Update tiny models for CanaryModel

chat_template.jinja

@@ -0,0 +1 @@
+{{- '<|startofcontext|><|startoftranscript|><|emo:undefined|>' -}}{%- for message in messages -%}{%- if message['role'] == 'user' -%}{%- for content in message['content'] if content['type'] == 'text' -%}{{- '<|' ~ content['source_language'] ~ '|>' -}}{{- '<|' ~ content['target_language'] ~ '|>' -}}{{- '<|pnc|>' if content['punctuation'] else '<|nopnc|>' -}}{{- '<|noitn|><|notimestamp|><|nodiarize|>' -}}{%- endfor -%}{%- elif message['role'] == 'assistant' -%}{{- message['content'] ~ '<|endoftext|>' -}}{%- endif -%}{%- endfor -%}

config.json

@@ -0,0 +1,61 @@
+{
+  "architectures": [
+    "CanaryModel"
+  ],
+  "bos_token_id": 4,
+  "decoder_config": {
+    "attention_bias": true,
+    "attention_dropout": 0.0,
+    "bos_token_id": 4,
+    "eos_token_id": 3,
+    "head_dim": 8,
+    "hidden_act": "relu",
+    "hidden_size": 16,
+    "initializer_range": 0.02,
+    "intermediate_size": 32,
+    "max_position_embeddings": 40,
+    "model_type": "canary_decoder",
+    "num_attention_heads": 2,
+    "num_hidden_layers": 2,
+    "num_key_value_heads": 2,
+    "pad_token_id": 2,
+    "use_cache": true,
+    "vocab_size": 16384
+  },
+  "decoder_start_token_id": 7,
+  "dtype": "float32",
+  "encoder_config": {
+    "activation_dropout": 0.1,
+    "attention_bias": true,
+    "attention_dropout": 0.1,
+    "conv_kernel_size": 9,
+    "convolution_bias": true,
+    "dropout": 0.1,
+    "dropout_positions": 0.0,
+    "hidden_act": "silu",
+    "hidden_size": 16,
+    "initializer_range": 0.02,
+    "intermediate_size": 32,
+    "layerdrop": 0.1,
+    "max_position_embeddings": 5000,
+    "model_type": "parakeet_encoder",
+    "num_attention_heads": 2,
+    "num_hidden_layers": 2,
+    "num_key_value_heads": 2,
+    "num_mel_bins": 80,
+    "scale_input": false,
+    "subsampling_conv_channels": 16,
+    "subsampling_conv_kernel_size": 3,
+    "subsampling_conv_stride": 2,
+    "subsampling_factor": 8
+  },
+  "eos_token_id": 3,
+  "initializer_range": 0.02,
+  "is_encoder_decoder": true,
+  "model_type": "canary",
+  "pad_token_id": 2,
+  "tie_word_embeddings": true,
+  "transformers_version": "5.19.0.dev0",
+  "use_cache": true,
+  "vocab_size": 16384
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7136cd32a946c25d3fb23fde2925619b879fe15e597a2f2966e92d8963a40d7c
+size 1143376

preprocessor_config.json

@@ -0,0 +1,12 @@
+{
+  "feature_extractor_type": "ParakeetFeatureExtractor",
+  "feature_size": 128,
+  "hop_length": 160,
+  "n_fft": 512,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "preemphasis": 0.97,
+  "return_attention_mask": true,
+  "sampling_rate": 16000,
+  "win_length": 400
+}

tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "backend": "tokenizers",
+  "bos_token": "<|startoftranscript|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|endoftext|>",
+  "is_local": true,
+  "local_files_only": false,
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "<pad>",
+  "processor_class": "CanaryProcessor",
+  "tokenizer_class": "TokenizersBackend",
+  "unk_token": "<unk>"
+}