Update README.md

README.md

@@ -1,72 +1,22 @@
 ---
-license: mit
-base_model: microsoft/deberta-v3-large
+language:
+- en
 tags:
-- generated_from_trainer
-metrics:
-- accuracy
-model-index:
-- name: deberta-v3-large-zeroshot-v1.1-all_except_nli
-  results: []
+- text-classification
+- zero-shot-classification
+pipeline_tag: zero-shot-classification
+library_name: transformers
+license: mit
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
-
-# deberta-v3-large-zeroshot-v1.1-all_except_nli
-
-This model is a fine-tuned version of [microsoft/deberta-v3-large](https://huggingface.co/microsoft/deberta-v3-large) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.3183
-- F1 Macro: 0.2132
-- F1 Micro: 0.2379
-- Accuracy Balanced: 0.2319
-- Accuracy: 0.2379
-- Precision Macro: 0.4070
-- Recall Macro: 0.2319
-- Precision Micro: 0.2379
-- Recall Micro: 0.2379
-
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
-
-More information needed
-
-## Training procedure
-
-### Training hyperparameters
-
-The following hyperparameters were used during training:
-- learning_rate: 9e-06
-- train_batch_size: 16
-- eval_batch_size: 64
-- seed: 42
-- gradient_accumulation_steps: 2
-- total_train_batch_size: 32
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_ratio: 0.06
-- num_epochs: 3
-
-### Training results
+# Model description:  deberta-v3-large-mnli-fever-anli-ling-wanli-binary
 
-| Training Loss | Epoch | Step  | Validation Loss | F1 Macro | F1 Micro | Accuracy Balanced | Accuracy | Precision Macro | Recall Macro | Precision Micro | Recall Micro |
-|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:|:-----------------:|:--------:|:---------------:|:------------:|:---------------:|:------------:|
-| 0.1929        | 1.0   | 27664 | 0.3072          | 0.8708   | 0.8831   | 0.8683            | 0.8831   | 0.8735          | 0.8683       | 0.8831          | 0.8831       |
-| 0.1426        | 2.0   | 55328 | 0.3692          | 0.8709   | 0.8839   | 0.8664            | 0.8839   | 0.8761          | 0.8664       | 0.8839          | 0.8839       |
-| 0.0935        | 3.0   | 82992 | 0.4419          | 0.8747   | 0.8864   | 0.8729            | 0.8864   | 0.8765          | 0.8729       | 0.8864          | 0.8864       |
+This model was mostly created as a comparative benchmark for another model, see here: https://huggingface.co/MoritzLaurer/deberta-v3-large-zeroshot-v1.1-all-33
 
+This model was only trained on five NLI datasets, while the other model was trained on many more datasets.
+I mostly recommend using the other model. 
+This NLI-only model might only be better for tasks that are not zeroshot classification 
+but that adhere more strictly to the original NLI task. 
 
-### Framework versions
+See the other model's model card for usage instructions, training data and the paper.
 
-- Transformers 4.33.3
-- Pytorch 1.11.0+cu113
-- Datasets 2.14.6
-- Tokenizers 0.12.1