File size: 2,109 Bytes
578c3d0 9a302a2 e00aa88 9a302a2 84bbd47 9a302a2 63d8b31 9a302a2 63d8b31 9a302a2 63d8b31 9a302a2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
---
license: apache-2.0
---
[Optimum Habana](https://github.com/huggingface/optimum-habana) is the interface between the Transformers library and Habana's Gaudi processor (HPU). It provides a set of tools enabling easy and fast model loading and fine-tuning on single- and multi-HPU settings for different downstream tasks.
Learn more about how to take advantage of the power of Habana HPUs to train Transformers models at [hf.co/hardware/habana](https://huggingface.co/hardware/habana).
## DistilBERT Base model HPU configuration
This model contains just the `GaudiConfig` file for running the [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) model on Habana's Gaudi processors (HPU).
**This model contains no model weights, only a GaudiConfig.**
This enables to specify:
- `use_habana_mixed_precision`: whether to use Habana Mixed Precision (HMP)
- `hmp_opt_level`: optimization level for HMP, see [here](https://docs.habana.ai/en/latest/PyTorch/PyTorch_User_Guide/PT_Mixed_Precision.html#configuration-options) for a detailed explanation
- `hmp_bf16_ops`: list of operators that should run in bf16
- `hmp_fp32_ops`: list of operators that should run in fp32
- `hmp_is_verbose`: verbosity
- `use_fused_adam`: whether to use Habana's custom AdamW implementation
- `use_fused_clip_norm`: whether to use Habana's fused gradient norm clipping operator
## Usage
The model is instantiated the same way as in the Transformers library.
The only difference is that there are a few new training arguments specific to HPUs:
```
from optimum.habana import GaudiTrainer, GaudiTrainingArguments
from transformers import DistilBertTokenizer, DistilBertModel
tokenizer = DistilBertTokenizer.from_pretrained("distilbert-base-uncased")
model = DistilBertModel.from_pretrained("distilbert-base-uncased")
args = GaudiTrainingArguments(
output_dir="/tmp/output_dir",
use_habana=True,
use_lazy_mode=True,
gaudi_config_name="Habana/distilbert-base-uncased",
)
trainer = GaudiTrainer(
model=model,
args=args,
tokenizer=tokenizer,
)
trainer.train()
```
|