Linear probe checkpoints for https://footprints.baulab.info
Paper: https://arxiv.org/abs/2406.20086
To load a Llama-2-7b checkpoint at layer 0 and target index -3:
import torch
import torch.nn as nn
from huggingface_hub import hf_hub_download
class LinearModel(nn.Module):
def __init__(self, input_size, output_size, bias=False):
super(LinearModel, self).__init__()
self.fc = nn.Linear(input_size, output_size, bias=bias)
def forward(self, x):
output = self.fc(x)
return output
# example: llama-2-7b probe at layer 0, predicting 3 tokens ago
# predicting the next token would be `layer0_tgtidx1.ckpt`
checkpoint_path = hf_hub_download(
repo_id="sfeucht/footprints",
filename="llama-2-7b/layer0_tgtidx-3.ckpt"
)
# model_size is 4096 for both models.
# vocab_size is 32000 for Llama-2-7b and 128256 for Llama-3-8b
probe = LinearModel(4096, 32000).cuda()
probe.load_state_dict(torch.load(checkpoint_path))