Training in progress, step 100

README.md

@@ -12,8 +12,8 @@ model-index:
 - name: w2v-bert-2.0-lmk_man_align_updated
   results:
   - task:
-      type: automatic-speech-recognition
       name: Automatic Speech Recognition
+      type: automatic-speech-recognition
     dataset:
       name: audiofolder
       type: audiofolder
@@ -21,9 +21,9 @@ model-index:
       split: test
       args: default
     metrics:
-    - type: wer
+    - name: Wer
+      type: wer
       value: 0.9895470383275261
-      name: Wer
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You

config.json

@@ -8,7 +8,7 @@
   "architectures": [
     "Wav2Vec2BertForCTC"
   ],
-  "attention_dropout": 0.0,
+  "attention_dropout": 0.1,
   "bos_token_id": 1,
   "classifier_proj_size": 768,
   "codevector_dim": 768,
@@ -18,25 +18,26 @@
   "ctc_loss_reduction": "mean",
   "ctc_zero_infinity": false,
   "diversity_loss_weight": 0.1,
+  "dtype": "float32",
   "eos_token_id": 2,
   "feat_proj_dropout": 0.0,
   "feat_quantizer_dropout": 0.0,
   "feature_projection_input_dim": 160,
   "final_dropout": 0.1,
   "hidden_act": "swish",
-  "hidden_dropout": 0.0,
+  "hidden_dropout": 0.1,
   "hidden_size": 1024,
   "initializer_range": 0.02,
   "intermediate_size": 4096,
   "layer_norm_eps": 1e-05,
-  "layerdrop": 0.0,
+  "layerdrop": 0.1,
   "left_max_position_embeddings": 64,
   "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
   "mask_feature_prob": 0.0,
   "mask_time_length": 10,
   "mask_time_min_masks": 2,
-  "mask_time_prob": 0.0,
+  "mask_time_prob": 0.05,
   "max_source_positions": 5000,
   "model_type": "wav2vec2-bert",
   "num_adapter_layers": 1,
@@ -72,8 +73,7 @@
     1,
     1
   ],
-  "torch_dtype": "float32",
-  "transformers_version": "4.52.2",
+  "transformers_version": "4.57.1",
   "use_intermediate_ffn_before_adapter": false,
   "use_weighted_layer_sum": false,
   "vocab_size": 32,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:be7cfbff7a9912925c9604c6b80a4fd8f0ec8b28074cf142a8cb8233d9996ba3
-size 2422945760
+oid sha256:80ea8dd3de07177468e6c901349e28dca16d9c7bae51e786eb44542ae1f8d78e
+size 2422949960

runs/Nov19_19-45-12_x1001c1s7b0n1/events.out.tfevents.1763599523.x1001c1s7b0n1.226985.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9e483acd06cab5e32648730a34bca413742657f82cf54ae0fa7b090ad75ba7f1
+size 7088

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:66fa8df3ac24ed128a27532ea114d2ad1efc9cd84b888ec930f0ff7d444bd9fa
-size 5777
+oid sha256:5754bea4554010570e4b649a8af6fe908d31b9892c904fc60888b08780735aec
+size 5905