Instructions to use tencent/Hunyuan-A13B-Instruct-FP8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tencent/Hunyuan-A13B-Instruct-FP8 with Transformers:

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

pipe = pipeline("text-generation", model="tencent/Hunyuan-A13B-Instruct-FP8", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("tencent/Hunyuan-A13B-Instruct-FP8", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tencent/Hunyuan-A13B-Instruct-FP8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tencent/Hunyuan-A13B-Instruct-FP8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tencent/Hunyuan-A13B-Instruct-FP8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tencent/Hunyuan-A13B-Instruct-FP8

SGLang

How to use tencent/Hunyuan-A13B-Instruct-FP8 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 "tencent/Hunyuan-A13B-Instruct-FP8" \
    --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": "tencent/Hunyuan-A13B-Instruct-FP8",
		"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 "tencent/Hunyuan-A13B-Instruct-FP8" \
        --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": "tencent/Hunyuan-A13B-Instruct-FP8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tencent/Hunyuan-A13B-Instruct-FP8 with Docker Model Runner:
```
docker model run hf.co/tencent/Hunyuan-A13B-Instruct-FP8
```

manaestras commited on Jun 26, 2025

Commit

03c370c

verified ·

1 Parent(s): 0226108

Upload modeling_hunyuan.py with huggingface_hub

Browse files

Files changed (1) hide show

modeling_hunyuan.py +3 -3

modeling_hunyuan.py CHANGED Viewed

@@ -74,7 +74,7 @@ _CONFIG_FOR_DOC = "HunYuanConfig"
 def topkgating(logits: Tensor, topk: int):
     logits = logits.float()
     gates = F.softmax(logits, dim=1)
-   # expert_capacity = topk * gates.shape[0]
     expert_capacity = max(topk, topk * gates.shape[0] // gates.shape[1])
     num_experts = int(gates.shape[1])
     # Top-k router probability and corresponding expert indices for each token.
@@ -1417,7 +1417,7 @@ class HunYuanModel(HunYuanPreTrainedModel):
         )
-class HunYuanForCausalLM(HunYuanPreTrainedModel):
     _tied_weights_keys = ["lm_head.weight"]
     def __init__(self, config: HunYuanConfig):
@@ -1547,7 +1547,7 @@ class HunYuanForCausalLM(HunYuanPreTrainedModel):
             if isinstance(past_key_values, Cache):
                 cache_length = past_key_values.get_seq_length()
                 past_length = past_key_values.seen_tokens
-                max_cache_length = past_key_values.get_max_length()
             else:
                 cache_length = past_length = past_key_values[0][0].shape[2]
                 max_cache_length = None

 def topkgating(logits: Tensor, topk: int):
     logits = logits.float()
     gates = F.softmax(logits, dim=1)
+    # expert_capacity = topk * gates.shape[0]
     expert_capacity = max(topk, topk * gates.shape[0] // gates.shape[1])
     num_experts = int(gates.shape[1])
     # Top-k router probability and corresponding expert indices for each token.
         )
+class HunYuanMoEV1ForCausalLM(HunYuanPreTrainedModel):
     _tied_weights_keys = ["lm_head.weight"]
     def __init__(self, config: HunYuanConfig):
             if isinstance(past_key_values, Cache):
                 cache_length = past_key_values.get_seq_length()
                 past_length = past_key_values.seen_tokens
+                max_cache_length = past_key_values.get_max_cache_shape()
             else:
                 cache_length = past_length = past_key_values[0][0].shape[2]
                 max_cache_length = None