Upload sliced model checkpoint

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,501 @@
+---
+license: gemma
+library_name: transformers
+pipeline_tag: image-text-to-text
+extra_gated_button_content: Acknowledge license
+base_model: google/gemma-3n-E2B
+tags:
+- automatic-speech-recognition
+- automatic-speech-translation
+- audio-text-to-text
+- video-text-to-text
+- matformer
+---
+
+> [!Note]
+> This is a submodel derived from `google/gemma-3n-E2B`. It has been modified by slicing specific layers and resizing FFN dimensions. It is not the original model.
+> To learn more about MatFormers, please review the [launch blog](https://developers.googleblog.com/en/introducing-gemma-3n-developer-guide) and generate your own submodels
+with the [MatFormer Lab](https://goo.gle/gemma3n-matformer-lab).
+>
+
+Skipped layers: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
+
+FFN hidden dimensions: [8192, 8192, 8192]
+
+
+
+> [!Note]
+> This repository corresponds to the launch version of Gemma 3n E2B, to be used with Hugging Face `transformers`,
+> supporting text, audio, and vision (image and video) inputs.
+> 
+> Gemma 3n models have multiple architecture innovations:
+>  * They are available in two sizes based on [effective parameters](https://ai.google.dev/gemma/docs/gemma-3n#parameters). While the raw parameter count of this model is 6B, the architecture design allows the model to be run with a memory footprint comparable to a traditional 2B model by offloading low-utilization matrices from the accelerator.
+>  * They use a MatFormer architecture that allows nesting sub-models within the [E4B model](https://huggingface.co/google/gemma-3n-E4B). We provide one sub-model (this model repository), or you can access a spectrum of custom-sized models using the [Mix-and-Match method](https://goo.gle/gemma3n-matformer-lab).
+>
+> Learn more about these techniques in the [technical blog post](https://developers.googleblog.com/en/introducing-gemma-3n-developer-guide)
+> and the [Gemma documentation](https://ai.google.dev/gemma/docs/gemma-3n). 
+
+
+# Gemma 3n model card
+
+**Model Page**: [Gemma 3n](https://ai.google.dev/gemma/docs/gemma-3n)
+
+**Resources and Technical Documentation**:
+
+-   [Responsible Generative AI Toolkit](https://ai.google.dev/responsible)
+-   [Gemma on Kaggle](https://www.kaggle.com/models/google/gemma-3n)
+-   [Gemma on HuggingFace](https://huggingface.co/collections/google/gemma-3n-685065323f5984ef315c93f4)
+-   [Gemma on Vertex Model Garden](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3n)
+
+**Terms of Use**: [Terms](https://ai.google.dev/gemma/terms)\
+**Authors**: Google DeepMind
+
+## Model Information
+
+Summary description and brief definition of inputs and outputs.
+
+### Description
+
+Gemma is a family of lightweight, state-of-the-art open models from Google,
+built from the same research and technology used to create the Gemini models.
+Gemma 3n models are designed for efficient execution on low-resource devices.
+They are capable of multimodal input, handling text, image, video, and audio
+input, and generating text outputs, with open weights for pre-trained and
+instruction-tuned variants. These models were trained with data in over 140
+spoken languages.
+
+Gemma 3n models use selective parameter activation technology to reduce resource
+requirements. This technique allows the models to operate at an effective size
+of 2B and 4B parameters, which is lower than the total number of parameters they
+contain. For more information on Gemma 3n's efficient parameter management
+technology, see the
+[Gemma 3n](https://ai.google.dev/gemma/docs/gemma-3n#parameters)
+page.
+
+### Inputs and outputs
+
+-   **Input:**
+    -   Text string, such as a question, a prompt, or a document to be
+        summarized
+    -   Images, normalized to 256x256, 512x512, or 768x768 resolution
+        and encoded to 256 tokens each
+    -   Audio data encoded to 6.25 tokens per second from a single channel
+    -   Total input context of 32K tokens
+-   **Output:**
+    -   Generated text in response to the input, such as an answer to a
+        question, analysis of image content, or a summary of a document
+    -   Total output length up to 32K tokens, subtracting the request
+        input tokens
+
+### Usage
+
+Below, there are some code snippets on how to get quickly started with running
+the model. First, install the Transformers library. Gemma 3n is supported
+starting from transformers 4.53.0.
+
+```sh
+$ pip install -U transformers
+```
+
+Then, copy the snippet from the section that is relevant for your use case.
+
+#### Running with the `pipeline` API
+
+You can initialize the model and processor for inference with `pipeline` as
+follows.
+
+```python
+from transformers import pipeline
+import torch
+
+pipe = pipeline(
+    "image-text-to-text",
+    model="google/gemma-3n-e2b",
+    device="cuda",
+    torch_dtype=torch.bfloat16,
+)
+output = pipe(
+    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg",
+    text="<image_soft_token> in this image, there is"
+)
+
+print(output)
+# [{'input_text': '<image_soft_token> in this image, there is',
+# 'generated_text': '<image_soft_token> in this image, there is a beautiful flower and a bee is sucking nectar and pollen from the flower.'}]
+```
+
+#### Running the model on a single GPU
+
+```python
+from transformers import AutoProcessor, Gemma3nForConditionalGeneration
+from PIL import Image
+import requests
+import torch
+
+model_id = "google/gemma-3n-e2b"
+
+model = Gemma3nForConditionalGeneration.from_pretrained(model_id, device="cuda", torch_dtype=torch.bfloat16,).eval()
+
+processor = AutoProcessor.from_pretrained(model_id)
+
+url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"
+image = Image.open(requests.get(url, stream=True).raw)
+prompt = "<image_soft_token> in this image, there is"
+model_inputs = processor(text=prompt, images=image, return_tensors="pt").to(model.device)
+
+input_len = model_inputs["input_ids"].shape[-1]
+
+with torch.inference_mode():
+    generation = model.generate(**model_inputs, max_new_tokens=10)
+    generation = generation[0][input_len:]
+
+decoded = processor.decode(generation, skip_special_tokens=True)
+print(decoded)
+# one picture of flowers which shows that the flower is
+```
+
+### Citation
+
+```
+@article{gemma_3n_2025,
+    title={Gemma 3n},
+    url={https://ai.google.dev/gemma/docs/gemma-3n},
+    publisher={Google DeepMind},
+    author={Gemma Team},
+    year={2025}
+}
+```
+
+## Model Data
+
+Data used for model training and how the data was processed.
+
+### Training Dataset
+
+These models were trained on a dataset that includes a wide variety of sources
+totalling approximately 11 trillion tokens. The knowledge cutoff date for the
+training data was June 2024. Here are the key components:
+
+-   **Web Documents**: A diverse collection of web text ensures the model
+    is exposed to a broad range of linguistic styles, topics, and vocabulary.
+    The training dataset includes content in over 140 languages.
+-   **Code**: Exposing the model to code helps it to learn the syntax and
+    patterns of programming languages, which improves its ability to generate
+    code and understand code-related questions.
+-   **Mathematics**: Training on mathematical text helps the model learn
+    logical reasoning, symbolic representation, and to address mathematical queries.
+-   **Images**: A wide range of images enables the model to perform image
+    analysis and visual data extraction tasks.
+-   Audio: A diverse set of sound samples enables the model to recognize
+    speech, transcribe text from recordings, and identify information in audio data.
+
+The combination of these diverse data sources is crucial for training a
+powerful multimodal model that can handle a wide variety of different tasks and
+data formats.
+
+### Data Preprocessing
+
+Here are the key data cleaning and filtering methods applied to the training
+data:
+
+-   **CSAM Filtering**: Rigorous CSAM (Child Sexual Abuse Material)
+    filtering was applied at multiple stages in the data preparation process to
+    ensure the exclusion of harmful and illegal content.
+-   **Sensitive Data Filtering**: As part of making Gemma pre-trained models
+    safe and reliable, automated techniques were used to filter out certain
+    personal information and other sensitive data from training sets.
+-   **Additional methods**: Filtering based on content quality and safety in
+    line with
+    [our policies](https://ai.google/static/documents/ai-responsibility-update-published-february-2025.pdf).
+
+## Implementation Information
+
+Details about the model internals.
+
+### Hardware
+
+Gemma was trained using [Tensor Processing Unit
+(TPU)](https://cloud.google.com/tpu/docs/intro-to-tpu) hardware (TPUv4p, TPUv5p
+and TPUv5e). Training generative models requires significant computational
+power. TPUs, designed specifically for matrix operations common in machine
+learning, offer several advantages in this domain:
+
+-   **Performance**: TPUs are specifically designed to handle the massive
+    computations involved in training generative models. They can speed up
+    training considerably compared to CPUs.
+-   **Memory**: TPUs often come with large amounts of high-bandwidth memory,
+    allowing for the handling of large models and batch sizes during training.
+    This can lead to better model quality.
+-   **Scalability**: TPU Pods (large clusters of TPUs) provide a scalable
+    solution for handling the growing complexity of large foundation models.
+    You can distribute training across multiple TPU devices for faster and more
+    efficient processing.
+-   **Cost-effectiveness**: In many scenarios, TPUs can provide a more
+    cost-effective solution for training large models compared to CPU-based
+    infrastructure, especially when considering the time and resources saved
+    due to faster training.
+
+These advantages are aligned with
+[Google's commitments to operate sustainably](https://sustainability.google/operating-sustainably/).
+
+### Software
+
+Training was done using [JAX](https://github.com/jax-ml/jax) and
+[ML Pathways](https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture/).
+JAX allows researchers to take advantage of the latest generation of hardware,
+including TPUs, for faster and more efficient training of large models. ML
+Pathways is Google's latest effort to build artificially intelligent systems
+capable of generalizing across multiple tasks. This is specially suitable for
+foundation models, including large language models like these ones.
+
+Together, JAX and ML Pathways are used as described in the
+[paper about the Gemini family of models](https://goo.gle/gemma2report):
+*"the 'single controller' programming model of Jax and Pathways allows a single
+Python process to orchestrate the entire training run, dramatically simplifying
+the development workflow."*
+
+## Evaluation
+
+Model evaluation metrics and results.
+
+### Benchmark Results
+
+These models were evaluated at full precision (float32) against a large
+collection of different datasets and metrics to cover different aspects of
+content generation. Evaluation results marked with **IT** are for
+instruction-tuned models. Evaluation results marked with **PT** are for
+pre-trained models.
+
+#### Reasoning and factuality
+
+| Benchmark                      | Metric         | n-shot   |  E2B PT  |  E4B PT  |
+| ------------------------------ |----------------|----------|:--------:|:--------:|
+| [HellaSwag][hellaswag]         | Accuracy       | 10-shot  |   72.2   |   78.6   |
+| [BoolQ][boolq]                 | Accuracy       | 0-shot   |   76.4   |   81.6   |
+| [PIQA][piqa]                   | Accuracy       | 0-shot   |   78.9   |   81.0   |
+| [SocialIQA][socialiqa]         | Accuracy       | 0-shot   |   48.8   |   50.0   |
+| [TriviaQA][triviaqa]           | Accuracy       | 5-shot   |   60.8   |   70.2   |
+| [Natural Questions][naturalq]  | Accuracy       | 5-shot   |   15.5   |   20.9   |
+| [ARC-c][arc]                   | Accuracy       | 25-shot  |   51.7   |   61.6   |
+| [ARC-e][arc]                   | Accuracy       | 0-shot   |   75.8   |   81.6   |
+| [WinoGrande][winogrande]       | Accuracy       | 5-shot   |   66.8   |   71.7   |
+| [BIG-Bench Hard][bbh]          | Accuracy       | few-shot |   44.3   |   52.9   |
+| [DROP][drop]                   | Token F1 score | 1-shot   |   53.9   |   60.8   |
+
+[hellaswag]: https://arxiv.org/abs/1905.07830
+[boolq]: https://arxiv.org/abs/1905.10044
+[piqa]: https://arxiv.org/abs/1911.11641
+[socialiqa]: https://arxiv.org/abs/1904.09728
+[triviaqa]: https://arxiv.org/abs/1705.03551
+[naturalq]: https://github.com/google-research-datasets/natural-questions
+[arc]: https://arxiv.org/abs/1911.01547
+[winogrande]: https://arxiv.org/abs/1907.10641
+[bbh]: https://paperswithcode.com/dataset/bbh
+[drop]: https://arxiv.org/abs/1903.00161
+
+#### Multilingual
+
+| Benchmark                           | Metric                  | n-shot   |  E2B IT  |  E4B IT  |
+| ------------------------------------|-------------------------|----------|:--------:|:--------:|
+| [MGSM][mgsm]                        | Accuracy                |  0-shot  |   53.1   |   60.7   |
+| [WMT24++][wmt24pp] (ChrF)           | Character-level F-score |  0-shot  |   42.7   |   50.1   |
+| [Include][include]                  | Accuracy                |  0-shot  |   38.6   |   57.2   |
+| [MMLU][mmlu] (ProX)                 | Accuracy                |  0-shot  |    8.1   |   19.9   |
+| [OpenAI MMLU][openai-mmlu]          | Accuracy                |  0-shot  |   22.3   |   35.6   |
+| [Global-MMLU][global-mmlu]          | Accuracy                |  0-shot  |   55.1   |   60.3   |
+| [ECLeKTic][eclektic]                | ECLeKTic score          |  0-shot  |    2.5   |    1.9   |
+
+[mgsm]: https://arxiv.org/abs/2210.03057
+[wmt24pp]: https://arxiv.org/abs/2502.12404v1
+[include]:https://arxiv.org/abs/2411.19799
+[mmlu]: https://arxiv.org/abs/2009.03300
+[openai-mmlu]: https://huggingface.co/datasets/openai/MMMLU
+[global-mmlu]: https://huggingface.co/datasets/CohereLabs/Global-MMLU
+[eclektic]: https://arxiv.org/abs/2502.21228
+
+#### STEM and code
+
+| Benchmark                           | Metric                   | n-shot   |  E2B IT  |  E4B IT  |
+| ------------------------------------|--------------------------|----------|:--------:|:--------:|
+| [GPQA][gpqa] Diamond                | RelaxedAccuracy/accuracy |  0-shot  |   24.8   |   23.7   |
+| [LiveCodeBench][lcb] v5             | pass@1                   |  0-shot  |   18.6   |   25.7   |
+| Codegolf v2.2                       | pass@1                   |  0-shot  |   11.0   |   16.8   |
+| [AIME 2025][aime-2025]              | Accuracy                 |  0-shot  |    6.7   |   11.6   |
+
+[gpqa]: https://arxiv.org/abs/2311.12022
+[lcb]: https://arxiv.org/abs/2403.07974
+[aime-2025]: https://www.vals.ai/benchmarks/aime-2025-05-09
+
+#### Additional benchmarks
+
+| Benchmark                            | Metric     | n-shot   |  E2B IT  |  E4B IT  |
+| ------------------------------------ |------------|----------|:--------:|:--------:|
+| [MMLU][mmlu]                         |  Accuracy  |  0-shot  |   60.1   |   64.9   |
+| [MBPP][mbpp]                         |  pass@1    |  3-shot  |   56.6   |   63.6   |
+| [HumanEval][humaneval]               |  pass@1    |  0-shot  |   66.5   |   75.0   |
+| [LiveCodeBench][lcb]                 |  pass@1    |  0-shot  |   13.2   |   13.2   |
+| HiddenMath                           |  Accuracy  |  0-shot  |   27.7   |   37.7   |
+| [Global-MMLU-Lite][global-mmlu-lite] |  Accuracy  |  0-shot  |   59.0   |   64.5   |
+| [MMLU][mmlu] (Pro)                   |  Accuracy  |  0-shot  |   40.5   |   50.6   |
+
+[gpqa]: https://arxiv.org/abs/2311.12022
+[mbpp]: https://arxiv.org/abs/2108.07732
+[humaneval]: https://arxiv.org/abs/2107.03374
+[lcb]: https://arxiv.org/abs/2403.07974
+[global-mmlu-lite]: https://huggingface.co/datasets/CohereForAI/Global-MMLU-Lite
+
+## Ethics and Safety
+
+Ethics and safety evaluation approach and results.
+
+### Evaluation Approach
+
+Our evaluation methods include structured evaluations and internal red-teaming
+testing of relevant content policies. Red-teaming was conducted by a number of
+different teams, each with different goals and human evaluation metrics. These
+models were evaluated against a number of different categories relevant to
+ethics and safety, including:
+
+-   **Child Safety**: Evaluation of text-to-text and image to text prompts
+    covering child safety policies, including child sexual abuse and
+    exploitation.
+-   **Content Safety:** Evaluation of text-to-text and image to text prompts
+    covering safety policies including, harassment, violence and gore, and hate
+    speech.
+-   **Representational Harms**: Evaluation of text-to-text and image to text
+    prompts covering safety policies including bias, stereotyping, and harmful
+    associations or inaccuracies.
+
+In addition to development level evaluations, we conduct "assurance
+evaluations" which are our 'arms-length' internal evaluations for responsibility
+governance decision making. They are conducted separately from the model
+development team, to inform decision making about release. High level findings
+are fed back to the model team, but prompt sets are held-out to prevent
+overfitting and preserve the results' ability to inform decision making. Notable
+assurance evaluation results are reported to our Responsibility & Safety Council
+as part of release review.
+
+### Evaluation Results
+
+For all areas of safety testing, we saw safe levels of performance across the
+categories of child safety, content safety, and representational harms relative
+to previous Gemma models. All testing was conducted without safety filters to
+evaluate the model capabilities and behaviors. For text-to-text,  image-to-text,
+and audio-to-text, and across all model sizes, the model produced minimal policy
+violations, and showed significant improvements over previous Gemma models'
+performance with respect to high severity violations. A limitation of our
+evaluations was they included primarily English language prompts.
+
+## Usage and Limitations
+
+These models have certain limitations that users should be aware of.
+
+### Intended Usage
+
+Open generative models have a wide range of applications across various
+industries and domains. The following list of potential uses is not
+comprehensive. The purpose of this list is to provide contextual information
+about the possible use-cases that the model creators considered as part of model
+training and development.
+
+-   Content Creation and Communication
+    -   **Text Generation**: Generate creative text formats such as
+        poems, scripts, code, marketing copy, and email drafts.
+    -   **Chatbots and Conversational AI**: Power conversational
+        interfaces for customer service, virtual assistants, or interactive
+        applications.
+    -   **Text Summarization**: Generate concise summaries of a text
+        corpus, research papers, or reports.
+    -   **Image Data Extraction**: Extract, interpret, and summarize
+        visual data for text communications.
+    -   **Audio Data Extraction**: Transcribe spoken language, translate speech
+        to text in other languages, and analyze sound-based data.
+-   Research and Education
+    -   **Natural Language Processing (NLP) and generative model
+        Research**: These models can serve as a foundation for researchers to
+        experiment with generative models and NLP techniques, develop
+        algorithms, and contribute to the advancement of the field.
+    -   **Language Learning Tools**: Support interactive language
+        learning experiences, aiding in grammar correction or providing writing
+        practice.
+    -   **Knowledge Exploration**: Assist researchers in exploring large
+        bodies of data by generating summaries or answering questions about
+        specific topics.
+
+### Limitations
+
+-   Training Data
+    -   The quality and diversity of the training data significantly
+        influence the model's capabilities. Biases or gaps in the training data
+        can lead to limitations in the model's responses.
+    -   The scope of the training dataset determines the subject areas
+        the model can handle effectively.
+-   Context and Task Complexity
+    -   Models are better at tasks that can be framed with clear
+        prompts and instructions. Open-ended or highly complex tasks might be
+        challenging.
+    -   A model's performance can be influenced by the amount of context
+        provided (longer context generally leads to better outputs, up to a
+        certain point).
+-   Language Ambiguity and Nuance
+    -   Natural language is inherently complex. Models might struggle
+        to grasp subtle nuances, sarcasm, or figurative language.
+-   Factual Accuracy
+    -   Models generate responses based on information they learned
+        from their training datasets, but they are not knowledge bases. They
+        may generate incorrect or outdated factual statements.
+-   Common Sense
+    -   Models rely on statistical patterns in language. They might
+        lack the ability to apply common sense reasoning in certain situations.
+
+### Ethical Considerations and Risks
+
+The development of generative models raises several ethical concerns. In
+creating an open model, we have carefully considered the following:
+
+-   Bias and Fairness
+    -   Generative models trained on large-scale, real-world text and image data
+        can reflect socio-cultural biases embedded in the training material.
+        These models underwent careful scrutiny, input data pre-processing
+        described and posterior evaluations reported in this card.
+-   Misinformation and Misuse
+    -   Generative models can be misused to generate text that is
+        false, misleading, or harmful.
+    -   Guidelines are provided for responsible use with the model, see the
+        [Responsible Generative AI Toolkit](https://ai.google.dev/responsible).
+-   Transparency and Accountability:
+    -   This model card summarizes details on the models' architecture,
+        capabilities, limitations, and evaluation processes.
+    -   A responsibly developed open model offers the opportunity to
+        share innovation by making generative model technology accessible to
+        developers and researchers across the AI ecosystem.
+
+Risks identified and mitigations:
+
+-   **Perpetuation of biases**: It's encouraged to perform continuous monitoring
+    (using evaluation metrics, human review) and the exploration of de-biasing
+    techniques during model training, fine-tuning, and other use cases.
+-   **Generation of harmful content**: Mechanisms and guidelines for content
+    safety are essential. Developers are encouraged to exercise caution and
+    implement appropriate content safety safeguards based on their specific
+    product policies and application use cases.
+-   **Misuse for malicious purposes**: Technical limitations and developer
+    and end-user education can help mitigate against malicious applications of
+    generative models. Educational resources and reporting mechanisms for users
+    to flag misuse are provided. Prohibited uses of Gemma models are outlined
+    in the
+    [Gemma Prohibited Use Policy](https://ai.google.dev/gemma/prohibited_use_policy).
+-   **Privacy violations**: Models were trained on data filtered for removal of
+    certain personal information and other sensitive data. Developers are
+    encouraged to adhere to privacy regulations with privacy-preserving
+    techniques.
+
+### Benefits
+
+At the time of release, this family of models provides high-performance open
+generative model implementations designed from the ground up for responsible AI
+development compared to similarly sized models.
+
+Using the benchmark evaluation metrics described in this document, these models
+have shown to provide superior performance to other, comparably-sized open model
+alternatives.

config.json

@@ -0,0 +1,150 @@
+{
+  "architectures": [
+    "Gemma3nForConditionalGeneration"
+  ],
+  "audio_config": {
+    "conf_attention_chunk_size": 12,
+    "conf_attention_context_left": 13,
+    "conf_attention_context_right": 0,
+    "conf_attention_logit_cap": 50.0,
+    "conf_conv_kernel_size": 5,
+    "conf_num_attention_heads": 8,
+    "conf_num_hidden_layers": 12,
+    "conf_reduction_factor": 4,
+    "conf_residual_weight": 0.5,
+    "dtype": "bfloat16",
+    "gradient_clipping": 10000000000.0,
+    "hidden_size": 1536,
+    "input_feat_size": 128,
+    "model_type": "gemma3n_audio",
+    "rms_norm_eps": 1e-06,
+    "sscp_conv_channel_size": [
+      128,
+      32
+    ],
+    "sscp_conv_group_norm_eps": 0.001,
+    "sscp_conv_kernel_size": [
+      [
+        3,
+        3
+      ],
+      [
+        3,
+        3
+      ]
+    ],
+    "sscp_conv_stride_size": [
+      [
+        2,
+        2
+      ],
+      [
+        2,
+        2
+      ]
+    ],
+    "vocab_offset": 262272,
+    "vocab_size": 128
+  },
+  "audio_soft_tokens_per_image": 188,
+  "audio_token_id": 262273,
+  "boa_token_id": 256000,
+  "boi_token_id": 255999,
+  "dtype": "bfloat16",
+  "eoa_token_id": 262272,
+  "eoi_token_id": 262144,
+  "image_token_id": 262145,
+  "initializer_range": 0.02,
+  "model_type": "gemma3n",
+  "text_config": {
+    "activation_sparsity_pattern": [
+      0.95,
+      0,
+      0
+    ],
+    "altup_active_idx": 0,
+    "altup_coef_clip": 120.0,
+    "altup_correct_scale": true,
+    "altup_num_inputs": 4,
+    "attention_bias": false,
+    "attention_dropout": 0.0,
+    "dtype": "bfloat16",
+    "final_logit_softcapping": 30.0,
+    "head_dim": 256,
+    "hidden_activation": "gelu_pytorch_tanh",
+    "hidden_size": 2048,
+    "hidden_size_per_layer_input": 256,
+    "initializer_range": 0.02,
+    "intermediate_size": [
+      8192,
+      8192,
+      8192
+    ],
+    "laurel_rank": 64,
+    "layer_types": [
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "sliding_attention",
+      "full_attention"
+    ],
+    "max_position_embeddings": 32768,
+    "model_type": "gemma3n_text",
+    "num_attention_heads": 8,
+    "num_hidden_layers": 3,
+    "num_key_value_heads": 2,
+    "num_kv_shared_layers": 0,
+    "rms_norm_eps": 1e-06,
+    "rope_local_base_freq": 10000.0,
+    "rope_scaling": null,
+    "rope_theta": 1000000.0,
+    "sliding_window": 512,
+    "use_cache": true,
+    "vocab_size": 262400,
+    "vocab_size_per_layer_input": 262144
+  },
+  "transformers_version": "4.57.1",
+  "vision_config": {
+    "architecture": "mobilenetv5_300m_enc",
+    "do_pooling": false,
+    "dtype": "bfloat16",
+    "hidden_size": 2048,
+    "initializer_range": 0.02,
+    "label_names": [
+      "LABEL_0",
+      "LABEL_1"
+    ],
+    "model_args": null,
+    "model_type": "gemma3n_vision",
+    "num_classes": 2,
+    "rms_norm_eps": 1e-06,
+    "vocab_offset": 262144,
+    "vocab_size": 128
+  },
+  "vision_soft_tokens_per_image": 256
+}

model-00001-of-00001.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:87979ba26147808043d0ac616d3e20a7f096e3ef05b2fe6f9137a1851808c59a
+size 3870532216

model.safetensors.index.json

@@ -0,0 +1,915 @@
+{
+  "metadata": {
+    "total_size": 3870532216
+  },
+  "weight_map": {
+    "model.audio_tower.conformer.0.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.0.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.1.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.10.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.11.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.2.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.3.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.4.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.5.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.6.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.7.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.8.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.attn.per_dim_scale": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.attn.relative_position_embedding.pos_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.post.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.post_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.attention.pre_attn_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_end.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_end.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_end.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_end.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_start.ffw_layer_1.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_start.ffw_layer_2.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_start.post_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.ffw_layer_start.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.lconv1d.conv_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.lconv1d.depthwise_conv1d.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.lconv1d.linear_end.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.lconv1d.linear_start.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.lconv1d.pre_layer_norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.conformer.9.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.subsample_conv_projection.conv_0.conv.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.subsample_conv_projection.conv_0.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.subsample_conv_projection.conv_1.conv.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.subsample_conv_projection.conv_1.norm.weight": "model-00001-of-00001.safetensors",
+    "model.audio_tower.subsample_conv_projection.input_proj_linear.weight": "model-00001-of-00001.safetensors",
+    "model.embed_audio.embedding.weight": "model-00001-of-00001.safetensors",
+    "model.embed_vision.embedding.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_projections.0.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_projections.1.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_projections.2.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_unembed_projections.0.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_unembed_projections.1.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.altup_unembed_projections.2.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.embed_tokens.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.0.bn1.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.0.bn2.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.0.conv_exp.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.0.conv_pwl.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.1.bn1.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.1.bn2.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.1.conv_exp.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.1.conv_pwl.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.2.bn1.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.2.bn2.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.2.conv_exp.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.0.2.conv_pwl.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.1.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.3.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.1.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.1.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.10.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.10.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.10.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.10.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.10.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.11.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.12.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.12.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.12.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.12.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.12.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.13.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.14.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.14.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.14.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.14.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.14.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.15.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.16.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.16.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.16.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.16.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.16.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.17.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.18.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.18.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.18.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.18.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.18.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.19.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.20.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.20.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.20.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.20.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.20.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.21.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.22.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.22.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.22.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.22.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.22.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.23.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.24.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.24.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.24.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.24.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.24.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.25.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.26.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.26.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.26.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.26.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.26.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.27.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.28.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.28.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.28.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.28.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.28.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.29.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.3.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.30.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.30.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.30.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.30.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.30.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.31.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.32.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.32.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.32.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.32.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.32.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.33.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.34.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.34.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.34.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.34.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.34.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.35.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.36.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.36.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.36.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.36.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.36.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.5.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.6.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.7.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.8.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.8.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.8.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.8.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.8.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.key.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.key.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.value.down_conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.value.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.2.9.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.dw_mid.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.dw_mid.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.dw_start.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.dw_start.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.0.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.1.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.10.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.10.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.10.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.10.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.10.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.11.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.12.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.12.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.12.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.12.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.12.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.13.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.14.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.14.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.14.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.14.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.14.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.15.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.16.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.16.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.16.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.16.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.16.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.17.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.18.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.18.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.18.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.18.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.18.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.19.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.2.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.2.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.2.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.2.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.2.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.20.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.20.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.20.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.20.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.20.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.21.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.22.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.22.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.22.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.22.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.22.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.23.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.24.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.24.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.24.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.24.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.24.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.25.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.26.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.26.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.26.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.26.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.26.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.27.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.28.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.28.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.28.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.28.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.28.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.29.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.3.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.30.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.30.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.30.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.30.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.30.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.31.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.32.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.32.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.32.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.32.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.32.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.33.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.34.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.34.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.34.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.34.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.34.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.35.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.36.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.36.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.36.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.36.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.36.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.37.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.38.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.38.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.38.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.38.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.38.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.4.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.4.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.4.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.4.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.4.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.5.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.6.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.6.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.6.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.6.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.6.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.7.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.8.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.8.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.8.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.8.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.8.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.attn.key.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.attn.output.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.attn.query.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.attn.value.proj.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.layer_scale.gamma": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.blocks.3.9.norm.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.conv_stem.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.conv_stem.conv.bias": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.conv_stem.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.msfa.ffn.pw_exp.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.msfa.ffn.pw_exp.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.msfa.ffn.pw_proj.bn.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.msfa.ffn.pw_proj.conv.weight": "model-00001-of-00001.safetensors",
+    "model.vision_tower.timm_model.msfa.norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.altup.correct_output_scale": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.altup.correction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.altup.modality_router.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.altup.router_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.input_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.laurel.linear_left.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.laurel.linear_right.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.mlp.down_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.mlp.gate_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.mlp.up_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.per_layer_input_gate.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.per_layer_projection.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.post_attention_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.k_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.o_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.q_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.1.self_attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.altup.correct_output_scale": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.altup.correction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.altup.modality_router.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.altup.router_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.input_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.laurel.linear_left.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.laurel.linear_right.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.mlp.down_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.mlp.gate_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.mlp.up_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.per_layer_input_gate.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.per_layer_projection.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.post_attention_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.k_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.o_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.q_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.2.self_attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.embed_audio.embedding_projection.weight": "model-00001-of-00001.safetensors",
+    "model.embed_audio.hard_embedding_norm.weight": "model-00001-of-00001.safetensors",
+    "model.embed_audio.soft_embedding_norm.weight": "model-00001-of-00001.safetensors",
+    "model.embed_vision.embedding_projection.weight": "model-00001-of-00001.safetensors",
+    "model.embed_vision.hard_embedding_norm.weight": "model-00001-of-00001.safetensors",
+    "model.embed_vision.soft_embedding_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.embed_tokens_per_layer.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.altup.correct_output_scale": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.altup.correction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.altup.modality_router.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.altup.prediction_coefs.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.altup.router_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.input_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.laurel.linear_left.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.laurel.linear_right.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.laurel.post_laurel_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.mlp.down_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.mlp.gate_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.mlp.up_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.per_layer_input_gate.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.per_layer_projection.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.post_attention_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.post_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.post_per_layer_input_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.pre_feedforward_layernorm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.k_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.k_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.o_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.q_norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.q_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.layers.0.self_attn.v_proj.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.norm.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.per_layer_model_projection.weight": "model-00001-of-00001.safetensors",
+    "model.language_model.per_layer_projection_norm.weight": "model-00001-of-00001.safetensors"
+  }
+}

special_tokens_map.json

@@ -0,0 +1,36 @@
+{
+  "audio_token": "<audio_soft_token>",
+  "boa_token": "<start_of_audio>",
+  "boi_token": "<start_of_image>",
+  "bos_token": {
+    "content": "<bos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "eoa_token": "<end_of_audio>",
+  "eoi_token": "<end_of_image>",
+  "eos_token": {
+    "content": "<eos>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "image_token": "<image_soft_token>",
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
+}

tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6c35ee648c07754b44cd9e371c75d4caa05c4504910b7ad29b1847ee9d8ba5d
+size 33442553

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea5f0cc48abfbfc04d14562270a32e02149a3e7035f368cc5a462786f4a59961
+size 4696020