Deploy TinyModel1 from GitHub Actions

README.md

@@ -0,0 +1,38 @@
+---
+license: apache-2.0
+library_name: transformers
+datasets:
+  - ag_news
+tags:
+  - tiny
+  - text-classification
+  - ag-news
+---
+
+# TinyModel1
+
+TinyModel1 is a lightweight news-topic classifier trained from scratch on `ag_news`.
+
+## Task
+
+Input: short news text  
+Output labels: World, Sports, Business, Sci/Tech
+
+## Training setup
+
+- Base model: tiny BERT from scratch
+- Train samples: 3000
+- Eval samples: 600
+- Epochs: 2
+- Batch size: 16
+- Learning rate: 0.0001
+
+## Quick metrics
+
+- Eval accuracy: 0.5333
+- Final train loss: 1.1577
+
+## Intended use
+
+- Fast baseline for category routing/classification
+- Starter model for domain adaptation and production experiments

artifact.json

@@ -0,0 +1,15 @@
+{
+  "name": "TinyModel1",
+  "version": "0.2.0",
+  "task": "text-classification",
+  "dataset": "ag_news",
+  "base_model": "tinymodel1-bert-scratch",
+  "labels": [
+    "World",
+    "Sports",
+    "Business",
+    "Sci/Tech"
+  ],
+  "eval_accuracy": 0.5333,
+  "train_loss": 1.1577
+}

config.json

@@ -0,0 +1,41 @@
+{
+  "add_cross_attention": false,
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": null,
+  "classifier_dropout": null,
+  "dtype": "float32",
+  "eos_token_id": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 128,
+  "id2label": {
+    "0": "World",
+    "1": "Sports",
+    "2": "Business",
+    "3": "Sci/Tech"
+  },
+  "initializer_range": 0.02,
+  "intermediate_size": 256,
+  "is_decoder": false,
+  "label2id": {
+    "Business": 2,
+    "Sci/Tech": 3,
+    "Sports": 1,
+    "World": 0
+  },
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 256,
+  "model_type": "bert",
+  "num_attention_heads": 4,
+  "num_hidden_layers": 2,
+  "pad_token_id": 0,
+  "problem_type": "single_label_classification",
+  "tie_word_embeddings": true,
+  "transformers_version": "5.5.4",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 8000
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ae0850b9449add792b6f2a310f04da54c51f41445970f6342fcffc4475893f4
+size 5361600

tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "backend": "tokenizers",
+  "cls_token": "[CLS]",
+  "do_lower_case": true,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "BertTokenizer",
+  "unk_token": "[UNK]"
+}