voice: add Marvis TTS 250m v0.2 (MLX 8bit/4bit) + Kyutai Mimi codec (CC-BY-4.0)

.gitattributes

@@ -11,3 +11,5 @@ tokenizer.json filter=lfs diff=lfs merge=lfs -text
 vocab.json filter=lfs diff=lfs merge=lfs -text
 merges.txt filter=lfs diff=lfs merge=lfs -text
 *.litertlm filter=lfs diff=lfs merge=lfs -text
+voice/**/*.safetensors filter=lfs diff=lfs merge=lfs -text
+voice/**/*.wav filter=lfs diff=lfs merge=lfs -text

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/README.md

@@ -0,0 +1,28 @@
+---
+license: apache-2.0
+datasets:
+- amphion/Emilia-Dataset
+language:
+- en
+- fr
+- de
+library_name: transformers
+tags:
+- mlx
+- mlx-audio
+- transformers
+- mlx
+---
+
+# Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit
+This model was converted to MLX format from [`Marvis-AI/marvis-tts-250m-v0.2`](https://huggingface.co/Marvis-AI/marvis-tts-250m-v0.2) using mlx-audio version **0.2.6**.
+Refer to the [original model card](https://huggingface.co/Marvis-AI/marvis-tts-250m-v0.2) for more details on the model.
+## Use with mlx
+
+```bash
+pip install -U mlx-audio
+```
+
+```bash
+python -m mlx_audio.tts.generate --model Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit --text "Describe this image."
+```

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/config.json

@@ -0,0 +1,307 @@
+{
+    "add_cross_attention": false,
+    "architectures": [
+        "CsmForConditionalGeneration"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "audio_eos_token_id": 128003,
+    "audio_num_codebooks": 32,
+    "audio_token_id": 128002,
+    "audio_vocab_size": 2051,
+    "backbone_flavor": "llama-250M",
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": 128000,
+    "chunk_size_feed_forward": 0,
+    "codebook_eos_token_id": 0,
+    "codebook_pad_token_id": 2050,
+    "codec_config": {
+        "sampling_rate": 24000,
+        "audio_channels": 1,
+        "hidden_size": 512,
+        "num_filters": 64,
+        "num_residual_layers": 1,
+        "upsampling_ratios": [
+            8,
+            6,
+            5,
+            4
+        ],
+        "kernel_size": 7,
+        "last_kernel_size": 3,
+        "residual_kernel_size": 3,
+        "dilation_growth_rate": 2,
+        "use_causal_conv": true,
+        "pad_mode": "constant",
+        "compress": 2,
+        "trim_right_ratio": 1.0,
+        "codebook_size": 2048,
+        "codebook_dim": 256,
+        "num_quantizers": 32,
+        "use_conv_shortcut": false,
+        "vector_quantization_hidden_dimension": 256,
+        "upsample_groups": 512,
+        "num_hidden_layers": 8,
+        "intermediate_size": 2048,
+        "num_attention_heads": 8,
+        "num_key_value_heads": 8,
+        "hidden_act": "gelu",
+        "max_position_embeddings": 8000,
+        "initializer_range": 0.02,
+        "norm_eps": 1e-05,
+        "use_cache": false,
+        "use_streaming": false,
+        "rope_theta": 10000.0,
+        "sliding_window": 250,
+        "attention_dropout": 0.0,
+        "head_dim": 64,
+        "layer_scale_initial_scale": 0.01,
+        "attention_bias": false,
+        "_frame_rate": 12.5,
+        "num_semantic_quantizers": 1,
+        "return_dict": true,
+        "output_hidden_states": false,
+        "torchscript": false,
+        "dtype": "float32",
+        "pruned_heads": {},
+        "tie_word_embeddings": true,
+        "chunk_size_feed_forward": 0,
+        "is_encoder_decoder": false,
+        "is_decoder": false,
+        "cross_attention_hidden_size": null,
+        "add_cross_attention": false,
+        "tie_encoder_decoder": false,
+        "architectures": [
+            "MimiModel"
+        ],
+        "finetuning_task": null,
+        "id2label": {
+            "0": "LABEL_0",
+            "1": "LABEL_1"
+        },
+        "label2id": {
+            "LABEL_0": 0,
+            "LABEL_1": 1
+        },
+        "task_specific_params": null,
+        "problem_type": null,
+        "tokenizer_class": null,
+        "prefix": null,
+        "bos_token_id": null,
+        "pad_token_id": null,
+        "eos_token_id": null,
+        "sep_token_id": null,
+        "decoder_start_token_id": null,
+        "max_length": 20,
+        "min_length": 0,
+        "do_sample": false,
+        "early_stopping": false,
+        "num_beams": 1,
+        "temperature": 1.0,
+        "top_k": 50,
+        "top_p": 1.0,
+        "typical_p": 1.0,
+        "repetition_penalty": 1.0,
+        "length_penalty": 1.0,
+        "no_repeat_ngram_size": 0,
+        "encoder_no_repeat_ngram_size": 0,
+        "bad_words_ids": null,
+        "num_return_sequences": 1,
+        "output_scores": false,
+        "return_dict_in_generate": false,
+        "forced_bos_token_id": null,
+        "forced_eos_token_id": null,
+        "remove_invalid_values": false,
+        "exponential_decay_length_penalty": null,
+        "suppress_tokens": null,
+        "begin_suppress_tokens": null,
+        "num_beam_groups": 1,
+        "diversity_penalty": 0.0,
+        "_name_or_path": "kyutai/mimi",
+        "normalize": false,
+        "tf_legacy_loss": false,
+        "use_bfloat16": false,
+        "model_type": "mimi",
+        "output_attentions": false
+    },
+    "cross_attention_hidden_size": null,
+    "decoder_flavor": "llama-60M",
+    "decoder_start_token_id": null,
+    "depth_decoder_config": {
+        "return_dict": true,
+        "output_hidden_states": false,
+        "torchscript": false,
+        "dtype": null,
+        "pruned_heads": {},
+        "tie_word_embeddings": false,
+        "chunk_size_feed_forward": 0,
+        "is_encoder_decoder": false,
+        "is_decoder": false,
+        "cross_attention_hidden_size": null,
+        "add_cross_attention": false,
+        "tie_encoder_decoder": false,
+        "architectures": null,
+        "finetuning_task": null,
+        "id2label": {
+            "0": "LABEL_0",
+            "1": "LABEL_1"
+        },
+        "label2id": {
+            "LABEL_0": 0,
+            "LABEL_1": 1
+        },
+        "task_specific_params": null,
+        "problem_type": null,
+        "tokenizer_class": null,
+        "prefix": null,
+        "bos_token_id": null,
+        "pad_token_id": null,
+        "eos_token_id": null,
+        "sep_token_id": null,
+        "decoder_start_token_id": null,
+        "max_length": 20,
+        "min_length": 0,
+        "do_sample": false,
+        "early_stopping": false,
+        "num_beams": 1,
+        "temperature": 1.0,
+        "top_k": 50,
+        "top_p": 1.0,
+        "typical_p": 1.0,
+        "repetition_penalty": 1.0,
+        "length_penalty": 1.0,
+        "no_repeat_ngram_size": 0,
+        "encoder_no_repeat_ngram_size": 0,
+        "bad_words_ids": null,
+        "num_return_sequences": 1,
+        "output_scores": false,
+        "return_dict_in_generate": false,
+        "forced_bos_token_id": null,
+        "forced_eos_token_id": null,
+        "remove_invalid_values": false,
+        "exponential_decay_length_penalty": null,
+        "suppress_tokens": null,
+        "begin_suppress_tokens": null,
+        "num_beam_groups": 1,
+        "diversity_penalty": 0.0,
+        "_name_or_path": "",
+        "model_type": "csm_depth_decoder_model",
+        "tf_legacy_loss": false,
+        "use_bfloat16": false,
+        "num_codebooks": 32,
+        "vocab_size": 2051,
+        "backbone_hidden_size": 1536,
+        "max_position_embeddings": 33,
+        "hidden_size": 1024,
+        "intermediate_size": 4096,
+        "num_hidden_layers": 4,
+        "num_attention_heads": 8,
+        "num_key_value_heads": 2,
+        "hidden_act": "silu",
+        "initializer_range": 0.02,
+        "rms_norm_eps": 1e-05,
+        "use_cache": true,
+        "rope_theta": 500000,
+        "rope_scaling": {
+            "factor": 32.0,
+            "high_freq_factor": 0.0078125,
+            "low_freq_factor": 0.001953125,
+            "original_max_position_embeddings": 16,
+            "rope_type": "llama3"
+        },
+        "attention_bias": false,
+        "attention_dropout": 0.0,
+        "mlp_bias": false,
+        "head_dim": 128,
+        "output_attentions": false
+    },
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "dtype": null,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 1536,
+    "id2label": {
+        "0": "LABEL_0",
+        "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 8192,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+        "LABEL_0": 0,
+        "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 2048,
+    "min_length": 0,
+    "mlp_bias": false,
+    "model_type": "csm",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 12,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_codebooks": 32,
+    "num_hidden_layers": 6,
+    "num_key_value_heads": 3,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 128002,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "quantization": {
+        "group_size": 64,
+        "bits": 4,
+        "mode": "affine"
+    },
+    "quantization_config": {
+        "group_size": 64,
+        "bits": 4,
+        "mode": "affine"
+    },
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+        "factor": 32.0,
+        "high_freq_factor": 0.5,
+        "low_freq_factor": 0.125,
+        "original_max_position_embeddings": 1024,
+        "rope_type": "llama3"
+    },
+    "rope_theta": 500000,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "text_tokenizer": "Marvis-AI/marvis-tts-250m-v0.2",
+    "text_vocab_size": 49152,
+    "tf_legacy_loss": false,
+    "tie_codebooks_embeddings": true,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": false,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torchscript": false,
+    "transformers_version": "4.57.1",
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": true,
+    "vocab_size": 2051
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/generation_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token_id": 1,
+  "depth_decoder__from_model_config": true,
+  "depth_decoder_do_sample": true,
+  "depth_decoder_temperature": 0.9,
+  "depth_decoder_top_k": 50,
+  "do_sample": true,
+  "max_new_tokens": 125,
+  "pad_token_id": 0,
+  "temperature": 0.9,
+  "transformers_version": "4.54.1"
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/model.safetensors.index.json

@@ -0,0 +1,253 @@
+{
+    "metadata": {
+        "total_size": 413620768,
+        "total_parameters": 599911265
+    },
+    "weight_map": {
+        "model.audio_embeddings.biases": "model.safetensors",
+        "model.audio_embeddings.scales": "model.safetensors",
+        "model.audio_embeddings.weight": "model.safetensors",
+        "model.audio_head": "model.safetensors",
+        "model.backbone.layers.0.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.norm.weight": "model.safetensors",
+        "model.codebook0_head.biases": "model.safetensors",
+        "model.codebook0_head.scales": "model.safetensors",
+        "model.codebook0_head.weight": "model.safetensors",
+        "model.decoder.layers.0.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.norm.weight": "model.safetensors",
+        "model.projection.biases": "model.safetensors",
+        "model.projection.scales": "model.safetensors",
+        "model.projection.weight": "model.safetensors",
+        "model.text_embeddings.biases": "model.safetensors",
+        "model.text_embeddings.scales": "model.safetensors",
+        "model.text_embeddings.weight": "model.safetensors"
+    }
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/prompts/conversational_a.txt

@@ -0,0 +1 @@
+Coffee's story likely begins in Ethiopia, where legend tells of a goat herder named Kaldi who noticed his goats became energetic after eating red berries from a particular bush. Curious, he tried them himself, and felt invigorated.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/prompts/conversational_b.txt

@@ -0,0 +1 @@
+Coffee's story likely begins in Ethiopia, where legend tells of a goat herder named Kaldi who noticed his goats became energetic after eating red berries from a particular bush. Curious, he tried them himself, and felt invigorated.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/prompts/conversational_de.txt

@@ -0,0 +1 @@
+Ich werde bald mehrere Sprachen sprechen, darunter auch Deutsch.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/prompts/conversational_fr.txt

@@ -0,0 +1 @@
+J'ai toujours cru qu’un étrange voleur joyeux mangeait rapidement cinq gâteaux exquis près du vieux phare.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|im_start|>",
+    "<|im_end|>",
+    "<repo_name>",
+    "<reponame>",
+    "<file_sep>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<jupyter_script>",
+    "<empty_output>"
+  ],
+  "bos_token": "<|im_start|>",
+  "eos_token": "<|im_end|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-4bit/tokenizer_config.json

@@ -0,0 +1,168 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<repo_name>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<file_sep>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<jupyter_script>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|im_start|>",
+    "<|im_end|>",
+    "<repo_name>",
+    "<reponame>",
+    "<file_sep>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<jupyter_script>",
+    "<empty_output>"
+  ],
+  "bos_token": "<|im_start|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/README.md

@@ -0,0 +1,28 @@
+---
+license: apache-2.0
+datasets:
+- amphion/Emilia-Dataset
+language:
+- en
+- fr
+- de
+library_name: transformers
+tags:
+- mlx
+- mlx-audio
+- transformers
+- mlx
+---
+
+# Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit
+This model was converted to MLX format from [`Marvis-AI/marvis-tts-250m-v0.2`](https://huggingface.co/Marvis-AI/marvis-tts-250m-v0.2) using mlx-audio version **0.2.6**.
+Refer to the [original model card](https://huggingface.co/Marvis-AI/marvis-tts-250m-v0.2) for more details on the model.
+## Use with mlx
+
+```bash
+pip install -U mlx-audio
+```
+
+```bash
+python -m mlx_audio.tts.generate --model Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit --text "Describe this image."
+```

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/config.json

@@ -0,0 +1,307 @@
+{
+    "add_cross_attention": false,
+    "architectures": [
+        "CsmForConditionalGeneration"
+    ],
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "audio_eos_token_id": 128003,
+    "audio_num_codebooks": 32,
+    "audio_token_id": 128002,
+    "audio_vocab_size": 2051,
+    "backbone_flavor": "llama-250M",
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": 128000,
+    "chunk_size_feed_forward": 0,
+    "codebook_eos_token_id": 0,
+    "codebook_pad_token_id": 2050,
+    "codec_config": {
+        "sampling_rate": 24000,
+        "audio_channels": 1,
+        "hidden_size": 512,
+        "num_filters": 64,
+        "num_residual_layers": 1,
+        "upsampling_ratios": [
+            8,
+            6,
+            5,
+            4
+        ],
+        "kernel_size": 7,
+        "last_kernel_size": 3,
+        "residual_kernel_size": 3,
+        "dilation_growth_rate": 2,
+        "use_causal_conv": true,
+        "pad_mode": "constant",
+        "compress": 2,
+        "trim_right_ratio": 1.0,
+        "codebook_size": 2048,
+        "codebook_dim": 256,
+        "num_quantizers": 32,
+        "use_conv_shortcut": false,
+        "vector_quantization_hidden_dimension": 256,
+        "upsample_groups": 512,
+        "num_hidden_layers": 8,
+        "intermediate_size": 2048,
+        "num_attention_heads": 8,
+        "num_key_value_heads": 8,
+        "hidden_act": "gelu",
+        "max_position_embeddings": 8000,
+        "initializer_range": 0.02,
+        "norm_eps": 1e-05,
+        "use_cache": false,
+        "use_streaming": false,
+        "rope_theta": 10000.0,
+        "sliding_window": 250,
+        "attention_dropout": 0.0,
+        "head_dim": 64,
+        "layer_scale_initial_scale": 0.01,
+        "attention_bias": false,
+        "_frame_rate": 12.5,
+        "num_semantic_quantizers": 1,
+        "return_dict": true,
+        "output_hidden_states": false,
+        "torchscript": false,
+        "dtype": "float32",
+        "pruned_heads": {},
+        "tie_word_embeddings": true,
+        "chunk_size_feed_forward": 0,
+        "is_encoder_decoder": false,
+        "is_decoder": false,
+        "cross_attention_hidden_size": null,
+        "add_cross_attention": false,
+        "tie_encoder_decoder": false,
+        "architectures": [
+            "MimiModel"
+        ],
+        "finetuning_task": null,
+        "id2label": {
+            "0": "LABEL_0",
+            "1": "LABEL_1"
+        },
+        "label2id": {
+            "LABEL_0": 0,
+            "LABEL_1": 1
+        },
+        "task_specific_params": null,
+        "problem_type": null,
+        "tokenizer_class": null,
+        "prefix": null,
+        "bos_token_id": null,
+        "pad_token_id": null,
+        "eos_token_id": null,
+        "sep_token_id": null,
+        "decoder_start_token_id": null,
+        "max_length": 20,
+        "min_length": 0,
+        "do_sample": false,
+        "early_stopping": false,
+        "num_beams": 1,
+        "temperature": 1.0,
+        "top_k": 50,
+        "top_p": 1.0,
+        "typical_p": 1.0,
+        "repetition_penalty": 1.0,
+        "length_penalty": 1.0,
+        "no_repeat_ngram_size": 0,
+        "encoder_no_repeat_ngram_size": 0,
+        "bad_words_ids": null,
+        "num_return_sequences": 1,
+        "output_scores": false,
+        "return_dict_in_generate": false,
+        "forced_bos_token_id": null,
+        "forced_eos_token_id": null,
+        "remove_invalid_values": false,
+        "exponential_decay_length_penalty": null,
+        "suppress_tokens": null,
+        "begin_suppress_tokens": null,
+        "num_beam_groups": 1,
+        "diversity_penalty": 0.0,
+        "_name_or_path": "kyutai/mimi",
+        "normalize": false,
+        "tf_legacy_loss": false,
+        "use_bfloat16": false,
+        "model_type": "mimi",
+        "output_attentions": false
+    },
+    "cross_attention_hidden_size": null,
+    "decoder_flavor": "llama-60M",
+    "decoder_start_token_id": null,
+    "depth_decoder_config": {
+        "return_dict": true,
+        "output_hidden_states": false,
+        "torchscript": false,
+        "dtype": null,
+        "pruned_heads": {},
+        "tie_word_embeddings": false,
+        "chunk_size_feed_forward": 0,
+        "is_encoder_decoder": false,
+        "is_decoder": false,
+        "cross_attention_hidden_size": null,
+        "add_cross_attention": false,
+        "tie_encoder_decoder": false,
+        "architectures": null,
+        "finetuning_task": null,
+        "id2label": {
+            "0": "LABEL_0",
+            "1": "LABEL_1"
+        },
+        "label2id": {
+            "LABEL_0": 0,
+            "LABEL_1": 1
+        },
+        "task_specific_params": null,
+        "problem_type": null,
+        "tokenizer_class": null,
+        "prefix": null,
+        "bos_token_id": null,
+        "pad_token_id": null,
+        "eos_token_id": null,
+        "sep_token_id": null,
+        "decoder_start_token_id": null,
+        "max_length": 20,
+        "min_length": 0,
+        "do_sample": false,
+        "early_stopping": false,
+        "num_beams": 1,
+        "temperature": 1.0,
+        "top_k": 50,
+        "top_p": 1.0,
+        "typical_p": 1.0,
+        "repetition_penalty": 1.0,
+        "length_penalty": 1.0,
+        "no_repeat_ngram_size": 0,
+        "encoder_no_repeat_ngram_size": 0,
+        "bad_words_ids": null,
+        "num_return_sequences": 1,
+        "output_scores": false,
+        "return_dict_in_generate": false,
+        "forced_bos_token_id": null,
+        "forced_eos_token_id": null,
+        "remove_invalid_values": false,
+        "exponential_decay_length_penalty": null,
+        "suppress_tokens": null,
+        "begin_suppress_tokens": null,
+        "num_beam_groups": 1,
+        "diversity_penalty": 0.0,
+        "_name_or_path": "",
+        "model_type": "csm_depth_decoder_model",
+        "tf_legacy_loss": false,
+        "use_bfloat16": false,
+        "num_codebooks": 32,
+        "vocab_size": 2051,
+        "backbone_hidden_size": 1536,
+        "max_position_embeddings": 33,
+        "hidden_size": 1024,
+        "intermediate_size": 4096,
+        "num_hidden_layers": 4,
+        "num_attention_heads": 8,
+        "num_key_value_heads": 2,
+        "hidden_act": "silu",
+        "initializer_range": 0.02,
+        "rms_norm_eps": 1e-05,
+        "use_cache": true,
+        "rope_theta": 500000,
+        "rope_scaling": {
+            "factor": 32.0,
+            "high_freq_factor": 0.0078125,
+            "low_freq_factor": 0.001953125,
+            "original_max_position_embeddings": 16,
+            "rope_type": "llama3"
+        },
+        "attention_bias": false,
+        "attention_dropout": 0.0,
+        "mlp_bias": false,
+        "head_dim": 128,
+        "output_attentions": false
+    },
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "dtype": null,
+    "early_stopping": false,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "head_dim": 128,
+    "hidden_act": "silu",
+    "hidden_size": 1536,
+    "id2label": {
+        "0": "LABEL_0",
+        "1": "LABEL_1"
+    },
+    "initializer_range": 0.02,
+    "intermediate_size": 8192,
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+        "LABEL_0": 0,
+        "LABEL_1": 1
+    },
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "max_position_embeddings": 2048,
+    "min_length": 0,
+    "mlp_bias": false,
+    "model_type": "csm",
+    "no_repeat_ngram_size": 0,
+    "num_attention_heads": 12,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_codebooks": 32,
+    "num_hidden_layers": 6,
+    "num_key_value_heads": 3,
+    "num_return_sequences": 1,
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": 128002,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "quantization": {
+        "group_size": 64,
+        "bits": 8,
+        "mode": "affine"
+    },
+    "quantization_config": {
+        "group_size": 64,
+        "bits": 8,
+        "mode": "affine"
+    },
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "rms_norm_eps": 1e-05,
+    "rope_scaling": {
+        "factor": 32.0,
+        "high_freq_factor": 0.5,
+        "low_freq_factor": 0.125,
+        "original_max_position_embeddings": 1024,
+        "rope_type": "llama3"
+    },
+    "rope_theta": 500000,
+    "sep_token_id": null,
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "text_tokenizer": "Marvis-AI/marvis-tts-250m-v0.2",
+    "text_vocab_size": 49152,
+    "tf_legacy_loss": false,
+    "tie_codebooks_embeddings": true,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": false,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torchscript": false,
+    "transformers_version": "4.57.1",
+    "typical_p": 1.0,
+    "use_bfloat16": false,
+    "use_cache": true,
+    "vocab_size": 2051
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/generation_config.json

@@ -0,0 +1,12 @@
+{
+  "bos_token_id": 1,
+  "depth_decoder__from_model_config": true,
+  "depth_decoder_do_sample": true,
+  "depth_decoder_temperature": 0.9,
+  "depth_decoder_top_k": 50,
+  "do_sample": true,
+  "max_new_tokens": 125,
+  "pad_token_id": 0,
+  "temperature": 0.9,
+  "transformers_version": "4.54.1"
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/model.safetensors.index.json

@@ -0,0 +1,253 @@
+{
+    "metadata": {
+        "total_size": 665486112,
+        "total_parameters": 599911265
+    },
+    "weight_map": {
+        "model.audio_embeddings.biases": "model.safetensors",
+        "model.audio_embeddings.scales": "model.safetensors",
+        "model.audio_embeddings.weight": "model.safetensors",
+        "model.audio_head": "model.safetensors",
+        "model.backbone.layers.0.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.4.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.input_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.down_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.gate_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.mlp.up_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.post_attention_layernorm.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.k_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.o_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.q_proj.weight": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.biases": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.scales": "model.safetensors",
+        "model.backbone.layers.5.self_attn.v_proj.weight": "model.safetensors",
+        "model.backbone.norm.weight": "model.safetensors",
+        "model.codebook0_head.biases": "model.safetensors",
+        "model.codebook0_head.scales": "model.safetensors",
+        "model.codebook0_head.weight": "model.safetensors",
+        "model.decoder.layers.0.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.0.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.1.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.2.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.input_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.down_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.gate_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.mlp.up_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.post_attention_layernorm.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.k_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.o_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.q_proj.weight": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.biases": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.scales": "model.safetensors",
+        "model.decoder.layers.3.self_attn.v_proj.weight": "model.safetensors",
+        "model.decoder.norm.weight": "model.safetensors",
+        "model.projection.biases": "model.safetensors",
+        "model.projection.scales": "model.safetensors",
+        "model.projection.weight": "model.safetensors",
+        "model.text_embeddings.biases": "model.safetensors",
+        "model.text_embeddings.scales": "model.safetensors",
+        "model.text_embeddings.weight": "model.safetensors"
+    }
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/prompts/conversational_a.txt

@@ -0,0 +1 @@
+Coffee's story likely begins in Ethiopia, where legend tells of a goat herder named Kaldi who noticed his goats became energetic after eating red berries from a particular bush. Curious, he tried them himself, and felt invigorated.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/prompts/conversational_b.txt

@@ -0,0 +1 @@
+Coffee's story likely begins in Ethiopia, where legend tells of a goat herder named Kaldi who noticed his goats became energetic after eating red berries from a particular bush. Curious, he tried them himself, and felt invigorated.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/prompts/conversational_de.txt

@@ -0,0 +1 @@
+Ich werde bald mehrere Sprachen sprechen, darunter auch Deutsch.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/prompts/conversational_fr.txt

@@ -0,0 +1 @@
+J'ai toujours cru qu’un étrange voleur joyeux mangeait rapidement cinq gâteaux exquis près du vieux phare.

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|im_start|>",
+    "<|im_end|>",
+    "<repo_name>",
+    "<reponame>",
+    "<file_sep>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<jupyter_script>",
+    "<empty_output>"
+  ],
+  "bos_token": "<|im_start|>",
+  "eos_token": "<|im_end|>",
+  "unk_token": {
+    "content": "<|endoftext|>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

voice/mlx/Marvis-AI/marvis-tts-250m-v0.2-MLX-8bit/tokenizer_config.json

@@ -0,0 +1,168 @@
+{
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<repo_name>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "4": {
+      "content": "<reponame>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "5": {
+      "content": "<file_sep>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "6": {
+      "content": "<filename>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "7": {
+      "content": "<gh_stars>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "8": {
+      "content": "<issue_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "9": {
+      "content": "<issue_comment>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "10": {
+      "content": "<issue_closed>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "11": {
+      "content": "<jupyter_start>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "12": {
+      "content": "<jupyter_text>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "13": {
+      "content": "<jupyter_code>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "14": {
+      "content": "<jupyter_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "15": {
+      "content": "<jupyter_script>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "16": {
+      "content": "<empty_output>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "additional_special_tokens": [
+    "<|endoftext|>",
+    "<|im_start|>",
+    "<|im_end|>",
+    "<repo_name>",
+    "<reponame>",
+    "<file_sep>",
+    "<filename>",
+    "<gh_stars>",
+    "<issue_start>",
+    "<issue_comment>",
+    "<issue_closed>",
+    "<jupyter_start>",
+    "<jupyter_text>",
+    "<jupyter_code>",
+    "<jupyter_output>",
+    "<jupyter_script>",
+    "<empty_output>"
+  ],
+  "bos_token": "<|im_start|>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "extra_special_tokens": {},
+  "model_max_length": 8192,
+  "tokenizer_class": "GPT2Tokenizer",
+  "unk_token": "<|endoftext|>",
+  "vocab_size": 49152
+}