Instructions to use tzchen07/Nemotron-Super-120B-sft-example with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tzchen07/Nemotron-Super-120B-sft-example with Transformers:

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

pipe = pipeline("text-generation", model="tzchen07/Nemotron-Super-120B-sft-example", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tzchen07/Nemotron-Super-120B-sft-example", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("tzchen07/Nemotron-Super-120B-sft-example", trust_remote_code=True)
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 Settings

vLLM

How to use tzchen07/Nemotron-Super-120B-sft-example with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tzchen07/Nemotron-Super-120B-sft-example"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tzchen07/Nemotron-Super-120B-sft-example",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tzchen07/Nemotron-Super-120B-sft-example

SGLang

How to use tzchen07/Nemotron-Super-120B-sft-example 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 "tzchen07/Nemotron-Super-120B-sft-example" \
    --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": "tzchen07/Nemotron-Super-120B-sft-example",
		"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 "tzchen07/Nemotron-Super-120B-sft-example" \
        --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": "tzchen07/Nemotron-Super-120B-sft-example",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tzchen07/Nemotron-Super-120B-sft-example with Docker Model Runner:
```
docker model run hf.co/tzchen07/Nemotron-Super-120B-sft-example
```

Nemotron-3-Super-120B - Example SFT

This is an example supervised fine-tune (SFT) of NVIDIA's Nemotron-3-Super-120B (NemotronHForCausalLM, a Mamba-2 / MoE hybrid with 512 routed experts, top-22). It was produced as an end-to-end pipeline validation run, not a production model.

Training summary

Base: Nemotron-3-Super-120B (~120.7B params, bf16)
Method: Full-parameter SFT (no LoRA), DeepSpeed ZeRO-3, 24x H200 over RDMA
Data: Alpaca (2,000 examples), chat format
Steps: 84 (one full epoch; 2000 / 24 ranks)
Loss: 2.88 -> best 0.3534
Key fix: set_z3_leaf_modules(model, [NemotronHMoE]) to avoid a ZeRO-3 parameter-gather deadlock caused by data-dependent MoE expert routing.

Notes / caveats

This checkpoint omits the auxiliary mtp (multi-token-prediction) head (dropped by save_16bit_model); it is not required for standard generation.
Behaviour: shifts the base reasoning model toward concise, Alpaca-style direct answers.
Use trust_remote_code=True when loading (custom Nemotron-H modeling code is included).

Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
m = AutoModelForCausalLM.from_pretrained('tzchen07/Nemotron-Super-120B-sft-example',
                                         trust_remote_code=True, torch_dtype='bfloat16', device_map='auto')
t = AutoTokenizer.from_pretrained('tzchen07/Nemotron-Super-120B-sft-example', trust_remote_code=True)

Downloads last month: 21

Safetensors

Model size

121B params

Tensor type

BF16

F32