Update README.md
Browse files
README.md
CHANGED
@@ -40,13 +40,13 @@ The model is fine-tuned by [UER-py](https://github.com/dbiir/UER-py/) on [Tencen
|
|
40 |
|
41 |
```
|
42 |
python3 run_ner.py --pretrained_model_path models/cluecorpussmall_roberta_base_seq512_model.bin-250000 \
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
```
|
51 |
|
52 |
Finally, we convert the pre-trained model into Huggingface's format:
|
|
|
40 |
|
41 |
```
|
42 |
python3 run_ner.py --pretrained_model_path models/cluecorpussmall_roberta_base_seq512_model.bin-250000 \
|
43 |
+
--vocab_path models/google_zh_vocab.txt \
|
44 |
+
--train_path datasets/cluener2020/train.tsv \
|
45 |
+
--dev_path datasets/cluener2020/dev.tsv \
|
46 |
+
--label2id_path datasets/cluener2020/label2id.json \
|
47 |
+
--output_model_path models/cluener2020_classifier_model.bin \
|
48 |
+
--learning_rate 3e-5 --batch_size 32 --epochs_num 5 --seq_length 512 \
|
49 |
+
--embedding word_pos_seg --encoder transformer --mask fully_visible
|
50 |
```
|
51 |
|
52 |
Finally, we convert the pre-trained model into Huggingface's format:
|