Hugging Face's logo

Spaces:
hf-accelerate
/
accelerate_examples
Running on CPU Upgrade

App Files Community
4
muellerzr HF staff commited on
Commit
1cbe8ed
•
1 Parent(s): 8d6d31e

Go to blank

Browse files
Files changed (5) hide show
  1. code_samples/basic +2 -2
  2. code_samples/calculating_metrics +3 -2
  3. code_samples/checkpointing +3 -3
  4. code_samples/experiment_tracking +3 -3
  5. code_samples/gradient_accumulation +3 -3
code_samples/basic CHANGED
@@ -27,5 +27,5 @@ is done automatically by `accelerator.prepare()`.
27
 
28
  ##
29
  To learn more checkout the related documentation:
30
- - [Basic Tutorial](https://huggingface.co/docs/accelerate/basic_tutorials/migration)
31
- - [The Accelerator](https://huggingface.co/docs/accelerate/package_reference/accelerator)
 
27
 
28
  ##
29
  To learn more checkout the related documentation:
30
+ - <a href="https://huggingface.co/docs/accelerate/basic_tutorials/migration" target="_blank">Basic Tutorial</a>
31
+ - <a href="https://huggingface.co/docs/accelerate/package_reference/accelerator" target="_blank">The Accelerator</a>
code_samples/calculating_metrics CHANGED
@@ -45,5 +45,6 @@ This will also *automatically* drop the padded values from the gathered tensors
45
  that all tensors have the same length. This ensures that the metric is calculated on the correct values.
46
  ##
47
  To learn more checkout the related documentation:
48
- - [API reference](https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.gather_for_metrics)
49
- - [Example script](https://github.com/huggingface/accelerate/blob/main/examples/by_feature/multi_process_metrics.py)
 
 
45
  that all tensors have the same length. This ensures that the metric is calculated on the correct values.
46
  ##
47
  To learn more checkout the related documentation:
48
+
49
+ - <a href="https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.gather_for_metrics" target="_blank">API reference</a>
50
+ - <a href="https://github.com/huggingface/accelerate/blob/main/examples/by_feature/multi_process_metrics.py" target="_blank">Example script</a>
code_samples/checkpointing CHANGED
@@ -23,6 +23,6 @@ any custom registered objects from the main process on each device to a passed i
23
  **This API is designed to save and resume training states only from within the same python script or training setup.**
24
  ##
25
  To learn more checkout the related documentation:
26
- - [`save_state` reference](https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.save_state)
27
- - [`load_state` reference](https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.load_state)
28
- - [Example script](https://github.com/huggingface/accelerate/blob/main/examples/by_feature/checkpointing.py)
 
23
  **This API is designed to save and resume training states only from within the same python script or training setup.**
24
  ##
25
  To learn more checkout the related documentation:
26
+ - <a href="https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.save_state" target="_blank">`save_state` reference</a>
27
+ - <a href="https://huggingface.co/docs/accelerate/v0.15.0/package_reference/accelerator#accelerate.Accelerator.load_state" target="_blank">`load_state` reference</a>
28
+ - <a href="https://github.com/huggingface/accelerate/blob/main/examples/by_feature/checkpointing.py" target="_blank">Example script</a>
code_samples/experiment_tracking CHANGED
@@ -27,6 +27,6 @@ At the end of training call `accelerator.end_training()` to call any finalizatio
27
  may need automatically.
28
  ##
29
  To learn more checkout the related documentation:
30
- - [Basic Tutorial](https://huggingface.co/docs/accelerate/usage_guides/tracking)
31
- - [Accelerator API Reference](https://huggingface.co/docs/accelerate/package_reference/accelerator#accelerate.Accelerator.log)
32
- - [Tracking API Reference](https://huggingface.co/docs/accelerate/package_reference/tracking)
 
27
  may need automatically.
28
  ##
29
  To learn more checkout the related documentation:
30
+ - <a href="https://huggingface.co/docs/accelerate/usage_guides/tracking" target="_blank">Basic Tutorial</a>
31
+ - <a href="https://huggingface.co/docs/accelerate/package_reference/accelerator#accelerate.Accelerator.log" target="_blank">Accelerator API Reference</a>
32
+ - <a href="https://huggingface.co/docs/accelerate/package_reference/tracking" target="_blank">Tracking API Reference</a>
code_samples/gradient_accumulation CHANGED
@@ -27,6 +27,6 @@ automatically when needed.
27
 
28
  ##
29
  To learn more checkout the related documentation:
30
- - [API reference](https://huggingface.co/docs/accelerate/package_reference/accelerator#accelerate.Accelerator.accumulate)
31
- - [Example script](https://github.com/huggingface/accelerate/blob/main/examples/by_feature/gradient_accumulation.py)
32
- - [Performing automatic gradient accumulation](https://github.com/huggingface/accelerate/blob/main/examples/by_feature/automatic_gradient_accumulation.py)
 
27
 
28
  ##
29
  To learn more checkout the related documentation:
30
+ - <a href="https://huggingface.co/docs/accelerate/package_reference/accelerator#accelerate.Accelerator.accumulate" target="_blank">API reference</a>
31
+ - <a href="https://github.com/huggingface/accelerate/blob/main/examples/by_feature/gradient_accumulation.py" target="_blank">Example script</a>
32
+ - <a href="https://github.com/huggingface/accelerate/blob/main/examples/by_feature/automatic_gradient_accumulation.py" target="_blank">Performing automatic gradient accumulation</a>