Update README.md

README.md

@@ -6,18 +6,18 @@ base_model: openai/whisper-medium
 tags:
 - generated_from_trainer
 datasets:
-- ivanlmh/NATI_audio_parapresentacion
+- ivanlmh/NATI_audio
 metrics:
 - wer
 model-index:
-- name: Whisper Medium Spanish-NATI - Ivan MH
+- name: Whisper Medium Spanish-NATI - Ivan MH - iidi
   results:
   - task:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
-      name: NATI_audio_parapresentacion
-      type: ivanlmh/NATI_audio_parapresentacion
+      name: NATI_audio
+      type: ivanlmh/NATI_audio
     metrics:
     - name: Wer
       type: wer
@@ -27,24 +27,27 @@ model-index:
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# Whisper Medium Spanish-NATI - Ivan MH
+# Whisper Medium Spanish-NATI - Ivan MH 
 
-This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the NATI_audio_parapresentacion dataset.
+This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the NATI_audio dataset.
 It achieves the following results on the evaluation set:
 - Loss: 2.2439
 - Wer: 66.7712
 
-## Model description
-
-More information needed
+This project is for the Instituto Interamericano para Discapacidad y Desarrollo Inclusivo (iiDi).
 
 ## Intended uses & limitations
 
-More information needed
+The model is fine-tuned for use on target speaker Natalia F.
+
+It is intended for research purposes, neither this model nor the dataset should be used without explicit permission from iiDi or Natalia.
 
 ## Training and evaluation data
 
-More information needed
+- 10 minutos aprox. de audio transcripto
+  - A partir de transcripción manual de audios de whatsapp
+- 5 minutos aprox.
+  - A partir de videos subtitulados por el iidi
 
 ## Training procedure