Instructions to use ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3 with Transformers:

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

pipe = pipeline("text-generation", model="ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3", trust_remote_code=True)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3

SGLang

How to use ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3 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 "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3 with Docker Model Runner:
```
docker model run hf.co/ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3
```

OLMo 3 with SiameseNorm and DepthAttention

This repository is one of five standalone OLMo 3 checkpoints trained with SiameseNorm and DepthAttention. All model, configuration, tokenizer, and custom modeling files are stored directly at the repository root.

Repository suffix	Training phase	Context length
`stage1`	Pretraining	8,192
`stage2`	Mid-training	8,192
`stage3`	Long-context training	65,536
`stage4-think`	Think SFT	65,536
`stage4-instruct`	Instruct SFT	65,536

Stage 3 and Stage 4 apply YaRN only to Full-attention layers. Sliding-window attention layers retain the original RoPE and a 4,096-token window.

Loading

Set repo_id to the desired repository from the table above:

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo_id = "ArchSpace-Collection/OLMo3-SiameseNorm-DepthAttention-<stage>"

tokenizer = AutoTokenizer.from_pretrained(
    repo_id,
    trust_remote_code=True,
    fix_mistral_regex=False,
)
model = AutoModelForCausalLM.from_pretrained(
    repo_id,
    trust_remote_code=True,
    dtype=torch.bfloat16,
    attn_implementation="sdpa",
)

fix_mistral_regex=False preserves the tokenizer behavior used during training. SDPA is the recommended and release-validated BF16 inference backend.

Architecture

16 transformer layers
hidden size 2,048
intermediate size 8,192
16 query heads and 16 key/value heads
128-dimensional attention heads
3:1 sliding-window/full-attention pattern
4,096-token sliding window
reordered RMSNorm, SiameseNorm, and DepthAttention

Downloads last month: -

Safetensors

Model size

1B params

Tensor type

BF16

Collection including ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-stage3

SiameseNorm-DepthAttention

Collection

5 items • Updated about 17 hours ago