Instructions to use KristianS7/Ouro-1.4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KristianS7/Ouro-1.4B with Transformers:

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

pipe = pipeline("text-generation", model="KristianS7/Ouro-1.4B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("KristianS7/Ouro-1.4B", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KristianS7/Ouro-1.4B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KristianS7/Ouro-1.4B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KristianS7/Ouro-1.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KristianS7/Ouro-1.4B

SGLang

How to use KristianS7/Ouro-1.4B 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 "KristianS7/Ouro-1.4B" \
    --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": "KristianS7/Ouro-1.4B",
		"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 "KristianS7/Ouro-1.4B" \
        --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": "KristianS7/Ouro-1.4B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KristianS7/Ouro-1.4B with Docker Model Runner:
```
docker model run hf.co/KristianS7/Ouro-1.4B
```

KristianS7 commited on May 25

Commit

09d3a07

verified ·

1 Parent(s): 880ce86

Update tied weight metadata for Transformers 5

Browse files

Transformers 5 expects `_tied_weights_keys` to be a target-to-source mapping.

This mirrors the built-in Qwen3/Qwen3.5 remote-code pattern and keeps Ouro compatible if `tie_word_embeddings` is enabled. With the default `tie_word_embeddings=False`, the mapping remains inert.

Files changed (1) hide show

modeling_ouro.py +1 -1

modeling_ouro.py CHANGED Viewed

@@ -605,7 +605,7 @@ class _FP32LMHead(nn.Module):
 class OuroForCausalLM(OuroPreTrainedModel, GenerationMixin):
-    _tied_weights_keys = ["lm_head.weight"]
     _tp_plan = {"lm_head": "colwise_rep"}
     _pp_plan = {"lm_head": (["hidden_states"], ["logits"])}

 class OuroForCausalLM(OuroPreTrainedModel, GenerationMixin):
+    _tied_weights_keys = {"lm_head.weight": "model.embed_tokens.weight"}
     _tp_plan = {"lm_head": "colwise_rep"}
     _pp_plan = {"lm_head": (["hidden_states"], ["logits"])}