Instructions to use suwesh/llamatron-1B-peft with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use suwesh/llamatron-1B-peft with Transformers:

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

pipe = pipeline("text-generation", model="suwesh/llamatron-1B-peft")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("suwesh/llamatron-1B-peft")
model = AutoModelForCausalLM.from_pretrained("suwesh/llamatron-1B-peft")
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]:]))

PEFT
How to use suwesh/llamatron-1B-peft with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use suwesh/llamatron-1B-peft with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "suwesh/llamatron-1B-peft"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "suwesh/llamatron-1B-peft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/suwesh/llamatron-1B-peft

SGLang

How to use suwesh/llamatron-1B-peft 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 "suwesh/llamatron-1B-peft" \
    --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": "suwesh/llamatron-1B-peft",
		"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 "suwesh/llamatron-1B-peft" \
        --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": "suwesh/llamatron-1B-peft",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use suwesh/llamatron-1B-peft with Docker Model Runner:
```
docker model run hf.co/suwesh/llamatron-1B-peft
```

suwesh commited on Aug 22, 2025

Commit

de3c8a3

verified ·

1 Parent(s): 781d08b

Update README.md

Browse files

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -87,7 +87,7 @@ model = PeftModel.from_pretrained(base_model, "suwesh/llamatron-1B-peft", subfol
 <pre>Checkpoint 11000 Training and Validation losses: 1.06 | 1.09</pre>
 # Evaluation details
-We use the [nvidia/Llama-3.1-Nemotron-Nano](https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-8B-v1) LLM as a Judge for evaluating the responses between the base llama 3.2 1b instruct and out PEFT model. The following are the judge's preference for each prompt to the two models:
 <pre>base: 122
 peft: 388
 tie: 29

 <pre>Checkpoint 11000 Training and Validation losses: 1.06 | 1.09</pre>
 # Evaluation details
+We use the [nvidia/Llama-3.1-Nemotron-Nano](https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-8B-v1) LLM as a Judge for evaluating the responses between the base llama 3.2 1b instruct and our PEFT model. The following are the judge's preference for each prompt to the two models, we also provide the ground truth in the prompt to the judge:
 <pre>base: 122
 peft: 388
 tie: 29