mobiuslabsgmbh

/

aanaphi2-v0.1

like 28

Follow

Mobius Labs GmbH 53

Hi! Can you try this, but might be a bit slow:

#Load model
import transformers, torch
compute_dtype = torch.float32
cache_path    = ''
device        = 'cpu'
model_id      = "mobiuslabsgmbh/aanaphi2-v0.1"
model         = transformers.AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=compute_dtype, 
                                                                  cache_dir=cache_path,
                                                                  device_map=device)
tokenizer     = transformers.AutoTokenizer.from_pretrained(model_id, cache_dir=cache_path)

#Set Prompt format
instruction_template = "### Human: "
response_template    = "### Assistant: "
def prompt_format(prompt):
    out = instruction_template + prompt + '\n' + response_template
    return out
model.eval();


def generate(prompt, max_length=1024):
    prompt_chat = prompt_format(prompt)
    inputs      = tokenizer(prompt_chat, return_tensors="pt", return_attention_mask=True).to(device)
    outputs     = model.generate(**inputs, max_length=max_length, eos_token_id= tokenizer.eos_token_id) 
    text        = tokenizer.batch_decode(outputs[:,:-1])[0]
    return text

#Generate
print(generate('If A+B=C and B=C, what would be the value of A?'))

Hi! Can you try this, but might be a bit slow:

#Load model
import transformers, torch
compute_dtype = torch.float32
cache_path    = ''
device        = 'cpu'
model_id      = "mobiuslabsgmbh/aanaphi2-v0.1"
model         = transformers.AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=compute_dtype, 
                                                                  cache_dir=cache_path,
                                                                  device_map=device)
tokenizer     = transformers.AutoTokenizer.from_pretrained(model_id, cache_dir=cache_path)

#Set Prompt format
instruction_template = "### Human: "
response_template    = "### Assistant: "
def prompt_format(prompt):
    out = instruction_template + prompt + '\n' + response_template
    return out
model.eval();


def generate(prompt, max_length=1024):
    prompt_chat = prompt_format(prompt)
    inputs      = tokenizer(prompt_chat, return_tensors="pt", return_attention_mask=True).to(device)
    outputs     = model.generate(**inputs, max_length=max_length, eos_token_id= tokenizer.eos_token_id) 
    text        = tokenizer.batch_decode(outputs[:,:-1])[0]
    return text

#Generate
print(generate('If A+B=C and B=C, what would be the value of A?'))

Thank you for your answer! I tried this, but I get a Killed signal. I checked my cpu and memory states while it was executing and none of them were close to being full. I wonder what might be the issue!

Strange, it shouldn't use too much RAM to get killed!

You can also just run it on Google colab with the free GPU, I have just tried there and it works fine. You'll need to install the following before you run the code:
pip install transformers, accelerate

Strange, it shouldn't use too much RAM to get killed!

You can also just run it on Google colab with the free GPU, I have just tried there and it works fine. You'll need to install the following before you run the code:
pip install transformers, accelerate

Thank you so much for all your help, I really appreciate it!

Happy to help!