Qwen2.5-Coder-14B-Instruct-FP8-dynamic

Model Overview

• Model Architecture: Qwen2ForCausalLM
  • Input: Text
  • Output: Text
• Model Optimizations:
  • Weight quantization: FP8
  • Activation quantization: FP8
• Release Date: 11/28/2024
• Version: 1.0
• Model Developers: Red Hat

Quantized version of Qwen/Qwen2.5-Coder-14B-Instruct.

Model Optimizations

This model was obtained by quantizing the weights and activations of Qwen/Qwen2.5-Coder-14B-Instruct to FP8 data type. This optimization reduces the number of bits per parameter from 16 to 8, reducing the disk size and GPU memory requirements by approximately 50%. Only the weights and activations of the linear operators within transformers blocks are quantized.

Deployment

Use with vLLM

1. Initialize vLLM server:

vllm serve RedHatAI/Qwen2.5-Coder-14B-Instruct-FP8-dynamic

1. Send requests to the server:

from openai import OpenAI

# Modify OpenAI's API key and API base to use vLLM's API server.
openai_api_key = "EMPTY"
openai_api_base = "http://<your-server-host>:8000/v1"

client = OpenAI(
    api_key=openai_api_key,
    base_url=openai_api_base,
)

model = "RedHatAI/Qwen2.5-Coder-14B-Instruct-FP8-dynamic"

messages = [
    {"role": "user", "content": "Write a quick sort algorithm."},
]

outputs = client.chat.completions.create(
    model=model,
    messages=messages,
)

generated_text = outputs.choices[0].message.content
print(generated_text)

Creation

This model was created with llm-compressor by running the code snippet below.

<details> <summary>Model Creation Code</summary>

from llmcompressor.modifiers.quantization import QuantizationModifier
from llmcompressor.transformers import oneshot
from transformers import AutoModelForCausalLM, AutoTokenizer

# Load model
model_stub = "Qwen/Qwen2.5-Coder-14B-Instruct"
model_name = model_stub.split("/")[-1]

model = AutoModelForCausalLM.from_pretrained(model_stub, dtype="auto")

tokenizer = AutoTokenizer.from_pretrained(model_stub)

# Configure the quantization algorithm and scheme
recipe = QuantizationModifier(
    ignore=["lm_head"],
    targets="Linear",
    scheme="FP8_dynamic",
)

# Apply quantization
oneshot(
    model=model,
    recipe=recipe,
)

# Save to disk in compressed-tensors format
save_path = model_name + "-FP8-dynamic"
model.save_pretrained(save_path)
tokenizer.save_pretrained(save_path)
print(f"Model and tokenizer saved to: {save_path}")

</details>