End of training

README.md

@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 22.6258
-- Wer: 0.9999
+- Loss: 1.0874
+- Wer: 0.8986
 
 ## Model description
 
@@ -38,28 +38,34 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.001
+- learning_rate: 0.0001
 - train_batch_size: 4
 - eval_batch_size: 8
 - seed: 42
 - optimizer: Use 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: 500
-- num_epochs: 25
+- num_epochs: 10
 - mixed_precision_training: Native AMP
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Wer    |
 |:-------------:|:------:|:----:|:---------------:|:------:|
-| 2.9169        | 0.4926 | 500  | 0.9563          | 0.6717 |
-| 1.1420        | 0.9852 | 1000 | 0.9029          | 0.6855 |
-| 0.9580        | 1.4778 | 1500 | 0.8356          | 0.6427 |
-| 1.1198        | 1.9704 | 2000 | 0.8504          | 0.6769 |
-| 1.6392        | 2.4631 | 2500 | 3.2183          | 1.0    |
-| 6.0871        | 2.9557 | 3000 | 14.2035         | 1.0315 |
-| 25.2255       | 3.4483 | 3500 | 22.6256         | 0.9999 |
-| 22.4894       | 3.9409 | 4000 | 22.6258         | 0.9999 |
+| 5.9632        | 0.4926 | 500  | 1.0587          | 0.7186 |
+| 1.2221        | 0.9852 | 1000 | 0.9062          | 0.6633 |
+| 1.0028        | 1.4778 | 1500 | 0.8629          | 0.6419 |
+| 1.1607        | 1.9704 | 2000 | 0.8534          | 0.6337 |
+| 1.0608        | 2.4631 | 2500 | 1.0133          | 0.7632 |
+| 1.3117        | 2.9557 | 3000 | 1.0389          | 0.7365 |
+| 1.3259        | 3.4483 | 3500 | 1.0874          | 0.8991 |
+| 1.2260        | 3.9409 | 4000 | 1.0875          | 0.8983 |
+| 1.3359        | 4.4335 | 4500 | 1.0875          | 0.8989 |
+| 1.2309        | 4.9261 | 5000 | 1.0875          | 0.8979 |
+| 1.2623        | 5.4187 | 5500 | 1.0875          | 0.8985 |
+| 1.2284        | 5.9113 | 6000 | 1.0875          | 0.8982 |
+| 1.3132        | 6.4039 | 6500 | 1.0875          | 0.8985 |
+| 1.3229        | 6.8966 | 7000 | 1.0874          | 0.8986 |
 
 
 ### Framework versions

adapter.khm.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b20706e75773ce132619bd8d6fecbd3e2685bb2893ef6c4b56bb24632aae4ba1
-size 9162368
+oid sha256:18e06ed91597a11d202f781ce7e567f5e368d2d5ef83df3349a0ac0498e1084b
+size 9029128

config.json

@@ -77,7 +77,7 @@
   "num_hidden_layers": 48,
   "num_negatives": 100,
   "output_hidden_size": 1280,
-  "pad_token_id": 99,
+  "pad_token_id": 73,
   "proj_codevector_dim": 1024,
   "tdnn_dilation": [
     1,
@@ -103,6 +103,6 @@
   "transformers_version": "5.0.0.dev0",
   "use_cache": false,
   "use_weighted_layer_sum": false,
-  "vocab_size": 102,
+  "vocab_size": 76,
   "xvector_output_dim": 512
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:164362f7e5aeaf3db122c877d43275e87cbe1b79baebd7f39c0474646db2e1e8
-size 3859254680
+oid sha256:1c24975998840862e3f61a01f98ebc32afe3591229979b25294fa9c5c33e1bdb
+size 3859121448

runs/Dec18_10-28-34_Hong/events.out.tfevents.1766028514.Hong

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8015f63eea21e6776acd09686a9da9010f15270b5c315f391e6245b684dd7d74
+size 13399

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4bb520140b6e1cb71e669ce7c3ebb14a3b966536e23b53b856f2a23f2741e26b
+oid sha256:a4dd97ab74010796555cc2b4db00f8eb2bdacb0bad31c58579c0fa23f46bd7c3
 size 4792