Instructions to use Xuezha/RecombinationTransformer-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Xuezha/RecombinationTransformer-base with Transformers:

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

pipe = pipeline("text-generation", model="Xuezha/RecombinationTransformer-base", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Xuezha/RecombinationTransformer-base", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Xuezha/RecombinationTransformer-base with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/Xuezha/RecombinationTransformer-base

SGLang

How to use Xuezha/RecombinationTransformer-base 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 "Xuezha/RecombinationTransformer-base" \
    --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": "Xuezha/RecombinationTransformer-base",
		"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 "Xuezha/RecombinationTransformer-base" \
        --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": "Xuezha/RecombinationTransformer-base",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Xuezha/RecombinationTransformer-base with Docker Model Runner:
```
docker model run hf.co/Xuezha/RecombinationTransformer-base
```

Xuezha commited on Jun 8, 2024

Commit

5ba2653

verified ·

1 Parent(s): 57c1748

Update modeling.py

Browse files

Files changed (1) hide show

modeling.py +3 -0

modeling.py CHANGED Viewed

@@ -185,6 +185,9 @@ class RecombinationTransformerForCausalLM(PreTrainedModel):
         if past:
             input_ids = input_ids[:, -1].unsqueeze(-1)
         return {"input_ids": input_ids, "attention_mask": attention_mask, "past_key_values": past}
     def generate(self, input_ids, attention_mask=None, max_length=512, min_length=None, num_return_sequences=1):

         if past:
             input_ids = input_ids[:, -1].unsqueeze(-1)
+        if attention_mask is None:
+            attention_mask = torch.ones(input_ids.shape, device=input_ids.device)
         return {"input_ids": input_ids, "attention_mask": attention_mask, "past_key_values": past}
     def generate(self, input_ids, attention_mask=None, max_length=512, min_length=None, num_return_sequences=1):