Training in progress, step 1000

config.json

@@ -0,0 +1,118 @@
+{
+  "_name_or_path": "facebook/seamless-m4t-v2-large",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "adaptor_dropout": 0.1,
+  "adaptor_kernel_size": 8,
+  "adaptor_stride": 8,
+  "add_adapter": true,
+  "architectures": [
+    "SeamlessM4Tv2Model"
+  ],
+  "attention_dropout": 0.1,
+  "bos_token_id": 2,
+  "char_vocab_size": 10943,
+  "conv_depthwise_kernel_size": 31,
+  "decoder_attention_heads": 16,
+  "decoder_ffn_dim": 8192,
+  "decoder_layerdrop": 0.05,
+  "decoder_layers": 24,
+  "decoder_start_token_id": 3,
+  "dropout": 0.1,
+  "encoder_attention_heads": 16,
+  "encoder_ffn_dim": 8192,
+  "encoder_layerdrop": 0.05,
+  "encoder_layers": 24,
+  "eos_token_id": 3,
+  "feature_projection_input_dim": 160,
+  "hidden_size": 1024,
+  "initializer_range": 0.02,
+  "is_encoder_decoder": true,
+  "lang_embed_dim": 256,
+  "layer_norm_eps": 1e-05,
+  "leaky_relu_slope": 0.1,
+  "left_max_position_embeddings": 64,
+  "max_new_tokens": 256,
+  "max_position_embeddings": 4096,
+  "model_type": "seamless_m4t_v2",
+  "num_adapter_layers": 1,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
+  "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": 8,
+  "sampling_rate": 16000,
+  "scale_embedding": true,
+  "speech_encoder_attention_heads": 16,
+  "speech_encoder_chunk_size": 20000,
+  "speech_encoder_dropout": 0.0,
+  "speech_encoder_hidden_act": "swish",
+  "speech_encoder_intermediate_size": 4096,
+  "speech_encoder_layerdrop": 0.1,
+  "speech_encoder_layers": 24,
+  "speech_encoder_left_chunk_num": 128,
+  "spkr_embed_dim": 256,
+  "t2u_bos_token_id": 0,
+  "t2u_decoder_attention_heads": 16,
+  "t2u_decoder_ffn_dim": 8192,
+  "t2u_decoder_layers": 6,
+  "t2u_encoder_attention_heads": 16,
+  "t2u_encoder_ffn_dim": 8192,
+  "t2u_encoder_layers": 6,
+  "t2u_eos_token_id": 2,
+  "t2u_max_position_embeddings": 4096,
+  "t2u_pad_token_id": 1,
+  "t2u_variance_pred_dropout": 0.5,
+  "t2u_variance_predictor_embed_dim": 1024,
+  "t2u_variance_predictor_hidden_dim": 256,
+  "t2u_variance_predictor_kernel_size": 3,
+  "t2u_vocab_size": 10082,
+  "torch_dtype": "float32",
+  "transformers_version": "4.44.0",
+  "unit_embed_dim": 1280,
+  "unit_hifi_gan_vocab_size": 10000,
+  "upsample_initial_channel": 512,
+  "upsample_kernel_sizes": [
+    11,
+    8,
+    8,
+    4,
+    4
+  ],
+  "upsample_rates": [
+    5,
+    4,
+    4,
+    2,
+    2
+  ],
+  "use_cache": false,
+  "var_pred_dropout": 0.5,
+  "variance_predictor_kernel_size": 3,
+  "vocab_size": 256102,
+  "vocoder_num_langs": 36,
+  "vocoder_num_spkrs": 200,
+  "vocoder_offset": 4
+}

model-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4319d8ddbe639a261e0fae2e93aefd693204bc605f042b36bf4c0f59816e678a
+size 4999163080

model-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:179be78454f4a4d9fa721301155573d6634f6cd96f445ad006fce5f73f0008f2
+size 4238114628

preprocessor_config.json

@@ -0,0 +1,111 @@
+{
+  "feature_extractor_type": "SeamlessM4TFeatureExtractor",
+  "feature_size": 80,
+  "language_code": [
+    "__afr__",
+    "__amh__",
+    "__arb__",
+    "__ary__",
+    "__arz__",
+    "__asm__",
+    "__azj__",
+    "__bel__",
+    "__ben__",
+    "__bos__",
+    "__bul__",
+    "__cat__",
+    "__ceb__",
+    "__ces__",
+    "__ckb__",
+    "__cmn__",
+    "__cmn_Hant__",
+    "__cym__",
+    "__dan__",
+    "__deu__",
+    "__ell__",
+    "__eng__",
+    "__est__",
+    "__eus__",
+    "__fin__",
+    "__fra__",
+    "__fuv__",
+    "__gaz__",
+    "__gle__",
+    "__glg__",
+    "__guj__",
+    "__heb__",
+    "__hin__",
+    "__hrv__",
+    "__hun__",
+    "__hye__",
+    "__ibo__",
+    "__ind__",
+    "__isl__",
+    "__ita__",
+    "__jav__",
+    "__jpn__",
+    "__kan__",
+    "__kat__",
+    "__kaz__",
+    "__khk__",
+    "__khm__",
+    "__kir__",
+    "__kor__",
+    "__lao__",
+    "__lit__",
+    "__lug__",
+    "__luo__",
+    "__lvs__",
+    "__mai__",
+    "__mal__",
+    "__mar__",
+    "__mkd__",
+    "__mlt__",
+    "__mni__",
+    "__mya__",
+    "__nld__",
+    "__nno__",
+    "__nob__",
+    "__npi__",
+    "__nya__",
+    "__ory__",
+    "__pan__",
+    "__pbt__",
+    "__pes__",
+    "__pol__",
+    "__por__",
+    "__ron__",
+    "__rus__",
+    "__sat__",
+    "__slk__",
+    "__slv__",
+    "__sna__",
+    "__snd__",
+    "__som__",
+    "__spa__",
+    "__srp__",
+    "__swe__",
+    "__swh__",
+    "__tam__",
+    "__tel__",
+    "__tgk__",
+    "__tgl__",
+    "__tha__",
+    "__tur__",
+    "__ukr__",
+    "__urd__",
+    "__uzn__",
+    "__vie__",
+    "__yor__",
+    "__yue__",
+    "__zlm__",
+    "__zul__"
+  ],
+  "num_mel_bins": 80,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "SeamlessM4TProcessor",
+  "return_attention_mask": true,
+  "sampling_rate": 16000,
+  "stride": 2
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:58c450e37eef62c33c4e066b23a9662a2b6fad87930b9bd11d80e0e0197af58a
+size 5304