Instructions to use DeerAi001/distilgpt2-imbd-finetunned with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DeerAi001/distilgpt2-imbd-finetunned with Transformers:

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

pipe = pipeline("text-generation", model="DeerAi001/distilgpt2-imbd-finetunned")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("DeerAi001/distilgpt2-imbd-finetunned", device_map="auto")

PEFT
How to use DeerAi001/distilgpt2-imbd-finetunned with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use DeerAi001/distilgpt2-imbd-finetunned with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DeerAi001/distilgpt2-imbd-finetunned"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DeerAi001/distilgpt2-imbd-finetunned",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/DeerAi001/distilgpt2-imbd-finetunned

SGLang

How to use DeerAi001/distilgpt2-imbd-finetunned 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 "DeerAi001/distilgpt2-imbd-finetunned" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DeerAi001/distilgpt2-imbd-finetunned",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "DeerAi001/distilgpt2-imbd-finetunned" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DeerAi001/distilgpt2-imbd-finetunned",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use DeerAi001/distilgpt2-imbd-finetunned with Docker Model Runner:
```
docker model run hf.co/DeerAi001/distilgpt2-imbd-finetunned
```

Model Card for DistilGPT2-IMDb-LoRA-Finetuned

A LoRA fine-tuned version of DistilGPT2 on IMDb movie review dataset for text generation.

Model Details

Model Description

This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model has been fine-tuned using LoRA (Low-Rank Adaptation) technique on IMDb dataset for movie review style text generation.

Developed by: Anurag Pandey
Funded by [optional]: Self-funded
Shared by [optional]: DeerAI001
Model type: Causal Language Model (Text Generation)
Language(s) (NLP): English
License: Apache-2.0
Finetuned from model [optional]: distilgpt2

Model Sources [optional]

Repository: https://huggingface.co/DeerAI001/distilgpt2-imbd-finetuned
Paper [optional]: N/A
Demo [optional]: N/A

Uses

Direct Use

This model can be used directly for text generation with movie review style completions.

Downstream Use [optional]

Can be further fine-tuned for specific sentiment analysis or review generation tasks.

Out-of-Scope Use

Not suitable for factual question answering or non-English text generation.

Bias, Risks, and Limitations

Model is trained on IMDb dataset which may contain biases present in movie reviews. Limited to English language only.

Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use for creative text generation purposes only.

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
from peft import PeftModel, PeftConfig

# Load adapter config
adapter_path = "DeerAI001/distilgpt2-imbd-finetuned"
peft_config = PeftConfig.from_pretrained(adapter_path)

# Load base model
base_model = AutoModelForCausalLM.from_pretrained(
    peft_config.base_model_name_or_path,
    trust_remote_code=True
)

# Load adapter on top of base model
model = PeftModel.from_pretrained(base_model, adapter_path)
model.eval()

# Tokenizer
tokenizer = AutoTokenizer.from_pretrained(peft_config.base_model_name_or_path)
tokenizer.pad_token = tokenizer.eos_token

# Inference pipeline
pipe = pipeline(
    "text-generation",
    model=model,
    tokenizer=tokenizer,
    max_new_tokens=100,
    do_sample=True,
    temperature=0.8
)

# Test input
prompt = "The movie was absolutely wonderful because"
result = pipe(prompt)
print(result[0]["generated_text"])

Training Details

Training Data

IMDb Movie Reviews Dataset - 500 samples subset used for training.

Training Procedure

Preprocessing [optional]

Tokenized with max_length=64, padding="max_length", truncation=True

Training Hyperparameters

Training regime: fp32 (CPU training)
Batch size: 1
Gradient accumulation steps: 8
Effective batch size: 8
Epochs: 1
LoRA rank (r): 4
LoRA alpha: 8
LoRA dropout: 0.1
Target modules: c_attn, c_proj

Speeds, Sizes, Times [optional]

Training time: ~30-60 minutes on CPU
Checkpoint size: ~1 MB (LoRA adapters only)

Evaluation

Testing Data, Factors & Metrics

Testing Data

IMDb test split

Factors

Text coherence and movie review style generation

Metrics

Qualitative evaluation of generated text

Results

Model generates coherent movie review style text completions.

Summary

Successfully fine-tuned DistilGPT2 using LoRA for movie review text generation.

Model Examination [optional]

LoRA adapters applied to attention layers (c_attn, c_proj) with rank 4.

Environmental Impact

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

Hardware Type: AMD Ryzen 5 5500U (CPU)
Hours used: ~1 hour
Cloud Provider: Local machine
Compute Region: India
Carbon Emitted: Minimal (CPU-only training)

Technical Specifications [optional]

Model Architecture and Objective

DistilGPT2 (82M parameters) with LoRA adapters (~50K trainable parameters)

Compute Infrastructure

Local CPU-based training

Hardware

CPU: AMD Ryzen 5 5500U
RAM: 8 GB
GPU: None (CPU training)

Software

transformers
peft
datasets
torch

Citation [optional]

BibTeX:

@misc{distilgpt2-imdb-lora-2025,
  author = {DeerAI001},
  title = {DistilGPT2 Fine-tuned with LoRA on IMDb},
  year = {2025},
  publisher = {Hugging Face},
  url = {https://huggingface.co/DeerAI001/distilgpt2-imbd-finetuned}
}

APA:

DeerAI001. (2025). DistilGPT2 Fine-tuned with LoRA on IMDb. Hugging Face. https://huggingface.co/DeerAI001/distilgpt2-imbd-finetuned

Glossary [optional]

LoRA: Low-Rank Adaptation - efficient fine-tuning technique
PEFT: Parameter-Efficient Fine-Tuning

More Information [optional]

For more details on LoRA fine-tuning, visit: https://huggingface.co/docs/peft

Model Card Authors [optional]

DeerAI001

Model Card Contact

Open an issue on the Hugging Face repository for questions or feedback.

Downloads last month: -; Downloads are not tracked for this model. How to track

Paper for DeerAi001/distilgpt2-imbd-finetunned

Quantifying the Carbon Emissions of Machine Learning

Paper • 1910.09700 • Published Oct 21, 2019 • 61