Model release
Browse files- .gitattributes +1 -0
- README.md +40 -0
- all_results.json +3 -0
- config.json +3 -0
- eval_nbest_predictions.json +3 -0
- eval_predictions.json +3 -0
- eval_results.json +3 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +3 -0
- tokenizer.json +3 -0
- tokenizer_config.json +3 -0
- train_results.json +3 -0
- trainer_state.json +3 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
.gitattributes
CHANGED
@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
25 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
25 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
26 |
*.zstandard filter=lfs diff=lfs merge=lfs -text
|
27 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
28 |
+
*.json filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1-v2
|
2 |
+
|
3 |
+
This model is obtained with [The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models](https://arxiv.org/abs/2203.07259).
|
4 |
+
|
5 |
+
|
6 |
+
It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 90%` (in the upcoming updated version of the paper).
|
7 |
+
|
8 |
+
```
|
9 |
+
Pruning method: oBERT upstream unstructured + sparse-transfer to downstream
|
10 |
+
Paper: https://arxiv.org/abs/2203.07259
|
11 |
+
Dataset: SQuADv1
|
12 |
+
Sparsity: 90%
|
13 |
+
Number of layers: 12
|
14 |
+
```
|
15 |
+
|
16 |
+
The dev-set performance reported in the paper is averaged over four seeds, and we release the best model (marked with `(*)`):
|
17 |
+
|
18 |
+
```
|
19 |
+
| oBERT 90% | F1 | EM |
|
20 |
+
| ------------ | ----- | ----- |
|
21 |
+
| seed=42 | 88.55 | 81.48 |
|
22 |
+
| seed=3407 | 88.34 | 81.25 |
|
23 |
+
| seed=123 (*)| 88.64 | 81.57 |
|
24 |
+
| seed=12345 | 88.44 | 81.43 |
|
25 |
+
| ------------ | ----- | ----- |
|
26 |
+
| mean | 88.49 | 81.43 |
|
27 |
+
| stdev | 0.130 | 0.134 |
|
28 |
+
```
|
29 |
+
|
30 |
+
Code: _coming soon_
|
31 |
+
|
32 |
+
## BibTeX entry and citation info
|
33 |
+
```bibtex
|
34 |
+
@article{kurtic2022optimal,
|
35 |
+
title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models},
|
36 |
+
author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan},
|
37 |
+
journal={arXiv preprint arXiv:2203.07259},
|
38 |
+
year={2022}
|
39 |
+
}
|
40 |
+
```
|
all_results.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1084b776556b7ff8013625ffb361458a0b72ad0a1dfa8ba93b988d00a98d46dd
|
3 |
+
size 298
|
config.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3274af34bf9376a6d05b260700b91a3c297dc3d2c36cdc0f48625c7544c5075a
|
3 |
+
size 773
|
eval_nbest_predictions.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:be5c804c18fdee5c004e4a99dffbde768fc5264cf643f9fb6c58a80b7128ee5f
|
3 |
+
size 45694273
|
eval_predictions.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2c07529741c541e6097e94146c4ea6a935f644d8d0683c78d74de479cdcdbc79
|
3 |
+
size 586538
|
eval_results.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cb49cdc4d4fefe04c289bcfbe44c46278edd1a280b59a98049e7a2c8dceb55b7
|
3 |
+
size 124
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cce02f9c8ac07308ca1a5d2ed4bba77bff9e41be27e986185970e046936da59a
|
3 |
+
size 435658993
|
special_tokens_map.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:303df45a03609e4ead04bc3dc1536d0ab19b5358db685b6f3da123d05ec200e3
|
3 |
+
size 112
|
tokenizer.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d241a60d5e8f04cc1b2b3e9ef7a4921b27bf526d9f6050ab90f9267a1f9e5c66
|
3 |
+
size 711396
|
tokenizer_config.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a2995bd0620bd945c6fedbfb8ec229865428cb1fb08d690e79e3d7715755af2e
|
3 |
+
size 441
|
train_results.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c6561cef679fa038a4076c61cd9eeef5a460f830d460e219fbb0ef9ba637efa8
|
3 |
+
size 194
|
trainer_state.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4aa82113c16c3d7e92d24f3c0c90b0e3ccb21e9a6cb1679b9e0c55f0ceb19568
|
3 |
+
size 12408
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d995d5779a5da16ba9516958193ba6446f645fd5e57afc148a8fb625f8daacf4
|
3 |
+
size 3375
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|