### Configure Training Script for AllenNLP

Source: https://github.com/allenai/scibert/blob/master/README.md

Example configuration for the `train_allennlp_local.sh` script, specifying the dataset and task for training. Modify paths and variables as needed for your specific experiment.

```shell
DATASET='bc5cdr'
TASK='ner'
```

--------------------------------

### Install Dependencies with Pip

Source: https://github.com/allenai/scibert/blob/master/README.md

Installs necessary Python packages for running experiments, including AllenNLP. Ensure you are using a Python 3.6 environment.

```shell
pip install -r requirements.txt
```

--------------------------------

### Programmatically Create Instances

Source: https://context7.com/allenai/scibert/llms.txt

Demonstrates how to create a single instance using a reader's text_to_instance method. This is useful for testing or inference on individual text samples.

```python
instance = reader.text_to_instance(
    text="CRISPR-Cas9 enables targeted gene knockout in mammalian cells.",
    label="method",
    metadata={"source": "abstract"}
)
```

--------------------------------

### AllenNLP NER Configuration

Source: https://context7.com/allenai/scibert/llms.txt

Example JSON configuration file for training a BERT-CRF named entity recognition model using SciBERT embeddings.

```APIDOC
## AllenNLP NER Configuration

JSON configuration for training a BERT-CRF named entity recognition model with SciBERT embeddings.

### Method
```jsonnet
// allennlp_config/ner.json
{
  "dataset_reader": {
    "type": "conll2003",
    "tag_label": "ner",
    "coding_scheme": "BIOUL",
    "token_indexers": {
      "bert": {
        "type": "bert-pretrained",
        "pretrained_model": std.extVar("BERT_VOCAB"),
        "do_lowercase": std.extVar("IS_LOWERCASE"),
        "use_starting_offsets": true
      }
    }
  },
  "train_data_path": std.extVar("TRAIN_PATH"),
  "validation_data_path": std.extVar("DEV_PATH"),
  "test_data_path": std.extVar("TEST_PATH"),
  "evaluate_on_test": true,
  "model": {
    "type": "crf_tagger",
    "label_encoding": "BIOUL",
    "constrain_crf_decoding": true,
    "calculate_span_f1": true,
    "dropout": 0.5,
    "text_field_embedder": {
      "allow_unmatched_keys": true,
      "embedder_to_indexer_map": {
        "bert": ["bert", "bert-offsets"]
      },
      "token_embedders": {
        "bert": {
          "type": "bert-pretrained",
          "pretrained_model": std.extVar("BERT_WEIGHTS")
        }
      }
    },
    "encoder": {
      "type": "lstm",
      "input_size": 768,
      "hidden_size": 200,
      "num_layers": 2,
      "dropout": 0.5,
      "bidirectional": true
    }
  },
  "trainer": {
    "optimizer": {"type": "bert_adam", "lr": std.extVar("LEARNING_RATE")},
    "validation_metric": "+f1-measure-overall",
    "num_epochs": std.parseInt(std.extVar("NUM_EPOCHS")),
    "patience": 10,
    "cuda_device": std.parseInt(std.extVar("CUDA_DEVICE"))
  }
}
```
```