Instructions to use elsagranger/VirtualCompiler with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use elsagranger/VirtualCompiler with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="elsagranger/VirtualCompiler")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("elsagranger/VirtualCompiler") model = AutoModelForCausalLM.from_pretrained("elsagranger/VirtualCompiler", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use elsagranger/VirtualCompiler with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "elsagranger/VirtualCompiler" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "elsagranger/VirtualCompiler", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/elsagranger/VirtualCompiler
- SGLang
How to use elsagranger/VirtualCompiler 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 "elsagranger/VirtualCompiler" \ --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": "elsagranger/VirtualCompiler", "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 "elsagranger/VirtualCompiler" \ --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": "elsagranger/VirtualCompiler", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use elsagranger/VirtualCompiler with Docker Model Runner:
docker model run hf.co/elsagranger/VirtualCompiler
| import requests | |
| import random | |
| CL = ['clang-11', 'clang-12', 'clang-9', 'gcc-11', 'gcc-7', 'gcc-9'] | |
| OP = ['O0', 'O1', 'O2', 'O3', 'Os'] | |
| ST = ['stripped', 'unstripped'] | |
| def process(source_code): | |
| compiler = random.choice(CL) | |
| optimizer = random.choice(OP) | |
| strip_type = random.choice(ST) | |
| prompt = f'Please compile this source code using {compiler} with optimization level {optimizer} into assembly code.' | |
| if strip_type == 'stripped': | |
| prompt += ' Strip the assembly code.' | |
| else: | |
| prompt += ' No strip the assembly code.' | |
| query_prompt = "<s>system\n" + prompt + \ | |
| "</s>\n<s>user\n" + source_code + "</s>\n<s>assistant\n" | |
| return query_prompt, compiler, optimizer, strip_type | |
| def do_request(src): | |
| url = "http://localhost:8080/v1/completions" | |
| query_prompt, _, _, _ = process(src) | |
| model_name = "VirtualCompiler" | |
| ret = requests.post(url, json={ | |
| "prompt": query_prompt, | |
| "max_tokens": 4096, | |
| "temperature": 0.3, | |
| "stop": ["</s>"], | |
| "model": model_name, | |
| "echo": False, | |
| "logprobs": True, | |
| }) | |
| return ret.json() | |
| src = '''static int | |
| layout_append(struct layout_cell *lc, char *buf, size_t len) | |
| { | |
| if (len == 0) | |
| return (-1); | |
| return (0); | |
| } | |
| ''' | |
| ret = do_request(src) | |
| print(ret['choices'][0]['text']) | |