update model card README.md
Browse files
README.md
CHANGED
|
@@ -2,8 +2,6 @@
|
|
| 2 |
license: cc-by-nc-4.0
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
| 5 |
-
metrics:
|
| 6 |
-
- accuracy
|
| 7 |
model-index:
|
| 8 |
- name: videomae-base-ipm_first_videos
|
| 9 |
results: []
|
|
@@ -16,8 +14,13 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
| 18 |
It achieves the following results on the evaluation set:
|
| 19 |
-
-
|
| 20 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
## Model description
|
| 23 |
|
|
@@ -43,28 +46,7 @@ The following hyperparameters were used during training:
|
|
| 43 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
- lr_scheduler_type: linear
|
| 45 |
- lr_scheduler_warmup_ratio: 0.1
|
| 46 |
-
- training_steps:
|
| 47 |
-
|
| 48 |
-
### Training results
|
| 49 |
-
|
| 50 |
-
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|
| 51 |
-
|:-------------:|:-----:|:----:|:---------------:|:--------:|
|
| 52 |
-
| No log | 0.07 | 3 | 1.7697 | 0.2286 |
|
| 53 |
-
| No log | 1.07 | 6 | 1.8755 | 0.1143 |
|
| 54 |
-
| No log | 2.07 | 9 | 1.9833 | 0.1143 |
|
| 55 |
-
| 1.7083 | 3.07 | 12 | 2.0962 | 0.1143 |
|
| 56 |
-
| 1.7083 | 4.07 | 15 | 2.1203 | 0.1143 |
|
| 57 |
-
| 1.7083 | 5.07 | 18 | 2.1553 | 0.1143 |
|
| 58 |
-
| 1.6139 | 6.07 | 21 | 2.1567 | 0.1143 |
|
| 59 |
-
| 1.6139 | 7.07 | 24 | 2.1460 | 0.1143 |
|
| 60 |
-
| 1.6139 | 8.07 | 27 | 2.1409 | 0.1143 |
|
| 61 |
-
| 1.6933 | 9.07 | 30 | 2.1007 | 0.1143 |
|
| 62 |
-
| 1.6933 | 10.07 | 33 | 2.0846 | 0.1429 |
|
| 63 |
-
| 1.6933 | 11.07 | 36 | 2.0487 | 0.1429 |
|
| 64 |
-
| 1.6933 | 12.07 | 39 | 2.0399 | 0.1429 |
|
| 65 |
-
| 1.6804 | 13.07 | 42 | 2.0327 | 0.1714 |
|
| 66 |
-
| 1.6804 | 14.05 | 44 | 2.0331 | 0.1714 |
|
| 67 |
-
|
| 68 |
|
| 69 |
### Framework versions
|
| 70 |
|
|
|
|
| 2 |
license: cc-by-nc-4.0
|
| 3 |
tags:
|
| 4 |
- generated_from_trainer
|
|
|
|
|
|
|
| 5 |
model-index:
|
| 6 |
- name: videomae-base-ipm_first_videos
|
| 7 |
results: []
|
|
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
+
- eval_loss: 0.6246
|
| 18 |
+
- eval_accuracy: 0.7895
|
| 19 |
+
- eval_runtime: 50.9021
|
| 20 |
+
- eval_samples_per_second: 0.747
|
| 21 |
+
- eval_steps_per_second: 0.196
|
| 22 |
+
- epoch: 48.01
|
| 23 |
+
- step: 146
|
| 24 |
|
| 25 |
## Model description
|
| 26 |
|
|
|
|
| 46 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 47 |
- lr_scheduler_type: linear
|
| 48 |
- lr_scheduler_warmup_ratio: 0.1
|
| 49 |
+
- training_steps: 330
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 50 |
|
| 51 |
### Framework versions
|
| 52 |
|