model checkpoint updated
Browse files- README.md +1 -1
- pytorch_model.bin +1 -1
README.md
CHANGED
|
@@ -49,7 +49,7 @@ The following hyperparameters were used during training:
|
|
| 49 |
- optimizer: adamw_bnb_8bit
|
| 50 |
- lr_scheduler_type: linear
|
| 51 |
- lr_scheduler_warmup_steps: 4000
|
| 52 |
-
- training_steps:
|
| 53 |
- mixed_precision_training: True
|
| 54 |
|
| 55 |
## Acknowledgement
|
|
|
|
| 49 |
- optimizer: adamw_bnb_8bit
|
| 50 |
- lr_scheduler_type: linear
|
| 51 |
- lr_scheduler_warmup_steps: 4000
|
| 52 |
+
- training_steps: 20532 (terminated upon convergence. Initially set to 21240 steps)
|
| 53 |
- mixed_precision_training: True
|
| 54 |
|
| 55 |
## Acknowledgement
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 967099139
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:456a50e26932efa33016bc674f81475fe3676d9b9d29d7f8d9e920088783ec0a
|
| 3 |
size 967099139
|