Instructions to use Kirim-ai/Kirim-Safeguard-R1-10B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Kirim-ai/Kirim-Safeguard-R1-10B with Transformers:

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

pipe = pipeline("text-generation", model="Kirim-ai/Kirim-Safeguard-R1-10B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Kirim-ai/Kirim-Safeguard-R1-10B", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Kirim-ai/Kirim-Safeguard-R1-10B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Kirim-ai/Kirim-Safeguard-R1-10B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Kirim-ai/Kirim-Safeguard-R1-10B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Kirim-ai/Kirim-Safeguard-R1-10B

SGLang

How to use Kirim-ai/Kirim-Safeguard-R1-10B 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 "Kirim-ai/Kirim-Safeguard-R1-10B" \
    --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": "Kirim-ai/Kirim-Safeguard-R1-10B",
		"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 "Kirim-ai/Kirim-Safeguard-R1-10B" \
        --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": "Kirim-ai/Kirim-Safeguard-R1-10B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Kirim-ai/Kirim-Safeguard-R1-10B with Docker Model Runner:
```
docker model run hf.co/Kirim-ai/Kirim-Safeguard-R1-10B
```

Kirim-Safeguard-R1-10B / generation_config.json

Song Yi

Create generation_config.json

d85f3d6 verified 9 months ago

Raw

History Blame Contribute Delete

1.23 kB

	{
	"_from_model_config": true,
	"bos_token_id": 128000,
	"eos_token_id": [
	128001,
	128004
	],
	"pad_token_id": null,
	"transformers_version": "4.45.0",

	"do_sample": true,
	"temperature": 0.7,
	"top_p": 0.9,
	"top_k": 50,
	"repetition_penalty": 1.1,
	"no_repeat_ngram_size": 3,
	"encoder_no_repeat_ngram_size": 0,

	"max_length": 8192,
	"max_new_tokens": null,
	"min_length": 0,
	"min_new_tokens": null,

	"early_stopping": false,
	"length_penalty": 1.0,
	"num_beams": 1,
	"num_beam_groups": 1,
	"diversity_penalty": 0.0,

	"num_return_sequences": 1,
	"output_attentions": false,
	"output_hidden_states": false,
	"output_scores": false,
	"return_dict_in_generate": false,

	"bad_words_ids": null,
	"force_words_ids": null,
	"renormalize_logits": false,
	"constraints": null,
	"forced_bos_token_id": null,
	"forced_eos_token_id": null,
	"remove_invalid_values": true,
	"exponential_decay_length_penalty": null,
	"suppress_tokens": null,
	"begin_suppress_tokens": null,

	"guidance_scale": null,
	"low_memory": null,

	"typical_p": null,
	"epsilon_cutoff": 0.0,
	"eta_cutoff": 0.0,

	"use_cache": true,
	"static_cache": false,
	"cache_implementation": null
	}