Upload README.md with huggingface_hub

README.md

@@ -1,68 +1,70 @@
 ---
-library_name: transformers
-license: mit
-base_model: prajjwal1/bert-medium
+language: en
+license: apache-2.0
 tags:
-- generated_from_trainer
+- text-classification
+- zero-shot-classification
+- nli
+datasets:
+- Pankaj8922/nli-high-quality-balanced
+base_model: prajjwal1/bert-medium
 metrics:
 - accuracy
+- f1
 model-index:
 - name: bert-medium-nli
-  results: []
+  results:
+  - task:
+      type: text-classification
+      name: Natural Language Inference
+    dataset:
+      name: Pankaj8922/nli-high-quality-balanced
+      type: Pankaj8922/nli-high-quality-balanced
+    metrics:
+    - type: accuracy
+      value: 0.8373
+      name: Test Accuracy
+    - type: f1
+      value: 0.8372
+      name: Test F1 (macro)
 ---
 
-<!-- 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. -->
-
 # bert-medium-nli
 
-This model is a fine-tuned version of [prajjwal1/bert-medium](https://huggingface.co/prajjwal1/bert-medium) on an unknown dataset.
-It achieves the following results on the evaluation set:
-- Loss: 0.4368
-- Accuracy: 0.8389
-- F1 Macro: 0.8389
-- Precision Macro: 0.8389
-- Recall Macro: 0.8389
-
-## Model description
-
-More information needed
-
-## Intended uses & limitations
-
-More information needed
-
-## Training and evaluation data
+Fine-tuned [`prajjwal1/bert-medium`](https://huggingface.co/prajjwal1/bert-medium) for natural language
+inference (entailment / neutral / contradiction), intended for use as a zero-shot
+text classification model via the entailment trick (hypothesis = "This text is
+about {label}.").
 
-More information needed
+Trained on [`Pankaj8922/nli-high-quality-balanced`](https://huggingface.co/datasets/Pankaj8922/nli-high-quality-balanced), a
+combined and filtered subset of MNLI, SNLI, FEVER-NLI, and ANLI:
+annotator-agreement filtered, deduplicated, teacher-confidence filtered,
+hypothesis-only artifact filtered, and class-balanced.
 
-## Training procedure
+## Results
 
-### Training hyperparameters
+| Split | Accuracy | F1 (macro) | Precision (macro) | Recall (macro) |
+|---|---|---|---|---|
+| Validation | 0.8389 | 0.8389 | 0.8389 | 0.8389 |
+| Test | 0.8373 | 0.8372 | 0.8372 | 0.8373 |
 
-The following hyperparameters were used during training:
-- learning_rate: 5e-05
-- train_batch_size: 64
-- eval_batch_size: 128
-- seed: 42
-- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: linear
-- lr_scheduler_warmup_steps: 0.06
-- num_epochs: 3
-- mixed_precision_training: Native AMP
+## Training details
 
-### Training results
+- Base model: `prajjwal1/bert-medium`
+- Epochs: 3
+- Batch size: 64 (train), 128 (eval)
+- Learning rate: 5e-05
+- Max sequence length: 256
 
-| Training Loss | Epoch | Step  | Validation Loss | Accuracy | F1 Macro | Precision Macro | Recall Macro |
-|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:|:---------------:|:------------:|
-| 0.5086        | 1.0   | 9979  | 0.4701          | 0.8077   | 0.8065   | 0.8122          | 0.8077       |
-| 0.3832        | 2.0   | 19958 | 0.4225          | 0.8324   | 0.8323   | 0.8329          | 0.8324       |
-| 0.2974        | 3.0   | 29937 | 0.4368          | 0.8389   | 0.8389   | 0.8389          | 0.8389       |
+## Labels
 
+- 0: entailment
+- 1: neutral
+- 2: contradiction
 
-### Framework versions
+## Intended use / limitations
 
-- Transformers 5.0.0
-- Pytorch 2.10.0+cu128
-- Datasets 5.0.0
-- Tokenizers 0.22.2
+This is a small (~41M parameter) model, so its ceiling on zero-shot performance
+against novel, unseen label sets is lower than larger NLI-tuned checkpoints
+(e.g. DeBERTa-v3-base or -large variants). Best suited for fast inference or
+resource-constrained settings rather than maximum accuracy.