amirbachar
commited on
Commit
•
6ec1050
1
Parent(s):
4549f19
End of training
Browse files
README.md
CHANGED
@@ -16,13 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
16 |
|
17 |
This model is a fine-tuned version of [google/flan-t5-xl](https://huggingface.co/google/flan-t5-xl) on an unknown dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
-
-
|
20 |
-
-
|
21 |
-
- eval_runtime: 1938.7518
|
22 |
-
- eval_samples_per_second: 0.533
|
23 |
-
- eval_steps_per_second: 0.067
|
24 |
-
- epoch: 13.0
|
25 |
-
- step: 85215
|
26 |
|
27 |
## Model description
|
28 |
|
@@ -49,6 +44,27 @@ The following hyperparameters were used during training:
|
|
49 |
- lr_scheduler_type: linear
|
50 |
- num_epochs: 15
|
51 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
52 |
### Framework versions
|
53 |
|
54 |
- PEFT 0.9.0
|
|
|
16 |
|
17 |
This model is a fine-tuned version of [google/flan-t5-xl](https://huggingface.co/google/flan-t5-xl) on an unknown dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.0936
|
20 |
+
- Execution Accuracy: 70.5029
|
|
|
|
|
|
|
|
|
|
|
21 |
|
22 |
## Model description
|
23 |
|
|
|
44 |
- lr_scheduler_type: linear
|
45 |
- num_epochs: 15
|
46 |
|
47 |
+
### Training results
|
48 |
+
|
49 |
+
| Training Loss | Epoch | Step | Execution Accuracy | Validation Loss |
|
50 |
+
|:-------------:|:-----:|:-----:|:------------------:|:---------------:|
|
51 |
+
| 0.068 | 1.0 | 6555 | 39.4584 | 0.0767 |
|
52 |
+
| 0.0432 | 2.0 | 13110 | 52.9981 | 0.0608 |
|
53 |
+
| 0.033 | 3.0 | 19665 | 60.3482 | 0.0612 |
|
54 |
+
| 0.0297 | 4.0 | 26220 | 62.8627 | 0.0589 |
|
55 |
+
| 0.0213 | 5.0 | 32775 | 64.1199 | 0.0605 |
|
56 |
+
| 0.0188 | 6.0 | 39330 | 64.3133 | 0.0619 |
|
57 |
+
| 0.0166 | 7.0 | 45885 | 66.441 | 0.0611 |
|
58 |
+
| 0.0162 | 8.0 | 52440 | 65.8607 | 0.0669 |
|
59 |
+
| 0.0109 | 9.0 | 58995 | 68.9555 | 0.0666 |
|
60 |
+
| 0.0101 | 10.0 | 65550 | 68.1818 | 0.0736 |
|
61 |
+
| 0.0085 | 11.0 | 72105 | 68.0851 | 0.0764 |
|
62 |
+
| 0.0069 | 12.0 | 78660 | 69.0522 | 0.0801 |
|
63 |
+
| 0.0068 | 13.0 | 85215 | 69.2456 | 0.0884 |
|
64 |
+
| 0.0052 | 14.0 | 91770 | 0.0883 | 70.793 |
|
65 |
+
| 0.0039 | 15.0 | 98325 | 0.0936 | 70.5029 |
|
66 |
+
|
67 |
+
|
68 |
### Framework versions
|
69 |
|
70 |
- PEFT 0.9.0
|