Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Error using the local llama3.1 model #609

Open
CzsGit opened this issue Aug 15, 2024 · 1 comment
Open

Error using the local llama3.1 model #609

CzsGit opened this issue Aug 15, 2024 · 1 comment

Comments

@CzsGit
Copy link

CzsGit commented Aug 15, 2024

I tried to load the local model and ran into this issue

Error:

raise ValueError(
ValueError: rope_scaling must be a dictionary with with two fields, type and factor, got {'factor': 8.0, 'low_freq_factor': 1.0,
'high_freq_factor': 4.0, 'original_max_position_embeddings': 8192, 'rope_type': 'llama3'}

Test code:

from transformers import AutoTokenizer
from petals import AutoDistributedModelForCausalLM

# Choose any model available at https://health.petals.dev
model_name = "./ckpt/Meta-Llama-3.1-405B-Instruct"

# Connect to a distributed network hosting model layers
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoDistributedModelForCausalLM.from_pretrained(model_name)

# Run the model as if it were on your computer
inputs = tokenizer("A cat sat", return_tensors="pt")["input_ids"]
outputs = model.generate(inputs, max_new_tokens=5)
print(tokenizer.decode(outputs[0]))  # A cat sat on a mat...

The current problem is that transformers are in the wrong version, 4.33.3 should be used to solve this problem, but 4.33.3 is not compatible with petals.

petals:2.2.0.post1

@xloem
Copy link

xloem commented Oct 19, 2024

i fixed this by using the latest petals from git
petals @ git+https://github.com/bigscience-workshop/petals@22afba627a7eb4fcfe9418c49472c6a51334b8ac

i also found it was super helpful to pass torch_dtype='bfloat16' as then it can map the weights straight to disk and loads immediately

but it says no servers are running ..?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants