Reduced number fo models available and extended the project description

app.py

@@ -7,24 +7,15 @@ from transformers import AutoTokenizer, AutoModelForMaskedLM, pipeline
 LOGO = "https://raw.githubusercontent.com/nlp-en-es/assets/main/logo.png"
 
 MODELS = {
-    "RoBERTa Base": {
-        "url": "bertin-project/bertin-roberta-base-spanish"
+    "RoBERTa Base Gaussian Seq Len 512": {
+        "url": "bertin-project/bertin-base-gaussian-exp-512seqlen"
     },
-    "RoBERTa Base Gaussian": {
+    "RoBERTa Base Gaussian Seq Len 128": {
         "url": "bertin-project/bertin-base-gaussian"
     },
-    "RoBERTa Base Random": {
+    "RoBERTa Base Random Seq Len 128": {
         "url": "bertin-project/bertin-base-random"
     },
-    "RoBERTa Base Stepwise": {
-        "url": "bertin-project/bertin-base-stepwise"
-    },
-    "RoBERTa Base Gaussian Experiment": {
-        "url": "bertin-project/bertin-base-gaussian-exp-512seqlen"
-    },
-    "RoBERTa Base Random Experiment": {
-        "url": "bertin-project/bertin-base-random-exp-512seqlen"
-    }
 }
 
 PROMPT_LIST = [
@@ -36,7 +27,9 @@ PROMPT_LIST = [
     "Mañana vienen mis amigos de <mask>.",
     "¿Te apetece venir a <mask> conmigo?",
     "En verano hace mucho <mask>.",
-    "En el bosque había <mask>."
+    "En el bosque había <mask>.",
+    "El ministro dijo que <mask> los impuestos.",
+    "Si no estuviera afónica, <mask> esa canción.",
 ]
 
 
@@ -63,6 +56,12 @@ st.markdown(
     The models are trained with Flax and using TPUs sponsored by Google since this is part of the
     [Flax/Jax Community Week](https://discuss.huggingface.co/t/open-to-the-community-community-week-using-jax-flax-for-nlp-cv/7104)
     organised by HuggingFace.
+    
+    All models are variations of RoBERTa-base trained from scratch in Spanish. 
+    We used the mc4 dataset. We reduced the dataset size to 50 million documents to keep training times shorter, and also to be able to bias training examples based on their perplexity.
+    The idea is to favour examples with perplexities that are neither too small (short, repetitive texts) or too long (potentially poor quality).
+    "Random" sampling simply takes documents at random to reduce the dataset size. "Gaussian" rejects documents with a higher probability for lower and larger perplexities, based on a Gaussian function.
+    The first models have been trained (250.000 steps) on sequence length 128, and training for Gaussian changed to sequence length 512 for the last 25.000 training steps.
     """
 )
 
@@ -88,12 +87,12 @@ if st.button("Fill the mask"):
 st.markdown(
     """
     ### Team members
-    - Javier de la Rosa ([versae](https://huggingface.co/versae))
     - Eduardo González ([edugp](https://huggingface.co/edugp))
-    - Paulo Villegas ([paulo](https://huggingface.co/paulo))
-    - Pablo González de Prado ([Pablogps](https://huggingface.co/Pablogps))
+    - Javier de la Rosa ([versae](https://huggingface.co/versae))
     - Manu Romero ([mrm8488](https://huggingface.co/mrm8488))
     - María Grandury ([mariagrandury](https://huggingface.co/mariagrandury))
+    - Pablo González de Prado ([Pablogps](https://huggingface.co/Pablogps))
+    - Paulo Villegas ([paulo](https://huggingface.co/paulo))
         
     ### More information
     You can find more information about these models

requirements.txt

@@ -1,4 +1,4 @@
 streamlit
 mtranslate
 transformers
-torch
+torch