Instructions to use talumbau/Qwen2.5-72B-Instruct-mxfp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use talumbau/Qwen2.5-72B-Instruct-mxfp8 with Transformers:

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

pipe = pipeline("text-generation", model="talumbau/Qwen2.5-72B-Instruct-mxfp8")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("talumbau/Qwen2.5-72B-Instruct-mxfp8")
model = AutoModelForCausalLM.from_pretrained("talumbau/Qwen2.5-72B-Instruct-mxfp8", 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 talumbau/Qwen2.5-72B-Instruct-mxfp8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "talumbau/Qwen2.5-72B-Instruct-mxfp8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "talumbau/Qwen2.5-72B-Instruct-mxfp8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/talumbau/Qwen2.5-72B-Instruct-mxfp8

SGLang

How to use talumbau/Qwen2.5-72B-Instruct-mxfp8 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 "talumbau/Qwen2.5-72B-Instruct-mxfp8" \
    --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": "talumbau/Qwen2.5-72B-Instruct-mxfp8",
		"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 "talumbau/Qwen2.5-72B-Instruct-mxfp8" \
        --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": "talumbau/Qwen2.5-72B-Instruct-mxfp8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use talumbau/Qwen2.5-72B-Instruct-mxfp8 with Docker Model Runner:
```
docker model run hf.co/talumbau/Qwen2.5-72B-Instruct-mxfp8
```

Qwen2.5-72B-Instruct — MXFP8 Quantization

This is an MXFP8-quantized derivative of Qwen/Qwen2.5-72B-Instruct, produced using llm-compressor v0.13.0.

Quantization details

Property	Value
Format	OCP MX FP8 (E4M3)
Block size	32 elements
Scale dtype	uint8 (E8M0)
Weight quantization	Static, per 32-element group
Activation quantization	Dynamic, per 32-element group
Layers quantized	All `nn.Linear` in the 80 decoder layers (560 total)
Layers skipped	`lm_head`, `embed_tokens`
Checkpoint format	`compressed-tensors` (auto-detected by vLLM)

The quantized layers cover all attention projections (q_proj, k_proj, v_proj, o_proj) and MLP projections (gate_proj, up_proj, down_proj) in every decoder layer.

Hardware requirements

MXFP8 hardware acceleration requires AMD MI350X / MI355X (CDNA4, gfx950) with ROCm ≥ 7.2.0, or NVIDIA Blackwell (SM100+).

Note: This checkpoint will NOT run correctly on AMD MI300X (gfx942) or NVIDIA Hopper (H100/H200). Those GPUs use different FP8 encodings (FNUZ and E4M3FNUZ respectively) that are incompatible with the OCP MX format stored here.

Usage with vLLM

from vllm import LLM, SamplingParams

llm = LLM(
    model="talumbau/Qwen2.5-72B-Instruct-mxfp8",
    # vLLM auto-detects compressed-tensors format — no --quantization flag needed
)

outputs = llm.generate(
    ["Tell me about AMD MI350X"],
    SamplingParams(max_tokens=200, temperature=0.7),
)
print(outputs[0].outputs[0].text)

Or from the command line:

vllm serve talumbau/Qwen2.5-72B-Instruct-mxfp8

Comparison variants

This checkpoint is part of a three-format study:

Variant	Repo	Format
bf16 (original)	`Qwen/Qwen2.5-72B-Instruct`	bfloat16
fp8 (per-channel/per-token)	`RedHatAI/Qwen2.5-72B-Instruct-FP8-dynamic`	compressed-tensors fp8
mxfp8 (this repo)	`talumbau/Qwen2.5-72B-Instruct-mxfp8`	OCP MX FP8, block_size=32

License

This derivative work is distributed under the Qwen License Agreement.

Modifications: The original bfloat16 weights were post-training quantized to OCP MX FP8 format using llm-compressor v0.13.0. No fine-tuning or other modifications were made.

Downloads last month: 32

Safetensors

Model size

73B params

Tensor type

BF16

F8_E4M3

Model tree for talumbau/Qwen2.5-72B-Instruct-mxfp8

Base model

Qwen/Qwen2.5-72B

Finetuned

Qwen/Qwen2.5-72B-Instruct

Quantized

(94)

this model