Update README.md
Browse files
README.md
CHANGED
|
@@ -60,7 +60,6 @@ tokenizer.batch_decode(outputs)
|
|
| 60 |
# Training details
|
| 61 |
* The model was loaded in 4-Bit
|
| 62 |
* The target modules include "q_proj", "k_proj", "v_proj", "o_proj"
|
| 63 |
-
* The training took about 2 hours approximately
|
| 64 |
* The fine-tuning was done on a free goggle colab with a single t4 GPU (huge thanks to unsloth for this)
|
| 65 |
* Even though the Full dataset was almost 3 million The lora model was finetuned on only 1 million row for each language
|
| 66 |
|
|
|
|
| 60 |
# Training details
|
| 61 |
* The model was loaded in 4-Bit
|
| 62 |
* The target modules include "q_proj", "k_proj", "v_proj", "o_proj"
|
|
|
|
| 63 |
* The fine-tuning was done on a free goggle colab with a single t4 GPU (huge thanks to unsloth for this)
|
| 64 |
* Even though the Full dataset was almost 3 million The lora model was finetuned on only 1 million row for each language
|
| 65 |
|