Instructions to use tencent/Hy4-preview with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tencent/Hy4-preview with Transformers:

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

pipe = pipeline("text-generation", model="tencent/Hy4-preview")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("tencent/Hy4-preview", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tencent/Hy4-preview with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tencent/Hy4-preview"
# 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/Hy4-preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tencent/Hy4-preview

SGLang

How to use tencent/Hy4-preview 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/Hy4-preview" \
    --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/Hy4-preview",
		"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/Hy4-preview" \
        --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/Hy4-preview",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tencent/Hy4-preview with Docker Model Runner:
```
docker model run hf.co/tencent/Hy4-preview
```

nielsr HF Staff commited on 1 day ago

Commit

7e11291

verified ·

1 Parent(s): 215d117

Add community evaluation results for DEEP-SWE, GPQA, SKILLSBENCH, SWE-BENCH_MULTILINGUAL, SWE-BENCH_PRO

Browse files

This PR adds community-provided evaluation results for the following benchmarks:

- **[DEEP-SWE](https://huggingface.co/datasets/datacurve/deep-swe?eval_result=tencent/Hy4-preview&leaderboard_task_id=deep_swe)**
- **[GPQA](https://huggingface.co/datasets/Idavidrein/gpqa?eval_result=tencent/Hy4-preview&leaderboard_task_id=diamond)**
- **[SKILLSBENCH](https://huggingface.co/datasets/benchflow/skillsbench?eval_result=tencent/Hy4-preview&leaderboard_task_id=skillsbench_v1_1)**
- **[SWE-BENCH_MULTILINGUAL](https://huggingface.co/datasets/SWE-bench/SWE-bench_Multilingual?eval_result=tencent/Hy4-preview&leaderboard_task_id=swe_bench_multilingual_%25_resolved)**
- **[SWE-BENCH_PRO](https://huggingface.co/datasets/ScaleAI/SWE-bench_Pro?eval_result=tencent/Hy4-preview&leaderboard_task_id=SWE_Bench_Pro)**

These results were extracted from the model card. This is based on the new [evaluation results feature](https://huggingface.co/docs/hub/eval-results).

*Note: This is an automated PR. Please review the evaluation results before merging.*

Files changed (5) hide show

.eval_results/deep-swe.yaml +7 -0
.eval_results/gpqa.yaml +7 -0
.eval_results/skillsbench.yaml +7 -0
.eval_results/swe-bench_multilingual.yaml +7 -0
.eval_results/swe-bench_pro.yaml +7 -0

.eval_results/deep-swe.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+- dataset:
+    id: datacurve/deep-swe
+    task_id: deep_swe
+  value: 64.3
+  source:
+    url: https://huggingface.co/tencent/Hy4-preview
+    name: Model Card

.eval_results/gpqa.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+- dataset:
+    id: Idavidrein/gpqa
+    task_id: diamond
+  value: 92.3
+  source:
+    url: https://huggingface.co/tencent/Hy4-preview
+    name: Model Card

.eval_results/skillsbench.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+- dataset:
+    id: benchflow/skillsbench
+    task_id: skillsbench_v1_1
+  value: 62.9
+  source:
+    url: https://huggingface.co/tencent/Hy4-preview
+    name: Model Card

.eval_results/swe-bench_multilingual.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+- dataset:
+    id: SWE-bench/SWE-bench_Multilingual
+    task_id: swe_bench_multilingual_%_resolved
+  value: 82.9
+  source:
+    url: https://huggingface.co/tencent/Hy4-preview
+    name: Model Card

.eval_results/swe-bench_pro.yaml ADDED Viewed

	@@ -0,0 +1,7 @@

+- dataset:
+    id: ScaleAI/SWE-bench_Pro
+    task_id: SWE_Bench_Pro
+  value: 65.7
+  source:
+    url: https://huggingface.co/tencent/Hy4-preview
+    name: Model Card