regisss HF staff skaulintel commited on
Commit
d63e57b
1 Parent(s): 9275f61

Enable torch autocast for albert xxlarge. (#1)

Browse files

- Enable torch autocast for albert large. (adc2b679b8ae948c8d504d51d7e2a40f5bb6ea9d)
- Remove mentions of habana mixed precision (c940e1a7c4a993f3fd1208c6c84946b8552380ae)


Co-authored-by: Shiv Kaul <skaulintel@users.noreply.huggingface.co>

Files changed (2) hide show
  1. README.md +2 -6
  2. gaudi_config.json +2 -25
README.md CHANGED
@@ -13,11 +13,7 @@ This model only contains the `GaudiConfig` file for running the [albert-xxlarge-
13
  **This model contains no model weights, only a GaudiConfig.**
14
 
15
  This enables to specify:
16
- - `use_habana_mixed_precision`: whether to use Habana Mixed Precision (HMP)
17
- - `hmp_opt_level`: optimization level for HMP, see [here](https://docs.habana.ai/en/latest/PyTorch/PyTorch_Mixed_Precision/PT_Mixed_Precision.html#configuration-options) for a detailed explanation
18
- - `hmp_bf16_ops`: list of operators that should run in bf16
19
- - `hmp_fp32_ops`: list of operators that should run in fp32
20
- - `hmp_is_verbose`: verbosity
21
  - `use_fused_adam`: whether to use Habana's custom AdamW implementation
22
  - `use_fused_clip_norm`: whether to use Habana's fused gradient norm clipping operator
23
 
@@ -45,4 +41,4 @@ python run_qa.py \
45
  --throughput_warmup_steps 2
46
  ```
47
 
48
- Check the [documentation](https://huggingface.co/docs/optimum/habana/index) out for more advanced usage and examples.
 
13
  **This model contains no model weights, only a GaudiConfig.**
14
 
15
  This enables to specify:
16
+ - `use_torch_autocast`: whether to use PyTorch's autocast mixed precision
 
 
 
 
17
  - `use_fused_adam`: whether to use Habana's custom AdamW implementation
18
  - `use_fused_clip_norm`: whether to use Habana's fused gradient norm clipping operator
19
 
 
41
  --throughput_warmup_steps 2
42
  ```
43
 
44
+ Check the [documentation](https://huggingface.co/docs/optimum/habana/index) out for more advanced usage and examples.
gaudi_config.json CHANGED
@@ -1,28 +1,5 @@
1
  {
2
- "use_habana_mixed_precision": true,
3
- "hmp_is_verbose": false,
4
  "use_fused_adam": true,
5
- "use_fused_clip_norm": true,
6
- "hmp_bf16_ops": [
7
- "add",
8
- "addmm",
9
- "bmm",
10
- "div",
11
- "dropout",
12
- "gelu",
13
- "iadd",
14
- "linear",
15
- "layer_norm",
16
- "matmul",
17
- "mm",
18
- "rsub",
19
- "softmax",
20
- "truediv"
21
- ],
22
- "hmp_fp32_ops": [
23
- "embedding",
24
- "nll_loss",
25
- "log_softmax",
26
- "cross_entropy"
27
- ]
28
  }
 
1
  {
2
+ "use_torch_autocast": true,
 
3
  "use_fused_adam": true,
4
+ "use_fused_clip_norm": true
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  }