Instructions to use baidu/ERNIE-4.5-21B-A3B-PT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use baidu/ERNIE-4.5-21B-A3B-PT with Transformers:

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

pipe = pipeline("text-generation", model="baidu/ERNIE-4.5-21B-A3B-PT")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("baidu/ERNIE-4.5-21B-A3B-PT")
model = AutoModelForCausalLM.from_pretrained("baidu/ERNIE-4.5-21B-A3B-PT", 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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use baidu/ERNIE-4.5-21B-A3B-PT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "baidu/ERNIE-4.5-21B-A3B-PT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "baidu/ERNIE-4.5-21B-A3B-PT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/baidu/ERNIE-4.5-21B-A3B-PT

SGLang

How to use baidu/ERNIE-4.5-21B-A3B-PT 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 "baidu/ERNIE-4.5-21B-A3B-PT" \
    --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": "baidu/ERNIE-4.5-21B-A3B-PT",
		"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 "baidu/ERNIE-4.5-21B-A3B-PT" \
        --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": "baidu/ERNIE-4.5-21B-A3B-PT",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use baidu/ERNIE-4.5-21B-A3B-PT with Docker Model Runner:
```
docker model run hf.co/baidu/ERNIE-4.5-21B-A3B-PT
```

self.loss_function = ErniePretrainingCriterion(config)没生效

by zhangliangx - opened Jul 2, 2025

Discussion

zhangliangx

Jul 2, 2025

在modeling_ernie4_5_moe.py 1323行，对loss_function的设置没生效，导致loss计算时报错

如图所示，执行完self.loss_function = ErniePretrainingCriterion(config)后loss function仍然是<function ForCausalLMLoss at 0x7fe9cbfbf250>。
我们发现原因是执行setattr时调用的是nn.Module的setattr，没有执行https://github.com/huggingface/transformers/blob/v4.53.0/src/transformers/modeling_utils.py#L5621 PretrainedModel的setter。进一步原因是ErniePretrainingCriterion是一个nn.Module，会优先执行到nn.Module的setattr将ErniePretrainingCriterion设置为一个模型。

最好的解决方案是将ErniePretrainingCriterion写成一个函数，因为所有的Module都会被注册到nn.Module的成员里。

其次，https://huggingface.co/baidu/ERNIE-4.5-21B-A3B-PT/blob/main/modeling_ernie4_5_moe.py#L962 ErniePretrainingCriterion中调用了self.config.get，而config是一个Ernie4_5_MoeConfig，没有get函数

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment