Training in progress, step 50

README.md

@@ -1,35 +1,90 @@
 ---
-base_model: openai/whisper-small
-datasets:
-- mozilla-foundation/common_voice_17_0
-language: sw
 library_name: transformers
 license: apache-2.0
+base_model: openai/whisper-small
+tags:
+- generated_from_trainer
+datasets:
+- common_voice_17_0
+metrics:
+- wer
 model-index:
-- name: Finetuned openai/whisper-small on Swahili
+- name: ASR-Swahili-Small
   results:
   - task:
+      name: Automatic Speech Recognition
       type: automatic-speech-recognition
-      name: Speech-to-Text
     dataset:
-      name: Common Voice (Swahili)
-      type: common_voice
+      name: common_voice_17_0
+      type: common_voice_17_0
+      config: sw
+      split: test
+      args: sw
     metrics:
-    - type: wer
-      value: 79.082
+    - name: Wer
+      type: wer
+      value: 79.08239916791864
 ---
 
-# Finetuned openai/whisper-small on 2000 Swahili training audio samples from mozilla-foundation/common_voice_17_0.
+<!-- 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. -->
+
+# ASR-Swahili-Small
+
+This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the common_voice_17_0 dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.4765
+- Model Preparation Time: 0.003
+- Wer: 79.0824
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 1e-05
+- train_batch_size: 4
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 16
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: linear
+- lr_scheduler_warmup_steps: 9
+- training_steps: 50
+- mixed_precision_training: Native AMP
+
+### Training results
 
-This model was created from the Mozilla.ai Blueprint:
-[speech-to-text-finetune](https://github.com/mozilla-ai/speech-to-text-finetune).
+| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer      |
+|:-------------:|:-----:|:----:|:---------------:|:----------------------:|:--------:|
+| 2.4744        | 0.04  | 5    | 2.4647          | 0.003                  | 133.2948 |
+| 2.3612        | 0.08  | 10   | 2.1132          | 0.003                  | 116.4914 |
+| 1.9312        | 0.12  | 15   | 1.8395          | 0.003                  | 90.6969  |
+| 1.5686        | 0.16  | 20   | 1.6659          | 0.003                  | 87.3628  |
+| 1.5144        | 0.2   | 25   | 1.5895          | 0.003                  | 82.8730  |
+| 1.4267        | 0.24  | 30   | 1.5432          | 0.003                  | 84.9070  |
+| 1.492         | 0.28  | 35   | 1.5229          | 0.003                  | 84.5718  |
+| 1.3699        | 0.32  | 40   | 1.4999          | 0.003                  | 82.3009  |
+| 1.2679        | 0.36  | 45   | 1.4842          | 0.003                  | 80.5848  |
+| 1.3613        | 0.4   | 50   | 1.4765          | 0.003                  | 79.0824  |
 
-## Evaluation results on 2000 audio samples of Swahili:
 
-### Baseline model (before finetuning) on Swahili
-- Word Error Rate: 139.455
-- Loss: 2.576
+### Framework versions
 
-### Finetuned model (after finetuning) on Swahili
-- Word Error Rate: 79.082
-- Loss: 1.477
+- Transformers 4.49.0
+- Pytorch 2.6.0+cu124
+- Datasets 3.3.1
+- Tokenizers 0.21.0

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d191909ce6fa88fa3008a69517b2c3055bfcdcf59a562aef101356f264f707b
+oid sha256:ec88b11a1d1ac7410edff5145bb7d994b2ead4a42cadc040d2d763bd59532f89
 size 966995080

runs/Feb27_01-47-15_ai4d-Lambda-Vector/events.out.tfevents.1740610489.ai4d-Lambda-Vector.1052975.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:07e6730f5589180413897b987a45b6440f072f1e91e95e2c34d1cd6ce9c9f73c
+size 7967

runs/Feb27_02-15-13_ai4d-Lambda-Vector/events.out.tfevents.1740613905.ai4d-Lambda-Vector.1113201.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d641afed01a135bbee18bfb404f24eb8aeb89d468c5693a2b7dcab914248b1b
+size 7384

runs/Feb27_03-09-55_ai4d-Lambda-Vector/events.out.tfevents.1740617299.ai4d-Lambda-Vector.1238951.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1cf5ba34a45efc696e171dac424a38e6690dde6c4ca114a819dd4207e70e2435
+size 7765

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70b0397ab7f149ad8d901fad9dc0d786c7b921134d84a5f3fadfb2168544b5f6
+oid sha256:48f921cced55cf902a140bcbd2244853c66a2b1fd78f76f701653cbf83d66c20
 size 5496