Instructions to use alfiannajih/g-retriever-resume-reviewer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use alfiannajih/g-retriever-resume-reviewer with Transformers:

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

pipe = pipeline("text-generation", model="alfiannajih/g-retriever-resume-reviewer", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("alfiannajih/g-retriever-resume-reviewer", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("alfiannajih/g-retriever-resume-reviewer", trust_remote_code=True)
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 alfiannajih/g-retriever-resume-reviewer with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/alfiannajih/g-retriever-resume-reviewer

SGLang

How to use alfiannajih/g-retriever-resume-reviewer 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 "alfiannajih/g-retriever-resume-reviewer" \
    --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": "alfiannajih/g-retriever-resume-reviewer",
		"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 "alfiannajih/g-retriever-resume-reviewer" \
        --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": "alfiannajih/g-retriever-resume-reviewer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use alfiannajih/g-retriever-resume-reviewer with Docker Model Runner:
```
docker model run hf.co/alfiannajih/g-retriever-resume-reviewer
```

Model Card for Model ID

This repository is created for submission to Compfest: Artificial Intelligence Challenge (AIC) 16.

G-Retriever integrates Graph Neural Networks (GNN), Large Language Model (LLM), and Retrieval-Augmented Generation(RAG) by using Knowledge Graph. This model was originaly developed by Xiaoxin He.

Model Details

Model Description

While the original method utilized Llama 2 family model as the LLM, this repository has experimented it with Llama 3.1 8B.

Model Sources

Repository: Repository
Training Script: G-Retriever Repository
Paper: G-Retriever Paper

Uses

This model is designed to be used as a resume reviewer. The approach involves retrieving a subgraph from a knowledge graph built from LinkedIn job postings and feeding it into a GNN. The features extracted from the subgraph are further processed and concatenated with the input embeddings from the query text. These concatenated features are then passed through the self-attention layer of Llama 3.1 8B to generate a resume review.