Instructions to use amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0 with Transformers:

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

pipe = pipeline("text-generation", model="amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0")
model = AutoModelForCausalLM.from_pretrained("amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0")
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 amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0

SGLang

How to use amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0 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 "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0" \
    --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": "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0",
		"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 "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0" \
        --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": "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0 with Docker Model Runner:
```
docker model run hf.co/amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0
```

Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0

Model Overview

Model Architecture: LlamaForCausalLM
- Input: Text
- Output: Text
Source Model: Llama-3.3-70B-Instruct
Supported Hardware: AMD EPYC (CPU inference)
Preferred Operating System: Linux
Inference Engine: vLLM v0.23.0
Quantization Framework: TorchAO v0.17.0
Quantization Method: 8-bit Dynamic Activation, 8-bit Weight Quantization, Symmetric
Compatible Stack:
- ZenDNN v6.0.0
- zentorch v2.11.0.2
- PyTorch v2.11.0
- TorchAO v0.17.0
- vLLM v0.23.0

zentorch v2.11.0.2 for PyTorch v2.11.0 has to be built from source.

This model was Built with Llama.

Quantization

The model was produced using torchao as shown in the example below. Both activations and weights are quantized to INT8 with symmetric mapping. Activation scales are computed dynamically at runtime per token. For Llama-3.3-70B-Instruct, the self-attention modules of layers 0, 1, and 3 are excluded from quantization in addition to lm_head.

import torch
from transformers import TorchAoConfig, AutoModelForCausalLM, AutoTokenizer
from torchao.quantization import Int8DynamicActivationInt8WeightConfig
from torchao.quantization.quant_primitives import MappingType

MODEL_ID   = "meta-llama/Llama-3.3-70B-Instruct"
OUTPUT_DIR = "amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0"
modules_to_skip = [
    "lm_head",
    "model.layers.0.self_attn",
    "model.layers.1.self_attn",
    "model.layers.3.self_attn",
]

quantization_config = TorchAoConfig(
    Int8DynamicActivationInt8WeightConfig(
        version=2,
        act_mapping_type=MappingType.SYMMETRIC,
    ),
    modules_to_not_convert=modules_to_skip,
)

model = AutoModelForCausalLM.from_pretrained(
    MODEL_ID,
    dtype=torch.bfloat16,
    device_map="cpu",
    quantization_config=quantization_config,
    trust_remote_code=True,
)
model.save_pretrained(OUTPUT_DIR)

tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
tokenizer.save_pretrained(OUTPUT_DIR)

# Smoke test
inputs = tokenizer("What are we having for dinner?", return_tensors="pt")
out = model.generate(**inputs, max_new_tokens=30, cache_implementation="static")
print(tokenizer.decode(out[0], skip_special_tokens=True))

Quick Start

Requirements

pip install --extra-index-url https://download.pytorch.org/whl/cpu \
            --extra-index-url https://wheels.vllm.ai/cpu/ \
    torch==2.11.0+cpu \
    vllm==0.23.0 \
    torchao==0.17.0 \
    "lm-eval[vllm]==0.4.12" \
    huggingface_hub

CPU runtime libraries (only needed if not already present):

conda install -c conda-forge gperftools=2.17.2 llvm-openmp=18.1.8 --no-deps -y

Recommended environment variables

# TorchInductor + zentorch
export TORCHINDUCTOR_FREEZING=1
export TORCHINDUCTOR_AUTOGRAD_CACHE=0
export VLLM_USE_AOT_COMPILE=0
export ZENDNNL_MATMUL_ALGO=1

# Required CPU runtime libraries
export LD_PRELOAD="<path to lib>/libtcmalloc_minimal.so.4:<path to lib>/libiomp5.so${LD_PRELOAD:+:$LD_PRELOAD}"

Locate the libraries with find / -name 'libtcmalloc_minimal.so.4' and find / -name 'libiomp5.so', then substitute the resulting directory for <path to lib>.

Evaluation

The model was evaluated against the BF16 (unquantized) baseline using lm-evaluation-harness with the vLLM engine.

Benchmark	BF16 Baseline	DA8W8 (this model)	Dynamic Quant Difference (baseline: BF16)
GSM8K (5-shot, exact-match flexible)	0.9477	0.9409	-0.72%

Evaluation Command

lm_eval \
    --model vllm \
    --model_args pretrained=amd/Llama-3.3-70B-Instruct-da8w8-torchao-v0.17.0,tokenizer=meta-llama/Llama-3.3-70B-Instruct,dtype=bfloat16 \
    --tasks gsm8k \
    --batch_size auto \
    --trust_remote_code \
    --num_fewshot 5 \
    --log_samples \
    --gen_kwargs "max_gen_toks=2048" \
    --apply_chat_template \
    --output_path .

Limitations

Version Lock: This model is quantized with TorchAO v0.17.0 and is compatible only with PyTorch v2.11.0 / ZenDNN v6.0.0. It will not load correctly on other PyTorch versions.
CPU Only: This model is optimized for AMD EPYC CPU inference via ZenDNN. It is not intended for GPU inference.