Training in progress, step 400

config.json

@@ -8,12 +8,12 @@
   "architectures": [
     "Wav2Vec2ForCTC"
   ],
-  "attention_dropout": 0.2,
+  "attention_dropout": 0.05,
   "bos_token_id": 1,
   "classifier_proj_size": 256,
   "codevector_dim": 768,
   "contrastive_logits_temperature": 0.1,
-  "conv_bias": false,
+  "conv_bias": true,
   "conv_dim": [
     512,
     512,
@@ -41,7 +41,7 @@
     2,
     2
   ],
-  "ctc_loss_reduction": "mean",
+  "ctc_loss_reduction": "sum",
   "ctc_zero_infinity": false,
   "diversity_loss_weight": 0.1,
   "do_stable_layer_norm": true,
@@ -53,12 +53,12 @@
   "feat_quantizer_dropout": 0.0,
   "final_dropout": 0.0,
   "hidden_act": "gelu",
-  "hidden_dropout": 0.2,
+  "hidden_dropout": 0.05,
   "hidden_size": 1024,
   "initializer_range": 0.02,
   "intermediate_size": 4096,
   "layer_norm_eps": 1e-05,
-  "layerdrop": 0.15,
+  "layerdrop": 0.2,
   "mask_channel_length": 10,
   "mask_channel_min_space": 1,
   "mask_channel_other": 0.0,

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3d409539beb094d9f0a79a19fc98bbcd7130c1b0d28ab588cd68988c3e356ef7
-size 1262180305
+oid sha256:e9b63a24031b5ec39f8d3f961e430dd4672befc291fd2ea92fbe3e0c8bd3dd99
+size 1262197037

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f73353db26f106ae805e0d4b5046ae8d42c5887615e4003f427861b571f1308c
+oid sha256:aedb057b875f61276bccb15c8d4e72cecefc56dbaa118155c1e631a9510704cc
 size 3579