alpaca-lora/export_hf_checkpoint.py
Lily 630d1146c8
Update export_hf_checkpoint.py (#302)
* Update export_hf_checkpoint.py

* Update finetune.py

New tokenizer base model for the current dev branch of transformers

* Update generate.py

* Update export_state_dict_checkpoint.py

* Update export_hf_checkpoint.py
2023-04-09 14:07:59 -07:00

56 lines
1.4 KiB
Python

import os
import torch
import transformers
from peft import PeftModel
from transformers import LlamaForCausalLM, LlamaTokenizer # noqa: F402
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`" # noqa: E501
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,
"tloen/alpaca-lora-7b",
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)
# merge weights - new merging method from peft
lora_model = lora_model.merge_and_unload()
lora_model.train(False)
# did we do anything?
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, "./hf_ckpt", state_dict=deloreanized_sd, max_shard_size="400MB"
)