We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
1 parent c89b231 commit 22a5e98Copy full SHA for 22a5e98
tests/baselines/gemma_2b_it.json
@@ -25,6 +25,18 @@
25
"--dataset_config_name wikitext-2-raw-v1",
26
"--use_hpu_graphs_for_inference"
27
]
28
+ },
29
+ "deepspeed": {
30
+ "learning_rate": 8e-4,
31
+ "train_batch_size": 4,
32
+ "perplexity": 924.062,
33
+ "train_runtime": 75.518,
34
+ "train_samples_per_second": 81.097,
35
+ "extra_arguments": [
36
+ "--dataset_config_name wikitext-2-raw-v1",
37
+ "--use_hpu_graphs_for_inference",
38
+ "--deepspeed tests/configs/deepspeed_zero_2.json"
39
+ ]
40
}
41
42
0 commit comments