Ivan Chelombiev
commited on
Commit
•
bf20149
1
Parent(s):
cb0616e
update model card README.md
Browse files
README.md
CHANGED
|
@@ -1,8 +1,6 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
| 4 |
-
datasets:
|
| 5 |
-
- Graphcore/wikipedia-bert-512
|
| 6 |
model-index:
|
| 7 |
- name: output-pretrain-groupbert-base-phase2
|
| 8 |
results: []
|
|
@@ -13,7 +11,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# output-pretrain-groupbert-base-phase2
|
| 15 |
|
| 16 |
-
This model was trained from scratch on the
|
| 17 |
|
| 18 |
## Model description
|
| 19 |
|
|
@@ -42,7 +40,8 @@ The following hyperparameters were used during training:
|
|
| 42 |
- total_eval_batch_size: 20
|
| 43 |
- optimizer: LAMB
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
-
-
|
|
|
|
| 46 |
- training precision: Mixed Precision
|
| 47 |
|
| 48 |
### Training results
|
|
@@ -53,5 +52,5 @@ The following hyperparameters were used during training:
|
|
| 53 |
|
| 54 |
- Transformers 4.20.1
|
| 55 |
- Pytorch 1.10.0+cpu
|
| 56 |
-
- Datasets 2.
|
| 57 |
- Tokenizers 0.12.1
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
model-index:
|
| 5 |
- name: output-pretrain-groupbert-base-phase2
|
| 6 |
results: []
|
|
|
|
| 11 |
|
| 12 |
# output-pretrain-groupbert-base-phase2
|
| 13 |
|
| 14 |
+
This model was trained from scratch on the None dataset.
|
| 15 |
|
| 16 |
## Model description
|
| 17 |
|
|
|
|
| 40 |
- total_eval_batch_size: 20
|
| 41 |
- optimizer: LAMB
|
| 42 |
- lr_scheduler_type: linear
|
| 43 |
+
- lr_scheduler_warmup_ratio: 0.15
|
| 44 |
+
- training_steps: 2038
|
| 45 |
- training precision: Mixed Precision
|
| 46 |
|
| 47 |
### Training results
|
|
|
|
| 52 |
|
| 53 |
- Transformers 4.20.1
|
| 54 |
- Pytorch 1.10.0+cpu
|
| 55 |
+
- Datasets 2.6.1
|
| 56 |
- Tokenizers 0.12.1
|