|
import os
|
|
|
|
import torch
|
|
import transformers
|
|
from peft import PeftModel
|
|
from transformers import LlamaForCausalLM, LlamaTokenizer
|
|
|
|
BASE_MODEL = os.environ.get("BASE_MODEL", None)
|
|
assert (
|
|
BASE_MODEL
|
|
), "Please specify a value for BASE_MODEL environment variable, e.g. `export BASE_MODEL=huggyllama/llama-7b`"
|
|
|
|
tokenizer = LlamaTokenizer.from_pretrained(BASE_MODEL)
|
|
|
|
base_model = LlamaForCausalLM.from_pretrained(
|
|
BASE_MODEL,
|
|
load_in_8bit=False,
|
|
torch_dtype=torch.float16,
|
|
device_map={"": "cpu"},
|
|
)
|
|
|
|
first_weight = base_model.model.layers[0].self_attn.q_proj.weight
|
|
first_weight_old = first_weight.clone()
|
|
|
|
lora_model = PeftModel.from_pretrained(
|
|
base_model,
|
|
"../outputs/lora-llama-clm-e2",
|
|
device_map={"": "cpu"},
|
|
torch_dtype=torch.float16,
|
|
)
|
|
|
|
lora_weight = lora_model.base_model.model.model.layers[0].self_attn.q_proj.weight
|
|
|
|
assert torch.allclose(first_weight_old, first_weight)
|
|
|
|
|
|
lora_model = lora_model.merge_and_unload()
|
|
|
|
lora_model.train(False)
|
|
|
|
|
|
assert not torch.allclose(first_weight_old, first_weight)
|
|
|
|
lora_model_sd = lora_model.state_dict()
|
|
deloreanized_sd = {
|
|
k.replace("base_model.model.", ""): v
|
|
for k, v in lora_model_sd.items()
|
|
if "lora" not in k
|
|
}
|
|
|
|
LlamaForCausalLM.save_pretrained(base_model, '../models/legal-base-7b', state_dict=deloreanized_sd, max_shard_size="400MB")
|
|
|