Instructions to use ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid with Transformers:

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

pipe = pipeline("text-generation", model="ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid

SGLang

How to use ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid 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 "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid" \
    --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": "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid",
		"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 "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid" \
        --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": "ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid with Docker Model Runner:
```
docker model run hf.co/ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Model Card for Echo-DSRN-Qwen2.5-0.5B-Hybrid

This repository contains experimental models designed strictly for academic evaluation and research purposes.

Critical Constraints:

No Production Deployment: Experimental models must not be deployed in commercial, enterprise, or mission-critical environments under any circumstances.

No Liability: Experimental models are provided "as-is" without warranties of any kind. The developers assume zero liability for downstream consequences, system integration failures, or regulatory non-compliance resulting from unauthorized deployment.

Echo-DSRN-Hybrid: Transformers Backbone + Surprise-Gated Dual-State Recurrent Architecture

Echo-DSRN-Hybrid is a hybrid recurrent architecture designed for resource-constrained deployment on narrow, well-defined tasks (e.g., intent routing, NER, semantic classification).

The Echo-DSRN memory-injectors combine three parallel computational paths within each block:

Fast GRU state: Tracks short-range token dynamics, updated every token.
Surprise-gated slow state: Selectively accumulates long-range information, write-protected by default and triggered by prediction error.
Sliding window attention: Handles fine-grained local dependencies within a bounded context window (128 tokens).

Echo-DSRN offers constant memory overhead (O(1) recurrent core + bounded O(window_size) attention) during generation.

Downloads last month: -

Safetensors

Model size

0.7B params

Tensor type

F32

Model tree for ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid

Finetunes

2 models

Datasets used to train ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid

Collection including ethicalabs/Echo-DSRN-Qwen2.5-0.5B-Hybrid

Echo-DSRN - Transformers Hybrid

Collection

Qwen2.5-0.5B-Instruct (backbone) + Echo-DSRN (memory) • 2 items • Updated Jul 21