Views
No views yet
pip install llama-cpp-python1from llama_cpp import Llama
2
3llm = Llama(
4 model_path="path/to/your/model.gguf",
5 n_ctx=8192,
6 penalize_nl=False
7)
8
9messages = [{"role": "user", "content": "Кога е основан Софийският университет?"}]
10response = llm.create_chat_completion(
11 messages=messages,
12 max_tokens=2048, # Choose maximum generated tokens
13 temperature=0.1,
14 top_p=0.9,
15 repeat_penalty=1.0,
16 stop=["<eos>", "<end_of_turn>"]
17)
18
191from llama_cpp import Llama
2
3llm = Llama(
4 model_path="path/to/your/model.gguf",
5 n_ctx=8192,
6 penalize_nl=False
7)
8
9prompt = "<start_of_turn>user\nКога е основан Софийският университет?<end_of_turn>\n<start_of_turn>model\n"
10response = llm(
11 prompt,
12 max_tokens=2048, # Choose maximum generated tokens
13 temperature=0.1,
14 top_p=0.9,
15 repeat_penalty=1.0,
16 stop=["<eos>","<end_of_turn>"]
17)