Upload tiny models for SeamlessM4Tv2ForTextToSpeech

config.json

@@ -0,0 +1,124 @@
+{
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "adaptor_dropout": 0.1,
+  "adaptor_kernel_size": 8,
+  "adaptor_stride": 8,
+  "add_adapter": true,
+  "architectures": [
+    "SeamlessM4Tv2ForTextToSpeech"
+  ],
+  "attention_dropout": 0.1,
+  "attention_probs_dropout_prob": 0.1,
+  "bos_token_id": 2,
+  "char_vocab_size": 4,
+  "conv_depthwise_kernel_size": 31,
+  "decoder_attention_heads": 2,
+  "decoder_ffn_dim": 6,
+  "decoder_layerdrop": 0.05,
+  "decoder_layers": 2,
+  "decoder_start_token_id": 3,
+  "dropout": 0.1,
+  "encoder_attention_heads": 2,
+  "encoder_ffn_dim": 6,
+  "encoder_layerdrop": 0.05,
+  "encoder_layers": 2,
+  "eos_token_id": 3,
+  "feature_projection_input_dim": 160,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 6,
+  "initializer_range": 0.02,
+  "is_encoder_decoder": true,
+  "lang_embed_dim": 6,
+  "layer_norm_eps": 1e-05,
+  "leaky_relu_slope": 0.1,
+  "left_max_position_embeddings": 2,
+  "max_new_tokens": null,
+  "max_position_embeddings": 256,
+  "model_type": "seamless_m4t_v2",
+  "num_adapter_layers": 1,
+  "num_attention_heads": 2,
+  "num_conv_pos_embeddings": 8,
+  "num_hidden_layers": 2,
+  "pad_token_id": 0,
+  "position_embeddings_type": "relative_key",
+  "resblock_dilation_sizes": [
+    [
+      1,
+      3,
+      5
+    ],
+    [
+      1,
+      3,
+      5
+    ],
+    [
+      1,
+      3,
+      5
+    ]
+  ],
+  "resblock_kernel_sizes": [
+    3,
+    7,
+    11
+  ],
+  "right_max_position_embeddings": 1,
+  "sampling_rate": 16000,
+  "scale_embedding": true,
+  "speech_encoder_attention_heads": 2,
+  "speech_encoder_chunk_size": 2,
+  "speech_encoder_dropout": 0.0,
+  "speech_encoder_hidden_act": "swish",
+  "speech_encoder_intermediate_size": 6,
+  "speech_encoder_layerdrop": 0.1,
+  "speech_encoder_layers": 2,
+  "speech_encoder_left_chunk_num": 1,
+  "spkr_embed_dim": 6,
+  "t2u_bos_token_id": 0,
+  "t2u_decoder_attention_heads": 2,
+  "t2u_decoder_ffn_dim": 6,
+  "t2u_decoder_layers": 2,
+  "t2u_encoder_attention_heads": 2,
+  "t2u_encoder_ffn_dim": 6,
+  "t2u_encoder_layers": 2,
+  "t2u_eos_token_id": 2,
+  "t2u_max_position_embeddings": 4096,
+  "t2u_num_langs": 0,
+  "t2u_offset_tgt_lang": 0,
+  "t2u_pad_token_id": 1,
+  "t2u_variance_pred_dropout": 0.5,
+  "t2u_variance_predictor_embed_dim": 6,
+  "t2u_variance_predictor_hidden_dim": 4,
+  "t2u_variance_predictor_kernel_size": 3,
+  "t2u_vocab_size": 20,
+  "torch_dtype": "float32",
+  "transformers_version": "4.36.0.dev0",
+  "unit_embed_dim": 25,
+  "unit_hifi_gan_vocab_size": 20,
+  "unit_hifigan_vocab_vise": 20,
+  "upsample_initial_channel": 32,
+  "upsample_kernel_sizes": [
+    11,
+    8,
+    8,
+    4,
+    4
+  ],
+  "upsample_rates": [
+    5,
+    4,
+    4,
+    2,
+    2
+  ],
+  "use_cache": true,
+  "var_pred_dropout": 0.5,
+  "variance_predictor_kernel_size": 3,
+  "vocab_size": 20,
+  "vocoder_num_langs": 5,
+  "vocoder_num_spkrs": 5,
+  "vocoder_offset": 0
+}

generation_config.json

@@ -0,0 +1,8 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 2,
+  "decoder_start_token_id": 3,
+  "eos_token_id": 3,
+  "pad_token_id": 0,
+  "transformers_version": "4.36.0.dev0"
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:044af3f5f7ef7075bdbb5ead30d3cfff5bb8b147b17a235dfc6533d1eb75e01e
+size 309184

preprocessor_config.json

@@ -0,0 +1,10 @@
+{
+  "feature_extractor_type": "SeamlessM4TFeatureExtractor",
+  "feature_size": 80,
+  "num_mel_bins": 80,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "return_attention_mask": true,
+  "sampling_rate": 16000,
+  "stride": 2
+}