finish model

config.json

@@ -2,6 +2,7 @@
   "architectures": [
     "SpeechEncoderDecoderModel"
   ],
+  "bos_token_id": 0,
   "decoder": {
     "_name_or_path": "",
     "activation_dropout": 0.1,
@@ -28,7 +29,6 @@
     "decoder_start_token_id": 2,
     "diversity_penalty": 0.0,
     "do_sample": false,
-    "do_stable_layer_norm": true,
     "dropout": 0.1,
     "early_stopping": true,
     "encoder_attention_heads": 4,
@@ -59,12 +59,12 @@
     "max_source_positions": 6000,
     "max_target_positions": 1024,
     "min_length": 0,
-    "model_type": "speech_to_text",
+    "model_type": "speech_to_text_2",
     "no_repeat_ngram_size": 0,
     "num_beam_groups": 1,
     "num_beams": 5,
     "num_conv_layers": 2,
-    "num_hidden_layers": 12,
+    "num_hidden_layers": 7,
     "num_return_sequences": 1,
     "output_attentions": false,
     "output_hidden_states": false,
@@ -219,8 +219,10 @@
     "use_bfloat16": false,
     "vocab_size": 32
   },
+  "eos_token_id": 2,
   "is_encoder_decoder": true,
   "model_type": "encoder-decoder",
+  "pad_token_id": 1,
   "tie_word_embeddings": false,
   "torch_dtype": "float32",
   "transformers_version": null

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:77645fc80a478180a14cb2026b47dac569fd1f1835f6b8bf1d944946bc26c2cd
+oid sha256:d778e96434ac418eb4b0410a06f8f2986a5e4549aeade96281f475ff96a5c2f6
 size 1329244376

special_tokens_map.json

@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}

tokenizer_config.json

@@ -0,0 +1 @@
+{"unk_token": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "pad_token": "<pad>", "tokenizer_class": "Speech2Text2Tokenizer"}