ONNX Covid 19 model

README.md

@@ -1,3 +1,20 @@
 ---
 license: mit
 ---
+This is a basic inference BERT model which has been fine-tuned to discriminate between covid19 and non-covid-19 relevant texts. 
+
+Unlike past models I have created this one raw and uploaded it as a standalone git repo to experiment with upload options. Not as streamlined as using the Huggingface card generation system but definitely simpler to do. 
+
+This is also my first experiment with ONNX.
+
+- The dataset came from Thiago Martins: https://github.com/thigm85
+
+Training data can be obtained as follows:
+```python
+import pandas as pd
+
+training_data = pd.read_csv("https://thigm85.github.io/data/cord19/cord19-query-title-label.csv")
+training_data.head()
+```
+
+Enjoy!

config.json

@@ -0,0 +1,26 @@
+{
+  "_name_or_path": "google/bert_uncased_L-4_H-512_A-8",
+  "architectures": [
+    "BertForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 512,
+  "initializer_range": 0.02,
+  "intermediate_size": 2048,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 8,
+  "num_hidden_layers": 4,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "torch_dtype": "float32",
+  "transformers_version": "4.34.0",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

model.onnx

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:304128df28249f5608aba2dfa15651ef81565803e568242811fd47a3c841fef7
+size 116639055

optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:026538e139cfbacd1f794b0198531f98129c67c516d8ca30cffc15fcc231b627
+size 10599

rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:061f4fbaefb1d4f27358e6ffb5061f83fdc82fc415eb4c8d60efe730c86df682
+size 14575

trainer_state.json

@@ -0,0 +1,125 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9182879377431905,
+  "eval_steps": 500,
+  "global_step": 7500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 0.6919,
+      "step": 500
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.6532593619972264e-05,
+      "loss": 0.6696,
+      "step": 1000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.306518723994452e-05,
+      "loss": 0.657,
+      "step": 1500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 3.959778085991678e-05,
+      "loss": 0.6521,
+      "step": 2000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.613037447988905e-05,
+      "loss": 0.6467,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.6400363445281982,
+      "eval_runtime": 33.0485,
+      "eval_samples_per_second": 311.027,
+      "eval_steps_per_second": 4.872,
+      "step": 2570
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.266296809986131e-05,
+      "loss": 0.6455,
+      "step": 3000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.9195561719833564e-05,
+      "loss": 0.6361,
+      "step": 3500
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.5728155339805826e-05,
+      "loss": 0.6448,
+      "step": 4000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.2260748959778087e-05,
+      "loss": 0.6413,
+      "step": 4500
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.879334257975035e-05,
+      "loss": 0.641,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.6325713992118835,
+      "eval_runtime": 33.21,
+      "eval_samples_per_second": 309.515,
+      "eval_steps_per_second": 4.848,
+      "step": 5140
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.5325936199722607e-05,
+      "loss": 0.64,
+      "step": 5500
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1858529819694868e-05,
+      "loss": 0.6391,
+      "step": 6000
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.39112343966713e-06,
+      "loss": 0.6442,
+      "step": 6500
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.92371705963939e-06,
+      "loss": 0.6312,
+      "step": 7000
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.4563106796116506e-06,
+      "loss": 0.6375,
+      "step": 7500
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 7710,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1186351718870016.0,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a376d2cae4c0f66886c32d18817d01dd4d18dc16f5450867a1fa144d623eb3b5
+size 4027