transformers-integration (#2)

- transformer integration (2071a8f0ddf71be08528b088890f41d8ca06359b)


Co-authored-by: Eustache Le Bihan <[email protected]>

README.md

@@ -50,6 +50,79 @@ The predicted text token timestamps can be recovered by subtracting the model's
 
 See the [GitHub repository](https://github.com/kyutai-labs/delayed-streams-modeling/).
 
+### Use with transformers
+
+Install transformers from source:
+```bash
+pip install git+https://github.com/huggingface/transformers
+```
+
+Inference:
+```python
+import torch
+from datasets import load_dataset, Audio
+from transformers import KyutaiSpeechToTextProcessor, KyutaiSpeechToTextForConditionalGeneration
+
+# 1. load the model and the processor
+torch_device = "cuda" if torch.cuda.is_available() else "cpu"
+model_id = "kyutai/stt-2.6b-en"
+
+processor = KyutaiSpeechToTextProcessor.from_pretrained(model_id)
+model = KyutaiSpeechToTextForConditionalGeneration.from_pretrained(model_id, device_map=torch_device)
+
+# 2. load audio samples
+ds = load_dataset(
+    "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation"
+)
+ds = ds.cast_column("audio", Audio(sampling_rate=24000))
+
+# 3. prepare the model inputs
+inputs = processor(
+    ds[0]["audio"]["array"],
+)
+inputs.to(torch_device)
+
+# 4. infer the model
+output_tokens = model.generate(**inputs)
+
+# 5. decode the generated tokens
+print(processor.batch_decode(output_tokens, skip_special_tokens=True))
+```
+
+Batched inference:
+```python
+import torch
+from datasets import load_dataset, Audio
+from transformers import KyutaiSpeechToTextProcessor, KyutaiSpeechToTextForConditionalGeneration
+
+# 1. load the model and the processor
+torch_device = "cuda" if torch.cuda.is_available() else "cpu"
+model_id = "kyutai/stt-2.6b-en"
+
+processor = KyutaiSpeechToTextProcessor.from_pretrained(model_id)
+model = KyutaiSpeechToTextForConditionalGeneration.from_pretrained(model_id, device_map=torch_device)
+
+# 2. load audio samples
+ds = load_dataset(
+    "hf-internal-testing/librispeech_asr_dummy", "clean", split="validation"
+)
+ds = ds.cast_column("audio", Audio(sampling_rate=24000))
+
+# 3. prepare the model inputs
+audio_arrays = [ds[i]["audio"]["array"] for i in range(4)]
+inputs = processor(audio_arrays, return_tensors="pt", padding=True)
+inputs = inputs.to(torch_device)
+
+# 4. infer the model
+output_tokens = model.generate(**inputs)
+
+# 5. decode the generated tokens
+decoded_outputs = processor.batch_decode(output_tokens, skip_special_tokens=True)
+for output in decoded_outputs:
+    print(output)
+```
+####
+
 ## Training Details
 
 ### Training Data

config.json

@@ -75,5 +75,80 @@
   },
   "model_type": "stt",
   "mimi_name": "[email protected]",
-  "tokenizer_name": "tokenizer_en_audio_4000.model"
+  "tokenizer_name": "tokenizer_en_audio_4000.model",
+  "architectures": [
+    "KyutaiSpeechToTextForConditionalGeneration"
+  ],
+  "attention_dropout": 0.0,
+  "audio_bos_token_id": 2048,
+  "audio_pad_token_id": 69569,
+  "bos_token_id": 48000,
+  "codebook_vocab_size": 2049,
+  "codec_config": {
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "audio_channels": 1,
+    "codebook_dim": 256,
+    "codebook_size": 2048,
+    "compress": 2,
+    "dilation_growth_rate": 2,
+    "frame_rate": 12.5,
+    "head_dim": 64,
+    "hidden_act": "gelu",
+    "hidden_size": 512,
+    "initializer_range": 0.02,
+    "intermediate_size": 2048,
+    "kernel_size": 7,
+    "last_kernel_size": 3,
+    "layer_scale_initial_scale": 0.01,
+    "max_position_embeddings": 8000,
+    "model_type": "mimi",
+    "norm_eps": 1e-05,
+    "num_attention_heads": 8,
+    "num_filters": 64,
+    "num_hidden_layers": 8,
+    "num_key_value_heads": 8,
+    "num_quantizers": 32,
+    "num_residual_layers": 1,
+    "num_semantic_quantizers": 1,
+    "pad_mode": "constant",
+    "residual_kernel_size": 3,
+    "rope_theta": 10000.0,
+    "sampling_rate": 24000,
+    "sliding_window": 250,
+    "trim_right_ratio": 1.0,
+    "upsample_groups": 512,
+    "upsampling_ratios": [
+      8,
+      6,
+      5,
+      4
+    ],
+    "use_cache": false,
+    "use_causal_conv": true,
+    "use_conv_shortcut": false,
+    "use_streaming": false,
+    "vector_quantization_hidden_dimension": 256
+  },
+  "ffn_dim": 11264,
+  "frame_size": 1920,
+  "head_dim": 64,
+  "hidden_act": "silu",
+  "hidden_size": 2048,
+  "initializer_range": 0.02,
+  "max_position_embeddings": 750,
+  "num_attention_heads": 32,
+  "num_codebooks": 32,
+  "num_hidden_layers": 48,
+  "num_key_value_heads": 32,
+  "pad_token_id": 3,
+  "rms_norm_eps": 1e-08,
+  "rope_theta": 100000.0,
+  "sliding_window": 375,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.53.0.dev0",
+  "use_cache": true,
+  "vocab_size": 4001,
+  "transformers_weights": "transformers.safetensors.index.json"
 }

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "audio_window_size": 1,
+  "bos_token_id": 48000,
+  "cache_implementation": "sliding_window",
+  "codec_cache_implementation": "sliding_window",
+  "codec_use_cache": true,
+  "pad_token_id": 3,
+  "transformers_version": "4.53.0.dev0"
+}

preprocessor_config.json

@@ -0,0 +1,13 @@
+{
+  "audio_delay_seconds": 2.5,
+  "audio_silence_prefix_seconds": 1.0,
+  "chunk_length_s": null,
+  "feature_extractor_type": "KyutaiSpeechToTextFeatureExtractor",
+  "feature_size": 1,
+  "overlap": null,
+  "padding_side": "right",
+  "padding_value": 0.0,
+  "processor_class": "KyutaiSpeechToTextProcessor",
+  "return_attention_mask": true,
+  "sampling_rate": 24000
+}

special_tokens_map.json

@@ -0,0 +1,3 @@
+{
+  "unk_token": "<unk>"
+}

tokenizer_config.json

@@ -0,0 +1,50 @@
+{
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "3": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "bos_token_id": null,
+  "chat_template": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token_id": null,
+  "extra_special_tokens": {},
+  "model_input_names": [
+    "input_ids",
+    "attention_mask"
+  ],
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token_id": null,
+  "processor_class": "KyutaiSpeechToTextProcessor",
+  "tokenizer_class": "PreTrainedTokenizerFast",
+  "unk_token": "<unk>"
+}

transformers-00001-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d10da996924f8718e69ad188d65e4334f6c82a12bcaf8c5df83e80c9f5fca98b
+size 4372392344

transformers-00002-of-00002.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fe01b3882f795cc9198c6e7f1fa4f08000e14f59ac8b181a294535ddb41cc33b
+size 1547899924

transformers.safetensors.index.json

@@ -0,0 +1,745 @@
+{
+  "metadata": {
+    "total_parameters": 2696431425,
+    "total_size": 6759058820
+  },
+  "weight_map": {
+    "codec_model.decoder.layers.0.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.0.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.11.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.11.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.12.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.12.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.12.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.12.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.14.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.14.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.2.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.2.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.3.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.3.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.3.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.3.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.5.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.5.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.6.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.6.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.6.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.6.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.8.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.8.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.9.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.9.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.9.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder.layers.9.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.0.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.1.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.2.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.3.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.4.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.5.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.6.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.decoder_transformer.layers.7.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.downsample.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.0.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.0.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.1.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.1.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.1.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.1.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.10.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.10.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.10.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.10.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.12.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.12.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.14.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.14.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.4.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.4.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.4.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.4.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.6.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.6.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.7.block.1.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.7.block.1.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.7.block.3.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.7.block.3.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.9.conv.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder.layers.9.conv.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.0.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.1.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.2.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.3.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.4.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.5.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.6.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.input_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.mlp_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.post_attention_layernorm.bias": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.self_attn.k_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.self_attn.o_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.self_attn.q_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.self_attn.v_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.encoder_transformer.layers.7.self_attn_layer_scale.scale": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.input_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.0.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.0.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.0.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.1.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.1.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.1.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.10.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.10.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.10.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.11.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.11.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.11.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.12.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.12.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.12.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.13.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.13.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.13.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.14.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.14.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.14.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.15.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.15.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.15.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.16.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.16.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.16.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.17.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.17.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.17.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.18.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.18.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.18.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.19.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.19.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.19.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.2.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.2.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.2.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.20.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.20.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.20.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.21.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.21.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.21.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.22.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.22.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.22.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.23.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.23.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.23.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.24.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.24.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.24.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.25.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.25.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.25.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.26.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.26.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.26.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.27.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.27.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.27.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.28.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.28.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.28.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.29.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.29.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.29.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.3.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.3.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.3.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.30.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.30.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.30.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.4.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.4.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.4.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.5.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.5.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.5.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.6.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.6.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.6.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.7.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.7.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.7.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.8.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.8.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.8.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.9.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.9.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.layers.9.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.acoustic_residual_vector_quantizer.output_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.semantic_residual_vector_quantizer.input_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.semantic_residual_vector_quantizer.layers.0.codebook.cluster_usage": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.semantic_residual_vector_quantizer.layers.0.codebook.embed_sum": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.semantic_residual_vector_quantizer.layers.0.codebook.initialized": "transformers-00002-of-00002.safetensors",
+    "codec_model.quantizer.semantic_residual_vector_quantizer.output_proj.weight": "transformers-00002-of-00002.safetensors",
+    "codec_model.upsample.conv.weight": "transformers-00002-of-00002.safetensors",
+    "lm_head.weight": "transformers-00002-of-00002.safetensors",
+    "model.embed_tokens.embed_tokens.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.0.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.1.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.10.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.11.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.12.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.13.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.14.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.15.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.16.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.17.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.18.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.19.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.2.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.20.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.21.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.22.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.23.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.24.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.25.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.26.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.27.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.28.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.29.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.3.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.30.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.31.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.32.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.33.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.34.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.35.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.36.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.37.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.37.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.38.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.39.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.4.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.4.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.40.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.40.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.41.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.42.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.43.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.44.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.45.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.46.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.input_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.mlp.fc1.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.mlp.fc2.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.post_attention_layernorm.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.self_attn.k_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.self_attn.o_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.self_attn.q_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.47.self_attn.v_proj.linear.weight": "transformers-00002-of-00002.safetensors",
+    "model.layers.5.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.5.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.6.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.7.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.8.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.input_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.mlp.fc1.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.mlp.fc2.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.post_attention_layernorm.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.k_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.o_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.q_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.layers.9.self_attn.v_proj.linear.weight": "transformers-00001-of-00002.safetensors",
+    "model.norm.weight": "transformers-00002-of-00002.safetensors"
+  }
+}