Instructions to use radicalnumerics/RND1-Base-0910 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use radicalnumerics/RND1-Base-0910 with Transformers:

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

pipe = pipeline("text-generation", model="radicalnumerics/RND1-Base-0910", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import RND1
model = RND1.from_pretrained("radicalnumerics/RND1-Base-0910", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use radicalnumerics/RND1-Base-0910 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "radicalnumerics/RND1-Base-0910"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "radicalnumerics/RND1-Base-0910",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/radicalnumerics/RND1-Base-0910

SGLang

How to use radicalnumerics/RND1-Base-0910 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 "radicalnumerics/RND1-Base-0910" \
    --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": "radicalnumerics/RND1-Base-0910",
		"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 "radicalnumerics/RND1-Base-0910" \
        --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": "radicalnumerics/RND1-Base-0910",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use radicalnumerics/RND1-Base-0910 with Docker Model Runner:
```
docker model run hf.co/radicalnumerics/RND1-Base-0910
```

RND1-Base-0910 / config.json

fedebotu

Update config.json for llama.cpp integration (#2)

f1b49af verified 6 months ago

raw

history blame contribute delete

1.2 kB

	{
	"architectures": ["RND1"],
	"attention_bias": false,
	"attention_dropout": 0.0,
	"auto_map": {
	"AutoConfig": "configuration_rnd.RND1Config",
	"AutoModel": "modeling_rnd.RND1Model",
	"AutoModelForMaskedLM": "modeling_rnd.RND1LM"
	},
	"decoder_sparse_step": 1,
	"dtype": "bfloat16",
	"eos_token_id": 151645,
	"head_dim": 128,
	"hidden_act": "silu",
	"hidden_size": 2048,
	"initializer_range": 0.02,
	"intermediate_size": 6144,
	"is_causal": false,
	"mask_token_id": 151669,
	"max_position_embeddings": 40960,
	"max_window_layers": 48,
	"mlp_only_layers": [],
	"model_type": "rnd1",
	"moe_backend": "hf",
	"moe_intermediate_size": 768,
	"norm_topk_prob": true,
	"num_attention_heads": 32,
	"num_diffusion_steps": 256,
	"num_experts": 128,
	"num_experts_per_tok": 8,
	"num_hidden_layers": 48,
	"num_key_value_heads": 4,
	"output_router_logits": false,
	"pad_token_id": 151643,
	"rms_norm_eps": 1e-06,
	"rope_scaling": false,
	"rope_theta": 1000000.0,
	"router_aux_loss_coef": 0.001,
	"sliding_window": false,
	"tie_word_embeddings": false,
	"transformers_version": "4.57.0",
	"use_cache": false,
	"use_sliding_window": false,
	"vocab_size": 151936
	}