Instructions to use NhatCuong22/gemma-4-e4b-scientific-reviewer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NhatCuong22/gemma-4-e4b-scientific-reviewer with Transformers:

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

pipe = pipeline("text-generation", model="NhatCuong22/gemma-4-e4b-scientific-reviewer")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("NhatCuong22/gemma-4-e4b-scientific-reviewer")
model = AutoModelForImageTextToText.from_pretrained("NhatCuong22/gemma-4-e4b-scientific-reviewer")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use NhatCuong22/gemma-4-e4b-scientific-reviewer with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "NhatCuong22/gemma-4-e4b-scientific-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": "NhatCuong22/gemma-4-e4b-scientific-reviewer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/NhatCuong22/gemma-4-e4b-scientific-reviewer

SGLang

How to use NhatCuong22/gemma-4-e4b-scientific-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 "NhatCuong22/gemma-4-e4b-scientific-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": "NhatCuong22/gemma-4-e4b-scientific-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 "NhatCuong22/gemma-4-e4b-scientific-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": "NhatCuong22/gemma-4-e4b-scientific-reviewer",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use NhatCuong22/gemma-4-e4b-scientific-reviewer with Docker Model Runner:
```
docker model run hf.co/NhatCuong22/gemma-4-e4b-scientific-reviewer
```

NhatCuong22 commited on Apr 18

Commit

16257fc

verified ·

1 Parent(s): 97af645

Upload benchmark/peerread_ft_metrics.json

Browse files

Files changed (1) hide show

benchmark/peerread_ft_metrics.json +85 -0

benchmark/peerread_ft_metrics.json ADDED Viewed

	@@ -0,0 +1,85 @@

+{
+  "label": "v3",
+  "input": "/root/benchmark/results/peerread_v3_ft_full.jsonl",
+  "n_papers": 95,
+  "venues": {
+    "conll_2016": 3,
+    "acl_2017": 14,
+    "iclr_2017": 78
+  },
+  "gen_time_s": {
+    "mean": 98.36,
+    "median": 96.75,
+    "total_min": 155.7
+  },
+  "review_words": {
+    "mean": 548.7,
+    "median": 546
+  },
+  "T2_decision_overall": {
+    "n": 95,
+    "accept_rate_true": 0.842,
+    "accept_rate_pred": 0.905,
+    "accuracy": 0.8105,
+    "precision": 0.8605,
+    "recall": 0.925,
+    "f1": 0.8916
+  },
+  "T2_decision_per_venue": {
+    "acl_2017": {
+      "n": 14,
+      "accept_rate_true": 0.643,
+      "accept_rate_pred": 0.857,
+      "accuracy": 0.6429,
+      "precision": 0.6667,
+      "recall": 0.8889,
+      "f1": 0.7619
+    },
+    "conll_2016": {
+      "n": 3,
+      "accept_rate_true": 0.333,
+      "accept_rate_pred": 1.0,
+      "accuracy": 0.3333,
+      "precision": 0.3333,
+      "recall": 1.0,
+      "f1": 0.5
+    },
+    "iclr_2017": {
+      "n": 78,
+      "accept_rate_true": 0.897,
+      "accept_rate_pred": 0.91,
+      "accuracy": 0.859,
+      "precision": 0.9155,
+      "recall": 0.9286,
+      "f1": 0.922
+    }
+  },
+  "T1_rouge_overall": {
+    "rouge1": {
+      "n": 95,
+      "mean_pct": 27.27
+    },
+    "rouge2": {
+      "n": 95,
+      "mean_pct": 6.89
+    },
+    "rougeL": {
+      "n": 95,
+      "mean_pct": 12.32
+    }
+  },
+  "T1_rouge_per_venue": {
+    "acl_2017": {
+      "n": 14,
+      "rougeL_mean_pct": 15.22
+    },
+    "conll_2016": {
+      "n": 3,
+      "rougeL_mean_pct": 16.26
+    },
+    "iclr_2017": {
+      "n": 78,
+      "rougeL_mean_pct": 11.65
+    }
+  }
+}