87 lines
3.8 KiB
Python
87 lines
3.8 KiB
Python
|
import sys
|
||
|
|
import shutil
|
||
|
|
import json
|
||
|
|
sys.path.insert(0, "/home/wangshuo1/projects/CPM-9G/gejiu_train")
|
||
|
|
import torch
|
||
|
|
from transformers import LlamaTokenizer, LlamaForCausalLM, LlamaConfig
|
||
|
|
from cpm.cpm9g.models import CPM9GTorch, CPM9GConfig
|
||
|
|
|
||
|
|
from transformers import AutoTokenizer, AutoConfig
|
||
|
|
from cpm.cpm9g import CPM9GTokenizer as BMTCPM9GTokenizer
|
||
|
|
from cpm.cpm9g import CPM9GTokenizer
|
||
|
|
from cpm.cpm9g.generation.cpm9g import CPM9GBeamSearch
|
||
|
|
|
||
|
|
source_path = "/data/public/zwl_data/11b-base/"
|
||
|
|
target_path = "/home/wangshuo1/projects/CPM-9G/convert_to_hf/11b-base-hf/"
|
||
|
|
file_name = "11b.pt"
|
||
|
|
|
||
|
|
def convert_pkl():
|
||
|
|
shutil.copyfile(f"{source_path}vocabs.txt", f"{target_path}vocabs.txt")
|
||
|
|
with open(f"{source_path}config.json") as f:
|
||
|
|
bmt_config = json.load(f)
|
||
|
|
config = {
|
||
|
|
"architectures": [
|
||
|
|
"LlamaForCausalLM"
|
||
|
|
],
|
||
|
|
"bos_token_id": 1,
|
||
|
|
"eos_token_id": 2,
|
||
|
|
"hidden_act": "silu",
|
||
|
|
"hidden_size": bmt_config['dim_model'],
|
||
|
|
"initializer_range": 0.02,
|
||
|
|
"intermediate_size": bmt_config['dim_ff'],
|
||
|
|
"max_length": 4096,
|
||
|
|
"max_position_embeddings": 4096,
|
||
|
|
"model_type": "llama",
|
||
|
|
"num_attention_heads": bmt_config['num_heads'],
|
||
|
|
"num_hidden_layers": bmt_config['num_layers'],
|
||
|
|
"num_key_value_heads": bmt_config['num_kv_heads'],
|
||
|
|
"pad_token_id": 0,
|
||
|
|
"pretraining_tp": 1,
|
||
|
|
"rms_norm_eps": 1e-05,
|
||
|
|
"rope_scaling": None,
|
||
|
|
"tie_word_embeddings": False,
|
||
|
|
"torch_dtype": "float32",
|
||
|
|
"transformers_version": "4.31.0",
|
||
|
|
"use_cache": True,
|
||
|
|
"vocab_size": bmt_config['vocab_size'],
|
||
|
|
}
|
||
|
|
with open(f"{target_path}config.json", "w") as f:
|
||
|
|
json.dump(config, f)
|
||
|
|
|
||
|
|
state = torch.load(f"{source_path}{file_name}")
|
||
|
|
new_state = {}
|
||
|
|
new_state["model.embed_tokens.weight"] = state["input_embedding.weight"]
|
||
|
|
new_state["lm_head.weight"] = state["lm_head.weight"]
|
||
|
|
new_state["model.norm.weight"] = state["encoder.output_layernorm.weight"]
|
||
|
|
layer_num = bmt_config['num_layers']
|
||
|
|
for lid in range(layer_num):
|
||
|
|
print(lid)
|
||
|
|
new_state[f"model.layers.{lid}.self_attn.q_proj.weight"] = state[f"encoder.layers.{lid}.self_att.self_attention.project_q.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.self_attn.k_proj.weight"] = state[f"encoder.layers.{lid}.self_att.self_attention.project_k.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.self_attn.v_proj.weight"] = state[f"encoder.layers.{lid}.self_att.self_attention.project_v.weight"]
|
||
|
|
|
||
|
|
new_state[f"model.layers.{lid}.self_attn.o_proj.weight"] = state[f"encoder.layers.{lid}.self_att.self_attention.attention_out.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.mlp.gate_proj.weight"] = state[f"encoder.layers.{lid}.ffn.ffn.w_in.w_0.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.mlp.up_proj.weight"] = state[f"encoder.layers.{lid}.ffn.ffn.w_in.w_1.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.mlp.down_proj.weight"] = state[f"encoder.layers.{lid}.ffn.ffn.w_out.weight"]
|
||
|
|
|
||
|
|
new_state[f"model.layers.{lid}.input_layernorm.weight"] = state[f"encoder.layers.{lid}.self_att.layernorm_before_attention.weight"]
|
||
|
|
new_state[f"model.layers.{lid}.post_attention_layernorm.weight"] = state[f"encoder.layers.{lid}.ffn.layernorm_before_ffn.weight"]
|
||
|
|
del state
|
||
|
|
state = None
|
||
|
|
torch.save(new_state, f"{target_path}pytorch_model.bin")
|
||
|
|
|
||
|
|
def test():
|
||
|
|
config = LlamaConfig.from_pretrained(f"{target_path}")
|
||
|
|
tokenizer = CPM9GTokenizer(f"{target_path}vocabs.txt")
|
||
|
|
model = LlamaForCausalLM.from_pretrained(f"{target_path}").cuda()
|
||
|
|
|
||
|
|
text = "请介绍一下清华大学："
|
||
|
|
inputs = torch.tensor([[tokenizer.bos_id] + tokenizer.encode(text)]).cuda()
|
||
|
|
output = model.generate(inputs, max_length=200)[0].tolist()
|
||
|
|
print(tokenizer.decode(output))
|
||
|
|
|
||
|
|
if __name__ == "__main__":
|
||
|
|
convert_pkl()
|
||
|
|
test()
|