Instructions to use EchoLabs33/olmoe-1b-7b-instruct-hxq with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use EchoLabs33/olmoe-1b-7b-instruct-hxq with Transformers:

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

pipe = pipeline("text-generation", model="EchoLabs33/olmoe-1b-7b-instruct-hxq")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("EchoLabs33/olmoe-1b-7b-instruct-hxq")
model = AutoModelForCausalLM.from_pretrained("EchoLabs33/olmoe-1b-7b-instruct-hxq", device_map="auto")
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 EchoLabs33/olmoe-1b-7b-instruct-hxq with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "EchoLabs33/olmoe-1b-7b-instruct-hxq"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "EchoLabs33/olmoe-1b-7b-instruct-hxq",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/EchoLabs33/olmoe-1b-7b-instruct-hxq

SGLang

How to use EchoLabs33/olmoe-1b-7b-instruct-hxq 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 "EchoLabs33/olmoe-1b-7b-instruct-hxq" \
    --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": "EchoLabs33/olmoe-1b-7b-instruct-hxq",
		"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 "EchoLabs33/olmoe-1b-7b-instruct-hxq" \
        --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": "EchoLabs33/olmoe-1b-7b-instruct-hxq",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use EchoLabs33/olmoe-1b-7b-instruct-hxq with Docker Model Runner:
```
docker model run hf.co/EchoLabs33/olmoe-1b-7b-instruct-hxq
```

voidstream commited on Apr 18

Commit

9c07e9d

verified ·

1 Parent(s): d1c8304

Update model card: LoRA fine-tuning now supported via HelixLinearSTE

Browse files

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -113,7 +113,7 @@ All deltas within standard error. Task performance is preserved after 1.9x compr
 - **GPU and CPU supported** — runs on any CUDA GPU or CPU via standard PyTorch.
 - **`trust_remote_code=True` required** — OLMoE uses custom modeling code.
-- **Not fine-tunable** — compressed weights are read-only (`is_trainable = False`).
 - **Requires `helix-substrate`** — the quantizer is not built into transformers. You need `pip install "helix-substrate[hf]"`.
 - **64 experts = slow eval** — lm-eval-harness takes ~5.5 hours on a 3090 due to MoE routing overhead. Inference speed is normal for interactive use.

 - **GPU and CPU supported** — runs on any CUDA GPU or CPU via standard PyTorch.
 - **`trust_remote_code=True` required** — OLMoE uses custom modeling code.
+- **Fine-tunable via LoRA** — compressed weights remain frozen, but LoRA adapters attach to each `HelixLinear` layer via `HelixLinearSTE`. See `helix-substrate` for training infrastructure.
 - **Requires `helix-substrate`** — the quantizer is not built into transformers. You need `pip install "helix-substrate[hf]"`.
 - **64 experts = slow eval** — lm-eval-harness takes ~5.5 hours on a 3090 due to MoE routing overhead. Inference speed is normal for interactive use.