Views
No views yet
deepseek-llm-7b-chat is a 7B parameter model initialized from deepseek-llm-7b-base and fine-tuned on extra instruction data.1import torch
2from transformers import AutoTokenizer, AutoModelForCausalLM, GenerationConfig
3
4model_name = "libraxiong/deepseek-zhenhuan-lora"
5tokenizer = AutoTokenizer.from_pretrained(model_name)
6model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.bfloat16, device_map="auto")
7model.generation_config = GenerationConfig.from_pretrained(model_name)
8model.generation_config.pad_token_id = model.generation_config.eos_token_id
9
10messages = [
11 {"role": "user", "content": "Who are you?"}
12]
13input_tensor = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt")
14outputs = model.generate(input_tensor.to(model.device), max_new_tokens=100)
15
16result = tokenizer.decode(outputs[0][input_tensor.shape[1]:], skip_special_tokens=True)
17print(result)apply_chat_template, you can also interact with our model following the sample template. Note that messages should be replaced by your input.User: {messages[0]['content']}
Assistant: {messages[1]['content']}<｜end▁of▁sentence｜>User: {messages[2]['content']}
Assistant:add_special_tokens=True), our tokenizer automatically adds a bos_token (<｜begin▁of▁sentence｜>) before the input text. Additionally, since the system prompt is not compatible with this version of our models, we DO NOT RECOMMEND including the system prompt in your input.