Instructions to use unlimitedbytes/gptoss-bigcodebench-20b-lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use unlimitedbytes/gptoss-bigcodebench-20b-lora with Transformers:

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

pipe = pipeline("text-generation", model="unlimitedbytes/gptoss-bigcodebench-20b-lora")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("unlimitedbytes/gptoss-bigcodebench-20b-lora", dtype="auto")

PEFT
How to use unlimitedbytes/gptoss-bigcodebench-20b-lora with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use unlimitedbytes/gptoss-bigcodebench-20b-lora with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "unlimitedbytes/gptoss-bigcodebench-20b-lora"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "unlimitedbytes/gptoss-bigcodebench-20b-lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/unlimitedbytes/gptoss-bigcodebench-20b-lora

SGLang

How to use unlimitedbytes/gptoss-bigcodebench-20b-lora 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 "unlimitedbytes/gptoss-bigcodebench-20b-lora" \
    --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": "unlimitedbytes/gptoss-bigcodebench-20b-lora",
		"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 "unlimitedbytes/gptoss-bigcodebench-20b-lora" \
        --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": "unlimitedbytes/gptoss-bigcodebench-20b-lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use unlimitedbytes/gptoss-bigcodebench-20b-lora with Docker Model Runner:
```
docker model run hf.co/unlimitedbytes/gptoss-bigcodebench-20b-lora
```

gptoss-bigcodebench-20b-lora

File size: 1,578 Bytes

---
license: apache-2.0
base_model: openai/gpt-oss-20b
library_name: transformers
pipeline_tag: text-generation
tags:
- peft
- lora
- bigcodebench
- gpt-oss
- code
- causal-lm
inference: false
---
# GPT-OSS-20B BigCodeBench LoRA Adapter

LoRA adapter weights fine-tuned from `openai/gpt-oss-20b` on BigCodeBench split `v0.1.4` (~1.1K samples).

## Training Summary

- Steps: 100
- Final train_loss: 0.7833267974853516
- Runtime (s): 3717.3139
- Samples/sec: 0.43
- Total FLOPs: 6.825417425085542e+16

## Usage

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base = 'openai/gpt-oss-20b'
adapter = 'unlimitedbytes/gptoss-bigcodebench-20b-lora'
model = AutoModelForCausalLM.from_pretrained(base, device_map='auto', torch_dtype='auto')
model = PeftModel.from_pretrained(model, adapter)
tokenizer = AutoTokenizer.from_pretrained(base)
messages = [
    {'role': 'system', 'content': 'You are a helpful coding assistant.'},
    {'role': 'user', 'content': 'Write a Python function to add two numbers.'}
]
input_ids = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors='pt').to(model.device)
out = model.generate(input_ids, max_new_tokens=128)
print(tokenizer.decode(out[0], skip_special_tokens=False))
```

Merge adapter:

```python
model = model.merge_and_unload()
model.save_pretrained('merged-model')
```

## Limitations

- 100 training steps only; not fully converged.
- Adapter only, no merged full weights.
- Outputs may include control tokens.

## License

Apache-2.0 (base) + dataset licenses.