Training in progress, epoch 1

config.json

@@ -0,0 +1,71 @@
+{
+  "_name_or_path": "vinai/PhoWhisper-medium",
+  "activation_dropout": 0.0,
+  "activation_function": "gelu",
+  "apply_spec_augment": false,
+  "architectures": [
+    "ViSpeechClassification"
+  ],
+  "attention_dropout": 0.0,
+  "begin_suppress_tokens": [
+    220,
+    50257
+  ],
+  "bos_token_id": 50257,
+  "classifier_proj_size": 256,
+  "d_model": 1024,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 4096,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 24,
+  "decoder_start_token_id": 50258,
+  "dropout": 0.0,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 4096,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 24,
+  "eos_token_id": 50257,
+  "forced_decoder_ids": null,
+  "id2label": {
+    "0": "male, northern dialect",
+    "1": "male, central dialect",
+    "2": "male, highland central dialect",
+    "3": "male, southern dialect",
+    "4": "female, northern dialect",
+    "5": "female, central dialect",
+    "6": "female, highland central dialect",
+    "7": "female, southern dialect"
+  },
+  "init_std": 0.02,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "female, central dialect": "5",
+    "female, highland central dialect": "6",
+    "female, northern dialect": "4",
+    "female, southern dialect": "7",
+    "male, central dialect": "1",
+    "male, highland central dialect": "2",
+    "male, northern dialect": "0",
+    "male, southern dialect": "3"
+  },
+  "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": 0.05,
+  "max_length": 448,
+  "max_source_positions": 1500,
+  "max_target_positions": 448,
+  "median_filter_width": 7,
+  "model_type": "whisper",
+  "num_hidden_layers": 24,
+  "num_mel_bins": 80,
+  "pad_token_id": 50257,
+  "scale_embedding": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.41.2",
+  "use_cache": true,
+  "use_weighted_layer_sum": false,
+  "vocab_size": 51865
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:57390c26ca78fa7df1428c07a8a26d1db0713746b8485ee909472e6cbe4f1e58
+size 1235737744

preprocessor_config.json

@@ -0,0 +1,14 @@
+{
+  "chunk_length": 30,
+  "feature_extractor_type": "WhisperFeatureExtractor",
+  "feature_size": 80,
+  "hop_length": 160,
+  "n_fft": 400,
+  "n_samples": 480000,
+  "nb_max_frames": 3000,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "WhisperProcessor",
+  "return_attention_mask": false,
+  "sampling_rate": 16000
+}

runs/Jul13_04-00-23_a32a8fc44f82/events.out.tfevents.1720843225.a32a8fc44f82.34.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c6c85921e25e1c0d52163660c47312d53ec384d914b7b19a0092d69c5f2a2494
+size 4184

runs/Jul13_04-03-01_a32a8fc44f82/events.out.tfevents.1720843382.a32a8fc44f82.247.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9471e9a2e4668855e309122a8ef1489d00137d9c12776453b78d7a2841127178
+size 4184

runs/Jul13_04-04-22_a32a8fc44f82/events.out.tfevents.1720843463.a32a8fc44f82.425.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2144cec0018af5c45268ebdd7c160d0c75e6bca40b4e0ab2aee02007b8197bd4
+size 7227

runs/Jul13_04-07-42_a32a8fc44f82/events.out.tfevents.1720843664.a32a8fc44f82.628.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:03383fb089b7a8268952fee161287d369f7dbc5ed0b8404d0a3ba2eb974b4cf4
+size 23885

runs/Jul13_05-53-57_a32a8fc44f82/events.out.tfevents.1720850039.a32a8fc44f82.3321.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f457f4e2adbd4da70cc4d74c5b454d60423bb6e26a4d54f50e972ff20d3ae19d
+size 27161

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2de8a88815cd7a4ce539dab9d23a753db8fa0d9213c6e02f1f86779a8bf8950f
+size 5176