uer commited on
Commit
820052c
1 Parent(s): 7ae0c0e

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +5 -4
README.md CHANGED
@@ -46,12 +46,13 @@ python3 pretrain.py --dataset_path lyric_dataset.pt \
46
  --pretrained_model_path gpt2-base-chinese-cluecorpussmall/pytorch_model.bin\
47
  --vocab_path models/google_zh_vocab.txt \
48
  --output_model_path models/lyric_gpt2_seq512_model.bin \
49
- --config_path models/bert_base_config.json --learning_rate 5e-5 \
50
- --world_size 8 --gpu_ranks 0 1 2 3 4 5 6 7 --tie_weight \
 
 
51
  --embedding word_pos --remove_embedding_layernorm \
52
  --encoder transformer --mask causal --layernorm_positioning pre \
53
- --target lm --batch_size 64 --total_steps 100000 \
54
- --save_checkpoint_steps 10000 --report_steps 5000
55
  ```
56
 
57
  Finally, we convert the pre-trained model into Huggingface's format:
 
46
  --pretrained_model_path gpt2-base-chinese-cluecorpussmall/pytorch_model.bin\
47
  --vocab_path models/google_zh_vocab.txt \
48
  --output_model_path models/lyric_gpt2_seq512_model.bin \
49
+ --config_path models/bert_base_config.json \
50
+ --world_size 8 --gpu_ranks 0 1 2 3 4 5 6 7 \
51
+ --total_steps 100000 --save_checkpoint_steps 10000 --report_steps 5000 \
52
+ --learning_rate 5e-5 --batch_size 64 \
53
  --embedding word_pos --remove_embedding_layernorm \
54
  --encoder transformer --mask causal --layernorm_positioning pre \
55
+ --target lm --tie_weight
 
56
  ```
57
 
58
  Finally, we convert the pre-trained model into Huggingface's format: