Training in progress, step 300

config.json

@@ -1,6 +1,5 @@
 {
   "_name_or_path": "ad019el/Kabyle_xlsr-finetuned-tamasheq.en",
-  "gradient_checkpointing": true,
   "activation_dropout": 0.05,
   "adapter_attn_dim": null,
   "adapter_kernel_size": 3,
@@ -10,7 +9,7 @@
   "architectures": [
     "Wav2Vec2ForCTC"
   ],
-  "attention_dropout": 0.1,
+  "attention_dropout": 0.2,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
   "codevector_dim": 768,
@@ -51,16 +50,16 @@
   "feat_extract_activation": "gelu",
   "feat_extract_dropout": 0.0,
   "feat_extract_norm": "layer",
-  "feat_proj_dropout": 0.0,
+  "feat_proj_dropout": 0.1,
   "feat_quantizer_dropout": 0.0,
   "final_dropout": 0.0,
   "hidden_act": "gelu",
-  "hidden_dropout": 0.1,
+  "hidden_dropout": 0.2,
   "hidden_size": 1024,
   "initializer_range": 0.02,
   "intermediate_size": 4096,
   "layer_norm_eps": 1e-05,
-  "layerdrop": 0.2,
+  "layerdrop": 0.4,
   "mask_feature_length": 10,
   "mask_feature_min_masks": 0,
   "mask_feature_prob": 0.0,
@@ -78,7 +77,7 @@
   "num_hidden_layers": 24,
   "num_negatives": 100,
   "output_hidden_size": 1024,
-  "pad_token_id": 29,
+  "pad_token_id": 31,
   "proj_codevector_dim": 768,
   "tdnn_dilation": [
     1,
@@ -104,6 +103,6 @@
   "torch_dtype": "float32",
   "transformers_version": "4.33.0",
   "use_weighted_layer_sum": false,
-  "vocab_size": 30,
+  "vocab_size": 32,
   "xvector_output_dim": 512
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b126ad797b0635deb678753911532f0d4779b545d8b03f6b8b8a5aa01c82dbf
-size 1262024813
+oid sha256:b9241009547bda1ca1601c79ba7d1e84b11d2e2db1ef8f245de563fa7029d6a0
+size 1262033069

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a572e40d9f760e26ad5c68b12aca2f753f6ae4394cddfb369c5e0ef220fd51a8
+oid sha256:8d3a0ab3e996c2bd91c1438deb204abcd6599fa6c2af500d250ae9d18fd61fdb
 size 4027