update training param and changed name

README.md

@@ -7,16 +7,16 @@ tags:
 - 
 license: agpl-3.0
 datasets:
-- EMBO/sd-nlp `PANELIZATION`
+- EMBO/sd-figures `PANELIZATION`
 metrics:
 -
 ---
 
-# sd-panels
+# sd-panelization
 
 ## Model description
 
-This model is a [RoBERTa base model](https://huggingface.co/roberta-base) that was further trained using a masked language modeling task on a compendium of english scientific textual examples from the life sciences using the [BioLang dataset](https://huggingface.co/datasets/EMBO/biolang). It was then fine-tuned for token classification on the SourceData [sd-nlp](https://huggingface.co/datasets/EMBO/sd-nlp) dataset with the `PANELIZATION` task to perform 'parsing' or 'segmentation' of figure legends into fragments corresponding to sub-panels.
+This model is a [RoBERTa base model](https://huggingface.co/roberta-base) that was further trained using a masked language modeling task on a compendium of english scientific textual examples from the life sciences using the [BioLang dataset](https://huggingface.co/datasets/EMBO/biolang). It was then fine-tuned for token classification on the SourceData [sd-figures](https://huggingface.co/datasets/EMBO/sd-figures) dataset with the `PANELIZATION` task to perform 'parsing' or 'segmentation' of figure legends into fragments corresponding to sub-panels.
 
 Figures are usually composite representations of results obtained with heterogenous experimental approaches and systems.  Breaking figures into  panels allows to identify more coherent descriptions of individual scientific experiments.
 
@@ -32,7 +32,7 @@ To have a quick check of the model:
 from transformers import pipeline, RobertaTokenizerFast, RobertaForTokenClassification
 example = """Fig 4. a, Volume density of early (Avi) and late (Avd) autophagic vacuoles.a, Volume density of early (Avi) and late (Avd) autophagic vacuoles from four independent cultures. Examples of Avi and Avd are shown in b and c, respectively. Bars represent 0.4����m. d, Labelling density of cathepsin-D as estimated in two independent experiments. e, Labelling density of LAMP-1."""
 tokenizer = RobertaTokenizerFast.from_pretrained('roberta-base', max_len=512)
-model = RobertaForTokenClassification.from_pretrained('EMBO/sd-panels')
+model = RobertaForTokenClassification.from_pretrained('EMBO/sd-panelization')
 ner = pipeline('ner', model, tokenizer=tokenizer)
 res = ner(example)
 for r in res: print(r['word'], r['entity'])
@@ -44,7 +44,7 @@ The model must be used with the `roberta-base` tokenizer.
 
 ## Training data
 
-The model was trained for token classification using the [EMBO/sd-nlp `PANELIZATION`](https://huggingface.co/datasets/EMBO/sd-nlp) dataset wich includes manually annotated examples.
+The model was trained for token classification using the [EMBO/sd-figures `PANELIZATION`](https://huggingface.co/datasets/EMBO/sd-panels) dataset wich includes manually annotated examples.
 
 ## Training procedure
 
@@ -52,15 +52,16 @@ The training was run on a NVIDIA DGX Station with 4XTesla V100 GPUs.
 
 Training code is available at https://github.com/source-data/soda-roberta
 
-- Command: `python -m tokcl.train PANELIZATION --num_train_epochs=10`
+- Model fine-tuned: EMMBO/bio-lm
 - Tokenizer vocab size: 50265
-- Training data: EMBO/sd-nlp NER
+- Training data: EMBO/sd-figures
+- Dataset configuration: PANELIZATION
 - TTraining with 2175 examples.                                  
 - Evaluating on 622 examples. 
 - Training on 2 features: `O`, `B-PANEL_START`
-- Epochs: 10.0
-- `per_device_train_batch_size`: 32
-- `per_device_eval_batch_size`: 32
+- Epochs: 1.3
+- `per_device_train_batch_size`: 16
+- `per_device_eval_batch_size`: 16
 - `learning_rate`: 0.0001
 - `weight_decay`: 0.0
 - `adam_beta1`: 0.9
@@ -70,14 +71,14 @@ Training code is available at https://github.com/source-data/soda-roberta
 
 ## Eval results
 
-Testing on 337 examples from test set with `sklearn.metrics`:
+Testing on 1802 examples from test set with `sklearn.metrics`:
 
-```
+```                                                                             
               precision    recall  f1-score   support
 
- PANEL_START       0.88      0.97      0.92       785
+ PANEL_START       0.89      0.95      0.92      5427
 
-   micro avg       0.88      0.97      0.92       785
-   macro avg       0.88      0.97      0.92       785
-weighted avg       0.88      0.97      0.92       785
+   micro avg       0.89      0.95      0.92      5427
+   macro avg       0.89      0.95      0.92      5427
+weighted avg       0.89      0.95      0.92      5427
 ```