File size: 2,335 Bytes
7865ccc
 
 
329f7ea
 
f3f6ab8
329f7ea
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
bca38f0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
329f7ea
bca38f0
 
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
49
---
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).

## ALBERT Large model HPU configuration

This model contains just the `GaudiConfig` file for running the [albert-large-v2](https://huggingface.co/albert-large-v2) 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.

[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 ALBERT Large with the following command:
```bash
python run_qa.py \
  --model_name_or_path albert-large-v2 \
  --gaudi_config_name Habana/albert-large-v2 \
  --dataset_name squad \
  --do_train \
  --do_eval \
  --per_device_train_batch_size 32 \
  --per_device_eval_batch_size 4 \
  --learning_rate 5e-5 \
  --num_train_epochs 2 \
  --max_seq_length 384 \
  --doc_stride 128 \
  --output_dir /tmp/squad/ \
  --use_habana \
  --use_lazy_mode \
  --throughput_warmup_steps 2
```

Check the [documentation](https://huggingface.co/docs/optimum/habana_index) out for more advanced usage and examples.