update

Signed-off-by: nithinraok <[email protected]>

config.json

@@ -2,19 +2,18 @@
   "architectures": [
     "ParakeetCTC"
   ],
-  "blank_token_id": 0,
+  "blank_token_id": 1024,
   "bos_token_id": 1,
   "ctc_loss_reduction": "mean",
   "ctc_zero_infinity": true,
-  "eos_token_id": 2,
-  "fastconformer_config": {
+  "encoder_config": {
     "activation_dropout": 0.1,
     "activation_function": "silu",
     "architectures": [
       "FastConformerModel"
     ],
-    "attention_bias": false,
     "attention_dropout": 0.1,
+    "attention_probs_dropout_prob": 0.1,
     "conv_kernel_size": 9,
     "d_model": 1024,
     "dropout": 0.1,
@@ -23,36 +22,29 @@
     "encoder_ffn_dim": 4096,
     "encoder_layerdrop": 0.1,
     "encoder_layers": 42,
-    "head_dim": 128,
     "hidden_act": "silu",
     "hidden_dropout_prob": 0.1,
     "hidden_size": 1024,
     "initializer_range": 0.02,
     "intermediate_size": 4096,
-    "max_position_embeddings": 2048,
-    "mlp_bias": false,
     "model_type": "fastconformer",
     "nemo_decoder_type": "nemo.collections.asr.modules.ConvASRDecoder",
     "nemo_encoder_type": "nemo.collections.asr.modules.ConformerEncoder",
     "nemo_model_type": "parakeet_ctc",
     "num_attention_heads": 8,
     "num_hidden_layers": 42,
-    "num_key_value_heads": 8,
     "num_mel_bins": 80,
-    "pretraining_tp": 1,
-    "rms_norm_eps": 1e-06,
-    "rope_scaling": null,
-    "rope_theta": 10000.0,
     "subsampling_conv_channels": 256,
     "subsampling_factor": 8,
     "use_bias": true,
-    "use_cache": true,
+    "use_cache": false,
     "vocab_size": 1024,
     "xscaling": true
   },
+  "eos_token_id": 2,
   "model_type": "parakeet_ctc",
   "pad_token_id": 0,
   "torch_dtype": "float32",
-  "transformers_version": "4.53.0.dev0",
+  "transformers_version": "4.54.0.dev0",
   "vocab_size": 1025
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55a59b38ff5e4773e5d836fb997e2b3529fe44406c86ba1d5009f1c29df06a3d
-size 4250700812
+oid sha256:69a77e28f784139d2c5460a99cb78f39511d15798c279495d1630d579bb08d44
+size 4250700828

tokenizer_config.json

@@ -0,0 +1,19 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "blank_token_id": 1024,
+  "clean_up_tokenization_spaces": false,
+  "do_lower_case": false,
+  "extra_special_tokens": {},
+  "model_max_length": 1000000000000000019884624838656,
+  "tokenizer_class": "ParakeetCTCTokenizer",
+  "unk_token": "<unk>"
+}