Commit
•
bf7d1d7
1
Parent(s):
feeec9d
Update README.md (#2)
Browse files- Update README.md (415055e3e5924a852f6ff44ac8195fe560e10b46)
- Update README.md (285dcfcfee0924093e43d766092e0be548ec83f9)
Co-authored-by: Adam Stachowicz <astachowicz@users.noreply.huggingface.co>
README.md
CHANGED
|
@@ -25,7 +25,7 @@ It is strongly recommended to train this model doing bf16 mixed-precision traini
|
|
| 25 |
|
| 26 |
[Here](https://github.com/huggingface/optimum-habana/blob/main/examples/question-answering/run_qa.py) is a question-answering example script to fine-tune a model on SQuAD. You can run it with BERT with the following command:
|
| 27 |
```bash
|
| 28 |
-
python run_qa.py \
|
| 29 |
--model_name_or_path bert-base-uncased \
|
| 30 |
--gaudi_config_name Habana/bert-base-uncased \
|
| 31 |
--dataset_name squad \
|
|
@@ -38,7 +38,9 @@ python run_qa.py \
|
|
| 38 |
--max_seq_length 384 \
|
| 39 |
--output_dir /tmp/squad/ \
|
| 40 |
--use_habana \
|
| 41 |
-
--
|
|
|
|
|
|
|
| 42 |
--throughput_warmup_steps 3 \
|
| 43 |
--bf16
|
| 44 |
```
|
|
|
|
| 25 |
|
| 26 |
[Here](https://github.com/huggingface/optimum-habana/blob/main/examples/question-answering/run_qa.py) is a question-answering example script to fine-tune a model on SQuAD. You can run it with BERT with the following command:
|
| 27 |
```bash
|
| 28 |
+
PT_HPU_LAZY_MODE=0 python run_qa.py \
|
| 29 |
--model_name_or_path bert-base-uncased \
|
| 30 |
--gaudi_config_name Habana/bert-base-uncased \
|
| 31 |
--dataset_name squad \
|
|
|
|
| 38 |
--max_seq_length 384 \
|
| 39 |
--output_dir /tmp/squad/ \
|
| 40 |
--use_habana \
|
| 41 |
+
--torch_compile_backend hpu_backend \
|
| 42 |
+
--torch_compile \
|
| 43 |
+
--use_lazy_mode false \
|
| 44 |
--throughput_warmup_steps 3 \
|
| 45 |
--bf16
|
| 46 |
```
|