Instructions to use naimul011/GlueQwen with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use naimul011/GlueQwen with Transformers:

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

pipe = pipeline("text-generation", model="naimul011/GlueQwen")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("naimul011/GlueQwen")
model = AutoModelForCausalLM.from_pretrained("naimul011/GlueQwen")
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

vLLM

How to use naimul011/GlueQwen with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "naimul011/GlueQwen"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "naimul011/GlueQwen",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/naimul011/GlueQwen

SGLang

How to use naimul011/GlueQwen 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 "naimul011/GlueQwen" \
    --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": "naimul011/GlueQwen",
		"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 "naimul011/GlueQwen" \
        --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": "naimul011/GlueQwen",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use naimul011/GlueQwen with Docker Model Runner:
```
docker model run hf.co/naimul011/GlueQwen
```

Model Card for Model ID

The GlueQwen model is fine-tuned on four distinct tasks from the GLUE benchmark: SST-2 (Sentiment Analysis), MRPC (Paraphrase Detection), CoLA (Linguistic Acceptability), and MNLI (Natural Language Inference). The base model used is Qwen/Qwen2.5-7B, which has 7 billion parameters. Qwen2.5-7B is designed to enhance various natural language understanding tasks through pre-training on diverse datasets, followed by fine-tuning for task-specific improvements.

The fine-tuning of GlueQwen involves optimizing the model for these GLUE tasks, aiming to measure catastrophic forgetting, model learning abilities, and overall training performance across different language tasks. These benchmarks provide insight into how well the model retains previous knowledge while learning new tasks sequentially.

Model Details

Model Description

Benchmark Table for GlueQwen Fine-Tuning Performance

Model	Parameter Size (B)	Pretrained Performance	Forgetting	Learning	Training Performance
Llama-3.2-1B	1	0.50	0.24	0.33	0.54
Llama-3.2-3B	3	0.56	0.225	0.36	0.61
Llama-3.1-8B	8	0.56	0.59	0.84	0.67
Llama-3-8B	8	0.53	0.39	0.98	0.70
Llama-2-7B	7	0.67	0.23	0.12	0.63
GPT-J-6B	6	0.50	0.39	0.45	0.54
Phi-2	2.7	0.59	0.10	0.15	0.61
Phi-3.5-mini	3.82	0.69	0.02	0.30	0.76
Orca-2-7b	7	0.76	0.185	0.33	0.81
Qwen2.5-0.5B	0.5	0.52	0.23	0.56	0.61
Qwen2.5-7B	7	0.56	0.51	1.12	0.77
Qwen2.5-14B	14	0.71	0.935	0.66	0.80
GlueQwen	7	0.59	0.42	0.97	0.73

Analysis

GlueQwen, fine-tuned on multiple tasks from the GLUE dataset, demonstrates a pre-trained performance of 0.59. Its forgetting rate is moderate at 0.42, reflecting some loss of previously learned information. However, the model exhibits a strong learning capability with a learning score of 0.97. The overall training performance stands at 0.73, positioning GlueQwen as a balanced model that manages forgetting while achieving significant improvements in task-specific learning.

Developed by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Model type: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Finetuned from model [optional]: [More Information Needed]

Model Sources [optional]

Repository: [More Information Needed]
Paper [optional]: [More Information Needed]
Demo [optional]: [More Information Needed]

Uses

Direct Use

[More Information Needed]

Downstream Use [optional]

[More Information Needed]

Out-of-Scope Use

[More Information Needed]

Bias, Risks, and Limitations

[More Information Needed]

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.

How to Get Started with the Model

Use the code below to get started with the model.

[More Information Needed]

Training Details

Training Data

[More Information Needed]

Training Procedure

Preprocessing [optional]

[More Information Needed]

Training Hyperparameters

Training regime: [More Information Needed]

Speeds, Sizes, Times [optional]

[More Information Needed]

Evaluation

Testing Data, Factors & Metrics

Testing Data

[More Information Needed]

Factors

[More Information Needed]

Metrics

[More Information Needed]

Results

[More Information Needed]

Summary

Model Examination [optional]

[More Information Needed]

Environmental Impact

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

Hardware Type: [More Information Needed]
Hours used: [More Information Needed]
Cloud Provider: [More Information Needed]
Compute Region: [More Information Needed]
Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

[More Information Needed]

Model Card Contact

[More Information Needed]

Downloads last month: 2

Safetensors

Model size

8B params

Tensor type

F32

Model tree for naimul011/GlueQwen

Base model

Qwen/Qwen2.5-7B

Quantized

(82)

this model

Paper for naimul011/GlueQwen

Quantifying the Carbon Emissions of Machine Learning

Paper • 1910.09700 • Published Oct 21, 2019 • 48