Views
No views yet
### System:
You are FF-LLM, a helpful assistant.
### Instruction:
{your question here}
### Response:1from transformers import GPT2LMHeadModel, GPT2Tokenizer
2
3model = GPT2LMHeadModel.from_pretrained("ff-llm/FF_3")
4tokenizer = GPT2Tokenizer.from_pretrained("ff-llm/FF_3")
5
6prompt = (
7 "### System:\nYou are FF-LLM, a helpful assistant.\n\n"
8 "### Instruction:\nWhat is the capital of France?\n\n### Response:\n"
9)
10input_ids = tokenizer.encode(prompt, return_tensors="pt")
11output = model.generate(
12 input_ids, max_new_tokens=256, do_sample=True,
13 temperature=0.7, top_p=0.9,
14 eos_token_id=tokenizer.eos_token_id,
15 pad_token_id=tokenizer.eos_token_id,
16)
17print(tokenizer.decode(output[0][input_ids.shape[1]:], skip_special_tokens=True))ollama run ff-llm/FF_3