End of training

README.md

@@ -1,7 +1,12 @@
 ---
+language:
+- ml
 license: cc-by-nc-4.0
 base_model: facebook/mms-1b-all
 tags:
+- automatic-speech-recognition
+- mozilla-foundation/common_voice_16_0
+- mms
 - generated_from_trainer
 datasets:
 - common_voice_16_0
@@ -14,15 +19,15 @@ model-index:
       name: Automatic Speech Recognition
       type: automatic-speech-recognition
     dataset:
-      name: common_voice_16_0
+      name: MOZILLA-FOUNDATION/COMMON_VOICE_16_0 - ML
       type: common_voice_16_0
       config: ml
       split: test
-      args: ml
+      args: 'Config: ml, Training split: train+validation, Eval split: test'
     metrics:
     - name: Wer
       type: wer
-      value: 0.5345542501727713
+      value: 0.5348997926744989
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -30,10 +35,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # breeze-listen-w2v2-ml
 
-This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_16_0 dataset.
+This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the MOZILLA-FOUNDATION/COMMON_VOICE_16_0 - ML dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2698
-- Wer: 0.5346
+- Loss: 0.2666
+- Wer: 0.5349
 
 ## Model description
 

adapter.mal.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b41817255284f3ae007dc04b8e4eb15d1b4e3feac18c049ba9bd1282af988ead
+size 9018880

all_results.json

@@ -0,0 +1,14 @@
+{
+    "epoch": 4.0,
+    "eval_loss": 0.2665550410747528,
+    "eval_runtime": 160.6538,
+    "eval_samples": 663,
+    "eval_samples_per_second": 4.127,
+    "eval_steps_per_second": 0.517,
+    "eval_wer": 0.5348997926744989,
+    "train_loss": 2.1205503649827913,
+    "train_runtime": 5112.0325,
+    "train_samples": 1968,
+    "train_samples_per_second": 1.54,
+    "train_steps_per_second": 0.385
+}

breeze-listen-w2v2-ml.log

@@ -132,3 +132,20 @@ weight_decay=0.0,
 {'eval_loss': 0.27460750937461853, 'eval_wer': 0.5393918451969593, 'eval_runtime': 160.7333, 'eval_samples_per_second': 4.125, 'eval_steps_per_second': 0.516, 'epoch': 3.25}
 {'eval_loss': 0.26981213688850403, 'eval_wer': 0.5345542501727713, 'eval_runtime': 160.1257, 'eval_samples_per_second': 4.14, 'eval_steps_per_second': 0.518, 'epoch': 3.66}
 {'train_runtime': 5112.0325, 'train_samples_per_second': 1.54, 'train_steps_per_second': 0.385, 'train_loss': 2.1205503649827913, 'epoch': 4.0}
+***** train metrics *****
+  epoch                    =        4.0
+  train_loss               =     2.1206
+  train_runtime            = 1:25:12.03
+  train_samples            =       1968
+  train_samples_per_second =       1.54
+  train_steps_per_second   =      0.385
+01/29/2024 21:22:32 - INFO - __main__ - *** Evaluate ***
+***** eval metrics *****
+  epoch                   =        4.0
+  eval_loss               =     0.2666
+  eval_runtime            = 0:02:40.65
+  eval_samples            =        663
+  eval_samples_per_second =      4.127
+  eval_steps_per_second   =      0.517
+  eval_wer                =     0.5349
+01/29/2024 21:25:13 - INFO - __main__ - Saving adapter weights under /cosmos/home/sp-operator/ai/training/models/simpragma/breeze-listen-w2v2-ml/adapter.mal.safetensors...

eval_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 4.0,
+    "eval_loss": 0.2665550410747528,
+    "eval_runtime": 160.6538,
+    "eval_samples": 663,
+    "eval_samples_per_second": 4.127,
+    "eval_steps_per_second": 0.517,
+    "eval_wer": 0.5348997926744989
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.0,
+    "train_loss": 2.1205503649827913,
+    "train_runtime": 5112.0325,
+    "train_samples": 1968,
+    "train_samples_per_second": 1.54,
+    "train_steps_per_second": 0.385
+}

trainer_state.json

@@ -0,0 +1,129 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.0,
+  "eval_steps": 200,
+  "global_step": 1968,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.41,
+      "eval_loss": 5.472805500030518,
+      "eval_runtime": 162.595,
+      "eval_samples_per_second": 4.078,
+      "eval_steps_per_second": 0.51,
+      "eval_wer": 1.075673807878369,
+      "step": 200
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 5.127437114715576,
+      "eval_runtime": 163.1607,
+      "eval_samples_per_second": 4.063,
+      "eval_steps_per_second": 0.509,
+      "eval_wer": 1.003800967519005,
+      "step": 400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 0.0007890792291220557,
+      "loss": 6.5037,
+      "step": 500
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 0.6166694760322571,
+      "eval_runtime": 161.3235,
+      "eval_samples_per_second": 4.11,
+      "eval_steps_per_second": 0.514,
+      "eval_wer": 0.8130615065653075,
+      "step": 600
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 0.328411728143692,
+      "eval_runtime": 162.053,
+      "eval_samples_per_second": 4.091,
+      "eval_steps_per_second": 0.512,
+      "eval_wer": 0.582930200414651,
+      "step": 800
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 0.0005214132762312634,
+      "loss": 1.0482,
+      "step": 1000
+    },
+    {
+      "epoch": 2.03,
+      "eval_loss": 0.3169207274913788,
+      "eval_runtime": 165.1028,
+      "eval_samples_per_second": 4.016,
+      "eval_steps_per_second": 0.503,
+      "eval_wer": 0.5666897028334485,
+      "step": 1000
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 0.28758111596107483,
+      "eval_runtime": 160.9496,
+      "eval_samples_per_second": 4.119,
+      "eval_steps_per_second": 0.516,
+      "eval_wer": 0.5425017277125086,
+      "step": 1200
+    },
+    {
+      "epoch": 2.85,
+      "eval_loss": 0.2846720516681671,
+      "eval_runtime": 161.8788,
+      "eval_samples_per_second": 4.096,
+      "eval_steps_per_second": 0.513,
+      "eval_wer": 0.5521769177608846,
+      "step": 1400
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 0.00025374732334047106,
+      "loss": 0.4314,
+      "step": 1500
+    },
+    {
+      "epoch": 3.25,
+      "eval_loss": 0.27460750937461853,
+      "eval_runtime": 160.7333,
+      "eval_samples_per_second": 4.125,
+      "eval_steps_per_second": 0.516,
+      "eval_wer": 0.5393918451969593,
+      "step": 1600
+    },
+    {
+      "epoch": 3.66,
+      "eval_loss": 0.26981213688850403,
+      "eval_runtime": 160.1257,
+      "eval_samples_per_second": 4.14,
+      "eval_steps_per_second": 0.518,
+      "eval_wer": 0.5345542501727713,
+      "step": 1800
+    },
+    {
+      "epoch": 4.0,
+      "step": 1968,
+      "total_flos": 2.984133759613272e+18,
+      "train_loss": 2.1205503649827913,
+      "train_runtime": 5112.0325,
+      "train_samples_per_second": 1.54,
+      "train_steps_per_second": 0.385
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 1968,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 200,
+  "total_flos": 2.984133759613272e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}