Hamid Shojanazeri
|
dbfea484c6
Feature : Enable Intel GPU/XPU finetuning and inference (#116)
|
1 jaar geleden |
Beto
|
1f5b202c18
Adding tests for the save_metrics param in the train function
|
1 jaar geleden |
Beto
|
7474514fe0
Merging with main
|
1 jaar geleden |
gaopengzhi
|
c7d410725b
Merge branch 'main' into grad_clip
|
1 jaar geleden |
Abhilash Majumder
|
4793f0fdf3
Merge branch 'main' into ipex_feature
|
1 jaar geleden |
gaopengzhi
|
e2797abe9b
Add gradient_clipping and gradient_clipping_threshold parameters
|
1 jaar geleden |
gaopengzhi
|
bb7c6c1e33
Support FSDP scenario
|
1 jaar geleden |
gaopengzhi
|
b1d9efd155
Refactor gradient clipping feature
|
1 jaar geleden |
Beto
|
17d02c3b44
Adding config to conditionally save stats
|
1 jaar geleden |
Beto
|
b974c87035
Merging latest from main
|
1 jaar geleden |
Jeremy Howard
|
eca8410b32
Use bf16 parameters in bf16 mixed prec
|
1 jaar geleden |
gaopengzhi
|
04befdef69
Add gradient clipping feature
|
1 jaar geleden |
Abhilash Majumder
|
11465d6329
remove duplicate import
|
1 jaar geleden |
Abhilash Majumder
|
6a78b96764
Merge branch 'main' into ipex_feature
|
1 jaar geleden |
Matthias Reso
|
e8bb7fbabc
Merge remote-tracking branch 'origin/main' into feature/length_based_batch_sampling
|
1 jaar geleden |
Matthias Reso
|
33925f71e6
Add missing amp context if use_fp16 is enabled
|
1 jaar geleden |
Beto
|
92be45b0fe
Adding matplotlib to requirements. Removing import from train_utils
|
1 jaar geleden |
Beto
|
80a6ac41b5
Adding the rank to the metrics filename to differentiate the run from each GPU
|
1 jaar geleden |
Beto
|
2f7256918e
Adding support to save fine tuning metrics to json. Adding script to plot saved metrics.
|
1 jaar geleden |
Matthias Reso
|
10f9367e56
fix missing labels in datasets
|
1 jaar geleden |
hongbo.mo
|
6217635e87
Fix tqdm bar not change length after terminal is resized
|
1 jaar geleden |
Matthias Reso
|
c33ea3cacb
Fix pbar update
|
1 jaar geleden |
hongbo.mo
|
5e910e6a42
Fix typo
|
1 jaar geleden |
hongbo.mo
|
0bc6a07a80
bugfix: update tqdm bar with the fixed gradient_accumulation_steps
|
1 jaar geleden |
abhilash1910
|
ad6b27d316
merge conflicts
|
1 jaar geleden |
abhilash1910
|
ace8b55b1b
revert gradient loss detachment
|
1 jaar geleden |
Abhilash Majumder
|
6e4455c420
fix memory bug
|
1 jaar geleden |
Abhilash Majumder
|
42a0ad1bab
fix bug on indent
|
1 jaar geleden |
abhilash1910
|
491d555c3c
Fix bugs in data loading
|
1 jaar geleden |
abhilash1910
|
33da341af5
upstream resolve conflict
|
1 jaar geleden |