Instructions to use CodyBontecou/llada-8b-instruct-duplicate with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use CodyBontecou/llada-8b-instruct-duplicate with Transformers:

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

pipe = pipeline("text-generation", model="CodyBontecou/llada-8b-instruct-duplicate", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("CodyBontecou/llada-8b-instruct-duplicate", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use CodyBontecou/llada-8b-instruct-duplicate with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CodyBontecou/llada-8b-instruct-duplicate"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "CodyBontecou/llada-8b-instruct-duplicate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/CodyBontecou/llada-8b-instruct-duplicate

SGLang

How to use CodyBontecou/llada-8b-instruct-duplicate 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 "CodyBontecou/llada-8b-instruct-duplicate" \
    --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": "CodyBontecou/llada-8b-instruct-duplicate",
		"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 "CodyBontecou/llada-8b-instruct-duplicate" \
        --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": "CodyBontecou/llada-8b-instruct-duplicate",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use CodyBontecou/llada-8b-instruct-duplicate with Docker Model Runner:
```
docker model run hf.co/CodyBontecou/llada-8b-instruct-duplicate
```

CodyBontecou commited on Mar 8, 2025

Commit

b38d2da

1 Parent(s): d19a591

output serialization

Browse files

Files changed (1) hide show

handler.py +8 -3

handler.py CHANGED Viewed

@@ -36,9 +36,14 @@ class EndpointHandler:
         with torch.no_grad():
             outputs = self.model(**inputs)
-        # Process outputs - this depends on your specific model and requirements
-        # For now, we'll just return the input as the output to fix the array format issue
-        return [{"input_text": input_text, "generated_text": outputs}]
 if __name__ == "__main__":

         with torch.no_grad():
             outputs = self.model(**inputs)
+        # Process outputs - convert tensors to serializable format
+        # Extract the last hidden state and convert to list for JSON serialization
+        last_hidden_state = outputs.last_hidden_state
+        # Convert to Python list (serializable) - using the mean of the embeddings as a simple approach
+        embedding = last_hidden_state.mean(dim=1).cpu().numpy().tolist()
+        return [{"input_text": input_text, "embedding": embedding}]
 if __name__ == "__main__":