Model release

.gitattributes

@@ -25,3 +25,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zstandard filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,39 @@
+# oBERT-12-upstream-pruned-unstructured-90-finetuned-squadv1
+
+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).
+
+
+It corresponds to the model presented in the `Table 2 - oBERT - SQuADv1 90%`.
+
+```
+Pruning method: oBERT upstream unstructured + sparse-transfer to downstream
+Paper: https://arxiv.org/abs/2203.07259
+Dataset: SQuADv1
+Sparsity: 90%
+Number of layers: 12
+```
+
+The dev-set performance reported in the paper is averaged over three seeds, and we release the best model (marked with `(*)`):
+
+```
+| oBERT 90%    | F1    | EM    |
+| ------------ | ----- | ----- |
+| seed=42   (*)| 88.47 | 81.43 |
+| seed=3407    | 88.32 | 81.13 |
+| seed=54321   | 88.47 | 81.38 |
+| ------------ | ----- | ----- |
+| mean         | 88.42 | 81.31 |
+| stdev        | 0.086 | 0.160 |
+```
+
+Code: _coming soon_
+
+## BibTeX entry and citation info
+```bibtex
+@article{kurtic2022optimal,
+  title={The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models},
+  author={Kurtic, Eldar and Campos, Daniel and Nguyen, Tuan and Frantar, Elias and Kurtz, Mark and Fineran, Benjamin and Goin, Michael and Alistarh, Dan},
+  journal={arXiv preprint arXiv:2203.07259},
+  year={2022}
+}
+```

config.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c8486603b13aa568cb30a3f12f76029a3365344a234ace7353010ea02ea15338
+size 659

eval_results.txt

@@ -0,0 +1,3 @@
+exact_match = 81.42857142857143
+f1 = 88.47400893004772
+epoch = 8.0

nbest_predictions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ffea4b0a133de6aee9d6e2674f613d9d35b38c10232e3d9493a4cb8b612029f
+size 49140930

predictions.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2028b74775a687bc8ff3dc91095ef21897a910d4e1096d30791a1df44930870c
+size 588238

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a79fb12aa6f84ed9bf575d09eed599fa9f0a9357c0f9d65caf9a360a9197a9f2
+size 435661303

special_tokens_map.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:303df45a03609e4ead04bc3dc1536d0ab19b5358db685b6f3da123d05ec200e3
+size 112

tokenizer_config.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0c6801eca952ba58ada6a33f687d17655bd3a9a7ba2052f79ac9282c10a27ff6
+size 456

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e060fe9fec6d2bb79f73caf4a82e6f78a66e11fec76411cfa3c9ffbd9cc7ef5b
+size 2415