Instructions to use nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1 with Transformers:

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

pipe = pipeline("text-generation", model="nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1")
model = AutoModelForCausalLM.from_pretrained("nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1")
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 nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1

SGLang

How to use nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1 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 "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1" \
    --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": "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1",
		"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 "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1" \
        --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": "nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1 with Docker Model Runner:
```
docker model run hf.co/nitky/Llama-3.3-SuperSwallowX-70B-Instruct-v0.1
```

Llama-3.3-SuperSwallowX-70B-Instruct-v0.1

File size: 4,998 Bytes

1a80318

merge_method: task_arithmetic
base_model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
models:
  - model: tokyotech-llm/Llama-3.1-Swallow-70B-Instruct-v0.1
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-swallow
---
merge_method: task_arithmetic
base_model: meta-llama/Llama-3.1-70B
models:
  - model: tokyotech-llm/Llama-3.1-Swallow-70B-Instruct-v0.1
    parameters:
      weight: 1.0
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-llamaswallow
---
merge_method: slerp
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-llamaswallow
models:
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-llamaswallow
  - model: cyberagent/Llama-3.1-70B-Japanese-Instruct-2407
parameters:
  t: 0.5
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-cyberswallow-preset
---
merge_method: task_arithmetic
base_model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-cyberswallow-preset
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-cyberswallow
---
merge_method: slerp
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-llamaswallow
models:
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-llamaswallow
  - model: AXCXEPT/Llama-3.1-70B-EZO-1.1-it
parameters:
  t: 0.5
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-ezoswallow-preset
---
merge_method: task_arithmetic
base_model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-ezoswallow-preset
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-ezoswallow
---
merge_method: model_stock
base_model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
models:
  - model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-swallow
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-cyberswallow
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-ezoswallow
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
---
merge_method: task_arithmetic
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: HPAI-BSC/Llama3.1-Aloe-Beta-70B
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-aloe
---
merge_method: task_arithmetic
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: marcelbinz/Llama-3.1-Centaur-70B
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-centaur
---
merge_method: task_arithmetic
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: hitachi-nlp/Llama-3.1-70B-FLDx2
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-fldx2
---
merge_method: task_arithmetic
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: Skywork/Skywork-Critic-Llama-3.1-70B
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-skywork
---
merge_method: task_arithmetic
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: meta-llama/Llama-3.1-70B-Instruct
    parameters:
      weight: -1.0
  - model: DISLab/SummLlama3.1-70B
    parameters:
      weight: 1.0
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-summ
---
merge_method: model_stock
base_model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-base
models:
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-aloe
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-centaur
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-fldx2
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-summ
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-preset-skywork
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-flavor
---
merge_method: slerp
base_model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
models:
  - model: nitky/Llama-3.3-SuperSwallow-70B-Instruct-v0.1
  - model: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1-flavor
parameters:
  t: 0.5
dtype: bfloat16
name: Llama-3.3-SuperSwallowX-70B-Instruct-v0.1