SGaleshchuk
/

Llama-2-13b-hf_uk_rank-32_ft

Text Generation

text-generation-inference

Model card Files Files and versions Community

SGaleshchuk commited on Apr 6

Commit

335e816

•

1 Parent(s): f36b052

Update README.md

Files changed (1) hide show

README.md +7 -2

README.md CHANGED Viewed

@@ -12,5 +12,10 @@ pipeline_tag: text-generation
 ---
 The following training arguments used for Llama-2 finetuning with Ukrainian corpora pf XL-SUM:
-learning-rate=2e-4,warm-up ratio = 0.03, maximum number of tokens =512, truncate otherwise, 5 epochs. Lora perf arguments:
-rank = 32, lora-alpha=16, dropout = 0.1.

 ---
 The following training arguments used for Llama-2 finetuning with Ukrainian corpora pf XL-SUM:
+- learning-rate=2e-4,
+- maximum number of tokens=512,
+- 5 epochs.
+  Lora perf arguments:
+- rank = 32,
+- lora-alpha=16,
+- dropout = 0.1.