Text Generation
Transformers
Safetensors
German
English
hanse
causal-lm
custom-code
research
custom_code
Instructions to use Evicka/HanseLM-78M-Base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Evicka/HanseLM-78M-Base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Evicka/HanseLM-78M-Base", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("Evicka/HanseLM-78M-Base", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Evicka/HanseLM-78M-Base with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Evicka/HanseLM-78M-Base" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Evicka/HanseLM-78M-Base", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/Evicka/HanseLM-78M-Base
- SGLang
How to use Evicka/HanseLM-78M-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 "Evicka/HanseLM-78M-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": "Evicka/HanseLM-78M-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 "Evicka/HanseLM-78M-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": "Evicka/HanseLM-78M-Base", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use Evicka/HanseLM-78M-Base with Docker Model Runner:
docker model run hf.co/Evicka/HanseLM-78M-Base
File size: 579 Bytes
e8ac551 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | {
"checkpoint": "hanse-78m-pretrain-phase2/final",
"kind": "held-out next-token language-model evaluation",
"sequence_length": 2048,
"batch_size": 6,
"batches": 200,
"seed": 1234,
"results": {
"fineweb2_de": {"loss": 3.310921, "perplexity": 27.410},
"finewiki_de": {"loss": 2.574947, "perplexity": 13.131},
"fineweb_en": {"loss": 2.943911, "perplexity": 18.990},
"mixed_validation": {"loss": 3.175663, "perplexity": 23.943}
},
"warning": "Corpus perplexities are tokenizer- and corpus-dependent, and are not downstream capability benchmarks."
}
|