Text Generation
Transformers
Safetensors
Chinese
English
yebb_moe
mixture-of-experts
pretrained
base-model
research
custom_code
Instructions to use sinksilk/Yebb-MoE-2B-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use sinksilk/Yebb-MoE-2B-Base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="sinksilk/Yebb-MoE-2B-Base", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("sinksilk/Yebb-MoE-2B-Base", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use sinksilk/Yebb-MoE-2B-Base with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "sinksilk/Yebb-MoE-2B-Base" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "sinksilk/Yebb-MoE-2B-Base", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/sinksilk/Yebb-MoE-2B-Base
- SGLang
How to use sinksilk/Yebb-MoE-2B-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 "sinksilk/Yebb-MoE-2B-Base" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "sinksilk/Yebb-MoE-2B-Base", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'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 "sinksilk/Yebb-MoE-2B-Base" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "sinksilk/Yebb-MoE-2B-Base", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use sinksilk/Yebb-MoE-2B-Base with Docker Model Runner:
docker model run hf.co/sinksilk/Yebb-MoE-2B-Base
Download export_validation.json from sinksilk/Yebb-MoE-2B-Base: direct link, hf CLI and curl.
- Browser
- Download file 1 kB
-
https://huggingface.co/sinksilk/Yebb-MoE-2B-Base/resolve/main/export_validation.json
- Command line
-
hf download hf://sinksilk/Yebb-MoE-2B-Base/export_validation.json
-
curl -L -o export_validation.json https://huggingface.co/sinksilk/Yebb-MoE-2B-Base/resolve/main/export_validation.json
1 kB
| { | |
| "status": "passed_CPU_FP32_export_verification", | |
| "torch_version": "2.10.0+cpu", | |
| "transformers_version": "5.20.2", | |
| "parameters": 1962164864, | |
| "all_state_tensors_exact": true, | |
| "state_tensor_count": 176, | |
| "all_state_tensors_finite": true, | |
| "reference_logits": [ | |
| { | |
| "text": "中国的首都是", | |
| "tokens": 4, | |
| "all_logits_bit_exact": true | |
| }, | |
| { | |
| "text": "The capital of France is", | |
| "tokens": 6, | |
| "all_logits_bit_exact": true | |
| }, | |
| { | |
| "text": "def is_even(n):\n return", | |
| "tokens": 10, | |
| "all_logits_bit_exact": true | |
| } | |
| ], | |
| "packed_document_mask_logits_bit_exact": true, | |
| "left_padded_batch_matches_single_inputs": true, | |
| "HF_generate_smoke_passed": true, | |
| "generate_smoke_output_ids": [ | |
| 1, | |
| 5297, | |
| 1851, | |
| 2431, | |
| 12935, | |
| 1439, | |
| 5297, | |
| 1851 | |
| ], | |
| "elapsed_seconds": 24.31835330836475, | |
| "device": "cpu", | |
| "scope": "Export/loading/numerical fidelity only, not a new capability evaluation." | |
| } | |