File size: 6,030 Bytes
e527d79 428b323 082b385 5678897 082b385 5678897 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 |
---
license: apache-2.0
---
### BioinspiredZephyr-7B: Large Language Model for the Mechanics of Biological and Bio-Inspired Materials
To accelerate discovery and guide insights, we report an open-source autoregressive transformer large language model (LLM), trained on expert knowledge in the biological materials field, especially focused on mechanics and structural properties.
The model is finetuned with a corpus of over a thousand peer-reviewed articles in the field of structural biological and bio-inspired materials and can be prompted to recall information, assist with research tasks, and function as an engine for creativity.
The model is based on HuggingFaceH4/zephyr-7b-beta.
This model is based on work reported in https://doi.org/10.1002/advs.202306724.
This repository includes both, Hugging Face transformers and GGUF files (in different versions, the q5_K_M is recommended).
#### Hugging Face transformers files: Loading and inference
```
from transformers import AutoModelForCausalLM, AutoTokenizer
from accelerate import infer_auto_device_map
model = AutoModelForCausalLM.from_pretrained(
model_name,
trust_remote_code=True,
device_map="auto", #device_map="cuda:0",
torch_dtype= torch.bfloat16,
# use_flash_attention_2=True,
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
```
Chat template
```
messages = [
{"role": "system", "content": "You are a friendly materials scientist."},
{"role": "user", "content": "What is the strongest spider silk material?"},
{"role": "assistant", "content": "Sample response."},
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
```
'<|system|>\nYou are a friendly materials scientist.</s>\n<|user|>\nWhat is the strongest spider silk material?</s>\n<|assistant|>\nSample response.</s>\n<|assistant|>\n'
```
device='cuda'
def generate_response (text_input="Biological materials offer amazing possibilities, such as",
num_return_sequences=1,
temperature=1.,
max_new_tokens=127,
num_beams=1,
top_k = 50,
top_p =0.9,repetition_penalty=1.,eos_token_id=2,verbatim=False,
exponential_decay_length_penalty_fac=None,
):
inputs = tokenizer.encode(text_input, add_special_tokens =False, return_tensors ='pt')
if verbatim:
print ("Length of input, tokenized: ", inputs.shape, inputs)
with torch.no_grad():
outputs = model.generate(input_ids=inputs.to(device),
max_new_tokens=max_new_tokens,
temperature=temperature, #value used to modulate the next token probabilities.
num_beams=num_beams,
top_k = top_k,
top_p =top_p,
num_return_sequences = num_return_sequences, eos_token_id=eos_token_id,
do_sample =True,
repetition_penalty=repetition_penalty,
)
return tokenizer.batch_decode(outputs[:,inputs.shape[1]:].detach().cpu().numpy(), skip_special_tokens=True)
```
Then:
```
messages = [
{"role": "system", "content": "You are a friendly materials scientist."},
{"role": "user", "content": "What is the strongest spider silk material?"},
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
output_text=generate_response (text_input=prompt, eos_token_id=eos_token,
num_return_sequences=1, repetition_penalty=1.,
top_p=0.9, top_k=512,
temperature=0.1,max_new_tokens=512, verbatim=False,
)
print (output_text)
```
#### GGUF files: Loading and inference
```
from llama_cpp import Llama
model_path='./BioinspiredZephyr-7B/ggml-model-q5_K_M.gguf'
chat_format="mistral-instruct"
llm = Llama(model_path=model_path,
n_gpu_layers=-1,verbose= True,
n_ctx=10000,
#main_gpu=0,
chat_format=chat_format,
#split_mode=llama_cpp.LLAMA_SPLIT_LAYER
)
```
Or, download directly from Hugging Face:
```
from llama_cpp import Llama
model_path='lamm-mit/BioinspiredZephyr-7B/ggml-model-q5_K_M.gguf'
chat_format="mistral-instruct"
llm = Llama.from_pretrained(
repo_id=model_path,
filename="*q5_K_M.gguf",
verbose=True,
n_gpu_layers=-1,
n_ctx=10000,
#main_gpu=0,
chat_format=chat_format,
)
```
For inference:
```
def generate_BioinspiredZephyr_7B(system_prompt='You are an expert in biological materials, mechanics and related topics.',
prompt="What is spider silk?",
temperature=0.0,
max_tokens=10000,
):
if system_prompt==None:
messages=[
{"role": "user", "content": prompt},
]
else:
messages=[
{"role": "system", "content": system_prompt},
{"role": "user", "content": prompt},
]
result=llm.create_chat_completion(
messages=messages,
temperature=temperature,
max_tokens=max_tokens,
)
start_time = time.time()
result=generate_BioinspiredZephyr_7B(system_prompt='You respond accurately.',
prompt="What is graphene? Answer with detail.",
max_tokens=512, temperature=0.7, )
print (result)
deltat=time.time() - start_time
print("--- %s seconds ---" % deltat)
toked=tokenizer(res)
print ("Tokens per second (generation): ", len (toked['input_ids'])/deltat)
``` |