,
-  "eos_token_id": 151645,
-  "pad_token_id": 151643,
-  "transformers_version": "4.51.3"
-}

 {
+    "bos_token_id": 151643,
+    "do_sample": true,
+    "eos_token_id": [
+        151645,
+        151643
+    ],
+    "pad_token_id": 151643,
+    "temperature": 0.6,
+    "top_k": 20,
+    "top_p": 0.95,
+    "transformers_version": "4.51.0"
+}

tokenizer_config.json CHANGED Viewed

@@ -231,7 +231,6 @@
   "clean_up_tokenization_spaces": false,
   "eos_token": "<|im_end|>",
   "errors": "replace",
-  "extra_special_tokens": {},
   "model_max_length": 131072,
   "pad_token": "<|endoftext|>",
   "split_special_tokens": false,

   "clean_up_tokenization_spaces": false,
   "eos_token": "<|im_end|>",
   "errors": "replace",
   "model_max_length": 131072,
   "pad_token": "<|endoftext|>",
   "split_special_tokens": false,