pqc90 commited on
Commit
7047ea2
1 Parent(s): 2ec771b

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -23
README.md CHANGED
@@ -35,11 +35,11 @@ It takes as input four features:
35
  This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
36
 
37
  - **Developed by:** [Fundacion CTIC](https://www.fundacionctic.org)
38
- - **Funded by [optional]:** [SATEC](https://www.satec.es)
39
  - **Model type:** Fine-tuned LM Encoder
40
  - **Language(s) (NLP):** Spanish
41
  - **License:** CC-BY-NC
42
- - **Finetuned from model [optional]:** [bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es)
43
 
44
  ### Model Sources [optional]
45
 
@@ -58,24 +58,17 @@ We DO NOT recommend to fine-tune this model. It is already meant to be a downstr
58
  This model can be directly used for classifying dermatological text data in Spanish EHRs.
59
 
60
 
61
- ### Downstream Use [optional]
62
  The model can be integrated into healthcare applications for automatic classification of dermatological conditions from patient records.
63
 
64
-
65
- [More Information Needed]
66
-
67
  ### Out-of-Scope Use
68
 
69
  The model is not suitable for non-medical text classification tasks or for texts in languages other than Spanish.
70
 
71
- [More Information Needed]
72
-
73
  ## Bias, Risks, and Limitations
74
 
75
  This model is fine-tuned on a specific dataset and may not generalize well to other types of medical texts or conditions. Users should be cautious of biases in the training data that could affect the model's performance.
76
 
77
- [More Information Needed]
78
-
79
  ### Recommendations
80
 
81
  <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
@@ -108,29 +101,25 @@ outputs = model(input_ids, attention_mask=attention_mask)
108
  ### Training Data
109
 
110
  The model was fine-tuned on the DermatES dataset from Fundación CTIC, which contains Spanish dermatological EHRs.
111
- [More Information Needed]
112
 
113
  ### Training Procedure
114
 
115
  The training used two NVIDIA GPUs (11gb and 49gb)
116
 
117
- #### Preprocessing [optional]
118
 
119
  Lowercased, anonymized and accents removed texts
120
 
121
-
122
  #### Training Hyperparameters
123
 
124
  - **Training regime:** fp32
125
 
126
- #### Speeds, Sizes, Times [optional]
127
 
128
  Epochs: 9
129
  Batch size: 64
130
  Learning rate: 0.0001
131
 
132
- [More Information Needed]
133
-
134
  ## Evaluation
135
 
136
 
@@ -142,8 +131,6 @@ Lowercased, anonymized and accents removed texts
142
 
143
  The evaluation was performed on 0.2 of the [DermatES](https://huggingface.co/datasets/fundacionctic/DermatES) dataset.
144
 
145
- [More Information Needed]
146
-
147
  #### Factors
148
 
149
  <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
@@ -180,7 +167,7 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
180
  - **Compute Region:** [More Information Needed]
181
  - **Carbon Emitted:** [More Information Needed]
182
 
183
- ## Technical Specifications [optional]
184
 
185
  ### Model Architecture and Objective
186
 
@@ -193,6 +180,7 @@ The model is based on the [RoBERTa](https://huggingface.co/FacebookAI/roberta-ba
193
  #### Hardware
194
 
195
  Two NVIDIA GPUs were used for the fine-tuning process.
 
196
  #### Software
197
 
198
  The fine-tuning was performed using the 🤗 Transformers library.
@@ -203,8 +191,6 @@ The fine-tuning was performed using the 🤗 Transformers library.
203
 
204
  **BibTeX:** Coming soon
205
 
206
- [More Information Needed]
207
-
208
  **APA:**
209
 
210
  [More Information Needed]
@@ -217,10 +203,11 @@ The fine-tuning was performed using the 🤗 Transformers library.
217
 
218
  [More Information Needed]
219
 
220
- ## Model Card Authors [optional]
221
 
222
  Leon-Paul Schaub Torre, Pelayo Quiros and Helena Garcia-Mieres
223
 
224
  ## Model Card Contact
225
 
226
- leon.schaub@fundacionctic.org
 
 
35
  This is the model card of a 🤗 transformers model that has been pushed on the Hub. This model card has been automatically generated.
36
 
37
  - **Developed by:** [Fundacion CTIC](https://www.fundacionctic.org)
38
+ - **Funded by:** [SATEC](https://www.satec.es)
39
  - **Model type:** Fine-tuned LM Encoder
40
  - **Language(s) (NLP):** Spanish
41
  - **License:** CC-BY-NC
42
+ - **Finetuned from model:** [bsc-bio-ehr-es](https://huggingface.co/PlanTL-GOB-ES/bsc-bio-ehr-es)
43
 
44
  ### Model Sources [optional]
45
 
 
58
  This model can be directly used for classifying dermatological text data in Spanish EHRs.
59
 
60
 
61
+ ### Downstream Use
62
  The model can be integrated into healthcare applications for automatic classification of dermatological conditions from patient records.
63
 
 
 
 
64
  ### Out-of-Scope Use
65
 
66
  The model is not suitable for non-medical text classification tasks or for texts in languages other than Spanish.
67
 
 
 
68
  ## Bias, Risks, and Limitations
69
 
70
  This model is fine-tuned on a specific dataset and may not generalize well to other types of medical texts or conditions. Users should be cautious of biases in the training data that could affect the model's performance.
71
 
 
 
72
  ### Recommendations
73
 
74
  <!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
 
101
  ### Training Data
102
 
103
  The model was fine-tuned on the DermatES dataset from Fundación CTIC, which contains Spanish dermatological EHRs.
 
104
 
105
  ### Training Procedure
106
 
107
  The training used two NVIDIA GPUs (11gb and 49gb)
108
 
109
+ #### Preprocessing
110
 
111
  Lowercased, anonymized and accents removed texts
112
 
 
113
  #### Training Hyperparameters
114
 
115
  - **Training regime:** fp32
116
 
117
+ #### Speeds, Sizes, Times
118
 
119
  Epochs: 9
120
  Batch size: 64
121
  Learning rate: 0.0001
122
 
 
 
123
  ## Evaluation
124
 
125
 
 
131
 
132
  The evaluation was performed on 0.2 of the [DermatES](https://huggingface.co/datasets/fundacionctic/DermatES) dataset.
133
 
 
 
134
  #### Factors
135
 
136
  <!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
 
167
  - **Compute Region:** [More Information Needed]
168
  - **Carbon Emitted:** [More Information Needed]
169
 
170
+ ## Technical Specifications
171
 
172
  ### Model Architecture and Objective
173
 
 
180
  #### Hardware
181
 
182
  Two NVIDIA GPUs were used for the fine-tuning process.
183
+
184
  #### Software
185
 
186
  The fine-tuning was performed using the 🤗 Transformers library.
 
191
 
192
  **BibTeX:** Coming soon
193
 
 
 
194
  **APA:**
195
 
196
  [More Information Needed]
 
203
 
204
  [More Information Needed]
205
 
206
+ ## Model Card Authors
207
 
208
  Leon-Paul Schaub Torre, Pelayo Quiros and Helena Garcia-Mieres
209
 
210
  ## Model Card Contact
211
 
212
+ leon.schaub@fundacionctic.org
213
+ pelayo.quiros@fundacionctic.org