Instructions to use pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210 with Transformers:

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

pipe = pipeline("text-generation", model="pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210")
model = AutoModelForMultimodalLM.from_pretrained("pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210")
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 pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210

SGLang

How to use pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210 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 "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210" \
    --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": "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210",
		"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 "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210" \
        --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": "pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210 with Docker Model Runner:
```
docker model run hf.co/pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210
```

4bit Quantized Model: Qwen2.5-32B-Instruct

A 4bit quantized variant of /mnt/d/Development/Libraries/Qwen2.5-32B-Instruct, optimized to reduce memory footprint and accelerate inference while maintaining high output similarity.

Overview

This checkpoint was quantized using BitsAndBytes and evaluated with standard text similarity metrics.

Model Architecture

Attribute	Value
Model class	Qwen2ForCausalLM
Number of parameters	17,161,065,472
Hidden size	5120
Number of layers	64
Attention heads	40
Vocabulary size	152064
Compute dtype	bfloat16

Quantization Configuration

{
  "quant_method": "bitsandbytes",
  "_load_in_8bit": false,
  "_load_in_4bit": true,
  "llm_int8_threshold": 6.0,
  "llm_int8_skip_modules": null,
  "llm_int8_enable_fp32_cpu_offload": false,
  "llm_int8_has_fp16_weight": false,
  "bnb_4bit_quant_type": "nf4",
  "bnb_4bit_use_double_quant": true,
  "bnb_4bit_compute_dtype": "bfloat16",
  "bnb_4bit_quant_storage": "uint8",
  "load_in_4bit": true,
  "load_in_8bit": false
}

Intended Use

Research and experimentation.
Instruction-following tasks in resource-constrained environments.
Demonstrations of quantized model capabilities.

Limitations

May reproduce biases from the original model.
Quantization may reduce generation diversity and factual accuracy.
Not intended for production without additional evaluation.

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210")
model = AutoModelForCausalLM.from_pretrained("pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210", device_map="auto")

prompt = "Explain the concept of reinforcement learning."
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Generation Settings

This model produces best results when generated with:

temperature: 0.3
top_p: 0.9

Model Files Metadata

Filename	Size (bytes)	SHA-256
`model-00001-of-00004.safetensors`	4,933,190,348	`b2a0e8a735e99b3a59bb3139541c444808aff3793a28c314c0f02bf17a00b5f7`
`model-00002-of-00004.safetensors`	4,958,587,236	`fd4b028d13261c8da0e29ed57b95189d666f62f3e8d4ab232c17c4e4e131543a`
`model-00003-of-00004.safetensors`	4,999,136,184	`0446d1c6da46a5daea91bed161fd62f2f48a658d879f58a14b7ab5528eb66935`
`model-00004-of-00004.safetensors`	4,324,534,021	`39002c4ed64520809793fb2b2023caf9bdbf0914feb4786d553c418139457018`
`quant_config.json`	426	`1bd2332861a3d1a8f387a9d04a1432b5bb57dec1a112ab6cfe594f67c5e66823`

Notes

Produced on 2026-05-27T12:33:55.921152.
Quantized automatically using BitsAndBytes.

Intended primarily for research and experimentation.

Citation

Qwen2.5-32B-Instruct

Qwen2.5 Technical Report

License

This model is distributed under the apache-2.0 license, consistent with the original /mnt/d/Development/Libraries/Qwen2.5-32B-Instruct.

Model Card Authors

This quantized model was prepared by PBH Applied Systems.

Downloads last month: 42

Safetensors

Model size

34B params

Tensor type

F32

F16

Paper for pbhappliedsystems/Qwen2.5-32B-Instruct-4bit-20260527_122210

Qwen2.5 Technical Report

Paper • 2412.15115 • Published Dec 19, 2024 • 380