Upload RobustRAG (sharded safetensors + tokenizer + config)
Browse files- README.md +179 -0
- config.json +45 -0
- generation_config.json +6 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +1070 -0
- special_tokens_map.json +44 -0
- tokenizer.model +3 -0
- tokenizer_config.json +161 -0
README.md
ADDED
|
@@ -0,0 +1,179 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
datasets:
|
| 4 |
+
- openbmb/VisRAG-Ret-Train-In-domain-data
|
| 5 |
+
- openbmb/VisRAG-Ret-Train-Synthetic-data
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
base_model:
|
| 9 |
+
- openbmb/MiniCPM-V-2
|
| 10 |
+
tags:
|
| 11 |
+
- VisRAG
|
| 12 |
+
pipeline_tag: feature-extraction
|
| 13 |
+
---
|
| 14 |
+
# VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents
|
| 15 |
+
<div style="display: flex; align-items: center;">
|
| 16 |
+
<a href="https://huggingface.co/openbmb/VisRAG-Ret" style="margin-right: 10px;">
|
| 17 |
+
<img src="https://img.shields.io/badge/VisRAG_Ret-fcd022?style=for-the-badge&logo=huggingface&logoColor=000" alt="VisRAG Ret">
|
| 18 |
+
</a>
|
| 19 |
+
<a href="https://huggingface.co/collections/openbmb/visrag-6717bbfb471bb018a49f1c69" style="margin-right: 10px;">
|
| 20 |
+
<img src="https://img.shields.io/badge/VisRAG_Collection-fcd022?style=for-the-badge&logo=huggingface&logoColor=000" alt="VisRAG Collection">
|
| 21 |
+
</a>
|
| 22 |
+
<a href="https://huggingface.co/spaces/tcy6/VisRAG_Pipeline" style="margin-right: 10px;">
|
| 23 |
+
<img src="https://img.shields.io/badge/VisRAG_Pipeline-fcd022?style=for-the-badge&logo=huggingface&logoColor=000" alt="VisRAG Pipeline">
|
| 24 |
+
</a>
|
| 25 |
+
<a href="https://arxiv.org/abs/2410.10594" style="margin-right: 10px;">
|
| 26 |
+
<img src="https://img.shields.io/badge/arXiv-2410.10594-ff0000.svg?style=for-the-badge" alt="arXiv">
|
| 27 |
+
</a>
|
| 28 |
+
<a href="https://colab.research.google.com/drive/11KV9adDNXPfHiuFAfXNOvtYJKcyR8JZH?usp=sharing" style="margin-right: 10px;">
|
| 29 |
+
<img src="https://img.shields.io/badge/VisRAG_Pipeline-ffffff?style=for-the-badge&logo=googlecolab&logoColor=f9ab00" alt="Google Colab">
|
| 30 |
+
</a>
|
| 31 |
+
<a href="https://github.com/openbmb/VisRAG" style="margin-right: 10px;">
|
| 32 |
+
<img src="https://img.shields.io/badge/VisRAG-000000?style=for-the-badge&logo=github&logoColor=white" alt="GitHub">
|
| 33 |
+
</a>
|
| 34 |
+
</div>
|
| 35 |
+
|
| 36 |
+
<p align="center">•
|
| 37 |
+
<a href="#📖-introduction"> 📖 Introduction </a> •
|
| 38 |
+
<a href="#🎉-news">🎉 News</a> •
|
| 39 |
+
<a href="#✨-visrag-pipeline">✨ VisRAG Pipeline</a> •
|
| 40 |
+
<a href="#⚡️-training">⚡️ Training</a>
|
| 41 |
+
</p>
|
| 42 |
+
<p align="center">•
|
| 43 |
+
<a href="#📦-requirements">📦 Requirements</a> •
|
| 44 |
+
<a href="#🔧-usage">🔧 Usage</a> •
|
| 45 |
+
<a href="#📄-license">📄 Lisense</a> •
|
| 46 |
+
<a href="#📑-citation">📑 Citation</a> •
|
| 47 |
+
<a href="#📧-contact">📧 Contact</a>
|
| 48 |
+
</p>
|
| 49 |
+
|
| 50 |
+
# 📖 Introduction
|
| 51 |
+
**VisRAG** is a novel vision-language model (VLM)-based RAG pipeline. In this pipeline, instead of first parsing the document to obtain text, the document is directly embedded using a VLM as an image and then retrieved to enhance the generation of a VLM.Compared to traditional text-based RAG, **VisRAG** maximizes the retention and utilization of the data information in the original documents, eliminating the information loss introduced during the parsing process.
|
| 52 |
+
<p align="center"><img width=800 src="https://github.com/openbmb/VisRAG/blob/master/assets/main_figure.png?raw=true"/></p>
|
| 53 |
+
|
| 54 |
+
# 🎉 News
|
| 55 |
+
|
| 56 |
+
* 20241104: Released our [VisRAG Pipeline](https://huggingface.co/spaces/tcy6/VisRAG_Pipeline) on Hugging Face Space.
|
| 57 |
+
* 20241031: Released our [VisRAG Pipeline](https://colab.research.google.com/drive/11KV9adDNXPfHiuFAfXNOvtYJKcyR8JZH?usp=sharing) on Colab.
|
| 58 |
+
* 20241015: Released our train data and test data on Hugging Face which can be found in the [VisRAG](https://huggingface.co/collections/openbmb/visrag-6717bbfb471bb018a49f1c69) Collection on Hugging Face. It is referenced at the beginning of this page.
|
| 59 |
+
* 20241014: Released our [Paper](https://arxiv.org/abs/2410.10594) on arXiv. Released our [Model](https://huggingface.co/openbmb/VisRAG-Ret) on Hugging Face. Released our [Code](https://github.com/OpenBMB/VisRAG) on GitHub.
|
| 60 |
+
|
| 61 |
+
# ✨ VisRAG Pipeline
|
| 62 |
+
|
| 63 |
+
## VisRAG-Ret
|
| 64 |
+
**VisRAG-Ret** is a document embedding model built on [MiniCPM-V 2.0](https://huggingface.co/openbmb/MiniCPM-V-2), a vision-language model that integrates [SigLIP](https://huggingface.co/google/siglip-so400m-patch14-384) as the vision encoder and [MiniCPM-2B](https://huggingface.co/openbmb/MiniCPM-2B-sft-bf16) as the language model.
|
| 65 |
+
|
| 66 |
+
## VisRAG-Gen
|
| 67 |
+
In the paper, We use MiniCPM-V 2.0, MiniCPM-V 2.6 and GPT-4o as the generators. Actually you can use any VLMs you like!
|
| 68 |
+
|
| 69 |
+
# ⚡️ Training
|
| 70 |
+
|
| 71 |
+
## VisRAG-Ret
|
| 72 |
+
Our training dataset of 362,110 Query-Document (Q-D) Pairs for **VisRAG-Ret** is comprised of train sets of openly available academic datasets (34%) and a synthetic dataset made up of pages from web-crawled PDF documents and augmented with VLM-generated (GPT-4o) pseudo-queries (66%). It can be found in the `VisRAG` Collection on Hugging Face, which is referenced at the beginning of this page.
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
## VisRAG-Gen
|
| 76 |
+
The generation part does not use any fine-tuning; we directly use off-the-shelf LLMs/VLMs for generation.
|
| 77 |
+
|
| 78 |
+
# 📦 Requirements
|
| 79 |
+
```
|
| 80 |
+
torch==2.1.2
|
| 81 |
+
torchvision==0.16.2
|
| 82 |
+
transformers==4.40.2
|
| 83 |
+
sentencepiece==0.1.99
|
| 84 |
+
decord==0.6.0
|
| 85 |
+
Pillow==10.1.0
|
| 86 |
+
```
|
| 87 |
+
|
| 88 |
+
# 🔧 Usage
|
| 89 |
+
|
| 90 |
+
## VisRAG-Ret
|
| 91 |
+
```python
|
| 92 |
+
from transformers import AutoModel, AutoTokenizer
|
| 93 |
+
import torch
|
| 94 |
+
import torch.nn.functional as F
|
| 95 |
+
from PIL import Image
|
| 96 |
+
import requests
|
| 97 |
+
from io import BytesIO
|
| 98 |
+
|
| 99 |
+
def weighted_mean_pooling(hidden, attention_mask):
|
| 100 |
+
attention_mask_ = attention_mask * attention_mask.cumsum(dim=1)
|
| 101 |
+
s = torch.sum(hidden * attention_mask_.unsqueeze(-1).float(), dim=1)
|
| 102 |
+
d = attention_mask_.sum(dim=1, keepdim=True).float()
|
| 103 |
+
reps = s / d
|
| 104 |
+
return reps
|
| 105 |
+
|
| 106 |
+
@torch.no_grad()
|
| 107 |
+
def encode(text_or_image_list):
|
| 108 |
+
|
| 109 |
+
if (isinstance(text_or_image_list[0], str)):
|
| 110 |
+
inputs = {
|
| 111 |
+
"text": text_or_image_list,
|
| 112 |
+
'image': [None] * len(text_or_image_list),
|
| 113 |
+
'tokenizer': tokenizer
|
| 114 |
+
}
|
| 115 |
+
else:
|
| 116 |
+
inputs = {
|
| 117 |
+
"text": [''] * len(text_or_image_list),
|
| 118 |
+
'image': text_or_image_list,
|
| 119 |
+
'tokenizer': tokenizer
|
| 120 |
+
}
|
| 121 |
+
outputs = model(**inputs)
|
| 122 |
+
attention_mask = outputs.attention_mask
|
| 123 |
+
hidden = outputs.last_hidden_state
|
| 124 |
+
|
| 125 |
+
reps = weighted_mean_pooling(hidden, attention_mask)
|
| 126 |
+
embeddings = F.normalize(reps, p=2, dim=1).detach().cpu().numpy()
|
| 127 |
+
return embeddings
|
| 128 |
+
|
| 129 |
+
model_name_or_path = "openbmb/VisRAG-Ret"
|
| 130 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=True)
|
| 131 |
+
model = AutoModel.from_pretrained(model_name_or_path, torch_dtype=torch.bfloat16, trust_remote_code=True).cuda()
|
| 132 |
+
model.eval()
|
| 133 |
+
|
| 134 |
+
queries = ["What does a dog look like?"]
|
| 135 |
+
INSTRUCTION = "Represent this query for retrieving relevant documents: "
|
| 136 |
+
queries = [INSTRUCTION + query for query in queries]
|
| 137 |
+
|
| 138 |
+
print("Downloading images...")
|
| 139 |
+
passages = [
|
| 140 |
+
Image.open(BytesIO(requests.get(
|
| 141 |
+
'https://github.com/OpenBMB/VisRAG/raw/refs/heads/master/scripts/demo/retriever/test_image/cat.jpeg'
|
| 142 |
+
).content)).convert('RGB'),
|
| 143 |
+
Image.open(BytesIO(requests.get(
|
| 144 |
+
'https://github.com/OpenBMB/VisRAG/raw/refs/heads/master/scripts/demo/retriever/test_image/dog.jpg'
|
| 145 |
+
).content)).convert('RGB')
|
| 146 |
+
]
|
| 147 |
+
print("Images downloaded.")
|
| 148 |
+
|
| 149 |
+
embeddings_query = encode(queries)
|
| 150 |
+
embeddings_doc = encode(passages)
|
| 151 |
+
|
| 152 |
+
scores = (embeddings_query @ embeddings_doc.T)
|
| 153 |
+
print(scores.tolist())
|
| 154 |
+
```
|
| 155 |
+
|
| 156 |
+
# 📄 License
|
| 157 |
+
|
| 158 |
+
* The code in this repo is released under the [Apache-2.0](https://github.com/OpenBMB/MiniCPM/blob/main/LICENSE) License.
|
| 159 |
+
* The usage of **VisRAG-Ret** model weights must strictly follow [MiniCPM Model License.md](https://github.com/OpenBMB/MiniCPM/blob/main/MiniCPM%20Model%20License.md).
|
| 160 |
+
* The models and weights of **VisRAG-Ret** are completely free for academic research. After filling out a ["questionnaire"](https://modelbest.feishu.cn/share/base/form/shrcnpV5ZT9EJ6xYjh3Kx0J6v8g) for registration, **VisRAG-Ret** weights are also available for free commercial use.
|
| 161 |
+
|
| 162 |
+
# 📑 Citation
|
| 163 |
+
|
| 164 |
+
```
|
| 165 |
+
@misc{yu2024visragvisionbasedretrievalaugmentedgeneration,
|
| 166 |
+
title={VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents},
|
| 167 |
+
author={Shi Yu and Chaoyue Tang and Bokai Xu and Junbo Cui and Junhao Ran and Yukun Yan and Zhenghao Liu and Shuo Wang and Xu Han and Zhiyuan Liu and Maosong Sun},
|
| 168 |
+
year={2024},
|
| 169 |
+
eprint={2410.10594},
|
| 170 |
+
archivePrefix={arXiv},
|
| 171 |
+
primaryClass={cs.IR},
|
| 172 |
+
url={https://arxiv.org/abs/2410.10594},
|
| 173 |
+
}
|
| 174 |
+
```
|
| 175 |
+
|
| 176 |
+
# 📧 Contact
|
| 177 |
+
|
| 178 |
+
- Shi Yu: yus21@mails.tsinghua.edu.cn
|
| 179 |
+
- Chaoyue Tang: tcy006@gmail.com
|
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "openbmb/VisIR",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"VisIR"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"auto_map": {
|
| 9 |
+
"AutoConfig": "configuration_minicpm.MiniCPMVConfig",
|
| 10 |
+
"AutoModel": "modeling_vis_ir.VisIR"
|
| 11 |
+
},
|
| 12 |
+
"bos_token_id": 1,
|
| 13 |
+
"dim_model_base": 256,
|
| 14 |
+
"drop_vision_last_layer": true,
|
| 15 |
+
"eos_token_id": 2,
|
| 16 |
+
"hidden_act": "silu",
|
| 17 |
+
"hidden_size": 2304,
|
| 18 |
+
"image_size": 448,
|
| 19 |
+
"initializer_range": 0.1,
|
| 20 |
+
"intermediate_size": 5760,
|
| 21 |
+
"max_position_embeddings": 4096,
|
| 22 |
+
"max_slice_nums": 9,
|
| 23 |
+
"mm_use_im_start_end": true,
|
| 24 |
+
"model_name_or_path": "openbmb/VisIR",
|
| 25 |
+
"model_type": "minicpmv",
|
| 26 |
+
"num_attention_heads": 36,
|
| 27 |
+
"num_hidden_layers": 40,
|
| 28 |
+
"num_key_value_heads": 36,
|
| 29 |
+
"patch_size": 14,
|
| 30 |
+
"pretraining_tp": 1,
|
| 31 |
+
"query_num": 64,
|
| 32 |
+
"rms_norm_eps": 1e-05,
|
| 33 |
+
"rope_scaling": null,
|
| 34 |
+
"rope_theta": 10000.0,
|
| 35 |
+
"scale_depth": 1.4,
|
| 36 |
+
"scale_emb": 12,
|
| 37 |
+
"scale_resolution": 448,
|
| 38 |
+
"slice_mode": true,
|
| 39 |
+
"tie_word_embeddings": false,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
+
"transformers_version": "4.40.2",
|
| 42 |
+
"use_cache": true,
|
| 43 |
+
"vision_encoder": "vit_so400m_patch14_siglip_384.webli",
|
| 44 |
+
"vocab_size": 122753
|
| 45 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 1,
|
| 4 |
+
"eos_token_id": 2,
|
| 5 |
+
"transformers_version": "4.40.2"
|
| 6 |
+
}
|
model-00001-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f2f1f0b353cd125705111c0363dd45baa26a586972152b2d55858bb7e8e53a2
|
| 3 |
+
size 4993235928
|
model-00002-of-00002.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98c53ff1bf18df15d896e702672160dd16df28ec2a0fca8be49f0b11fd18028d
|
| 3 |
+
size 2741383120
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1,1070 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"metadata": {
|
| 3 |
+
"total_size": 7734496384
|
| 4 |
+
},
|
| 5 |
+
"weight_map": {
|
| 6 |
+
"coupled_vpm.vpm1.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 7 |
+
"coupled_vpm.vpm1.blocks.0.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 8 |
+
"coupled_vpm.vpm1.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 9 |
+
"coupled_vpm.vpm1.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 10 |
+
"coupled_vpm.vpm1.blocks.0.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 11 |
+
"coupled_vpm.vpm1.blocks.0.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 12 |
+
"coupled_vpm.vpm1.blocks.0.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 13 |
+
"coupled_vpm.vpm1.blocks.0.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 14 |
+
"coupled_vpm.vpm1.blocks.0.norm1.bias": "model-00002-of-00002.safetensors",
|
| 15 |
+
"coupled_vpm.vpm1.blocks.0.norm1.weight": "model-00002-of-00002.safetensors",
|
| 16 |
+
"coupled_vpm.vpm1.blocks.0.norm2.bias": "model-00002-of-00002.safetensors",
|
| 17 |
+
"coupled_vpm.vpm1.blocks.0.norm2.weight": "model-00002-of-00002.safetensors",
|
| 18 |
+
"coupled_vpm.vpm1.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 19 |
+
"coupled_vpm.vpm1.blocks.1.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 20 |
+
"coupled_vpm.vpm1.blocks.1.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 21 |
+
"coupled_vpm.vpm1.blocks.1.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 22 |
+
"coupled_vpm.vpm1.blocks.1.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 23 |
+
"coupled_vpm.vpm1.blocks.1.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 24 |
+
"coupled_vpm.vpm1.blocks.1.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 25 |
+
"coupled_vpm.vpm1.blocks.1.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 26 |
+
"coupled_vpm.vpm1.blocks.1.norm1.bias": "model-00002-of-00002.safetensors",
|
| 27 |
+
"coupled_vpm.vpm1.blocks.1.norm1.weight": "model-00002-of-00002.safetensors",
|
| 28 |
+
"coupled_vpm.vpm1.blocks.1.norm2.bias": "model-00002-of-00002.safetensors",
|
| 29 |
+
"coupled_vpm.vpm1.blocks.1.norm2.weight": "model-00002-of-00002.safetensors",
|
| 30 |
+
"coupled_vpm.vpm1.blocks.10.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 31 |
+
"coupled_vpm.vpm1.blocks.10.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 32 |
+
"coupled_vpm.vpm1.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 33 |
+
"coupled_vpm.vpm1.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 34 |
+
"coupled_vpm.vpm1.blocks.10.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 35 |
+
"coupled_vpm.vpm1.blocks.10.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 36 |
+
"coupled_vpm.vpm1.blocks.10.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 37 |
+
"coupled_vpm.vpm1.blocks.10.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 38 |
+
"coupled_vpm.vpm1.blocks.10.norm1.bias": "model-00002-of-00002.safetensors",
|
| 39 |
+
"coupled_vpm.vpm1.blocks.10.norm1.weight": "model-00002-of-00002.safetensors",
|
| 40 |
+
"coupled_vpm.vpm1.blocks.10.norm2.bias": "model-00002-of-00002.safetensors",
|
| 41 |
+
"coupled_vpm.vpm1.blocks.10.norm2.weight": "model-00002-of-00002.safetensors",
|
| 42 |
+
"coupled_vpm.vpm1.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 43 |
+
"coupled_vpm.vpm1.blocks.11.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 44 |
+
"coupled_vpm.vpm1.blocks.11.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 45 |
+
"coupled_vpm.vpm1.blocks.11.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 46 |
+
"coupled_vpm.vpm1.blocks.11.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 47 |
+
"coupled_vpm.vpm1.blocks.11.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 48 |
+
"coupled_vpm.vpm1.blocks.11.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 49 |
+
"coupled_vpm.vpm1.blocks.11.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 50 |
+
"coupled_vpm.vpm1.blocks.11.norm1.bias": "model-00002-of-00002.safetensors",
|
| 51 |
+
"coupled_vpm.vpm1.blocks.11.norm1.weight": "model-00002-of-00002.safetensors",
|
| 52 |
+
"coupled_vpm.vpm1.blocks.11.norm2.bias": "model-00002-of-00002.safetensors",
|
| 53 |
+
"coupled_vpm.vpm1.blocks.11.norm2.weight": "model-00002-of-00002.safetensors",
|
| 54 |
+
"coupled_vpm.vpm1.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 55 |
+
"coupled_vpm.vpm1.blocks.12.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 56 |
+
"coupled_vpm.vpm1.blocks.12.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 57 |
+
"coupled_vpm.vpm1.blocks.12.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 58 |
+
"coupled_vpm.vpm1.blocks.12.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 59 |
+
"coupled_vpm.vpm1.blocks.12.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 60 |
+
"coupled_vpm.vpm1.blocks.12.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 61 |
+
"coupled_vpm.vpm1.blocks.12.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 62 |
+
"coupled_vpm.vpm1.blocks.12.norm1.bias": "model-00002-of-00002.safetensors",
|
| 63 |
+
"coupled_vpm.vpm1.blocks.12.norm1.weight": "model-00002-of-00002.safetensors",
|
| 64 |
+
"coupled_vpm.vpm1.blocks.12.norm2.bias": "model-00002-of-00002.safetensors",
|
| 65 |
+
"coupled_vpm.vpm1.blocks.12.norm2.weight": "model-00002-of-00002.safetensors",
|
| 66 |
+
"coupled_vpm.vpm1.blocks.13.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 67 |
+
"coupled_vpm.vpm1.blocks.13.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 68 |
+
"coupled_vpm.vpm1.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 69 |
+
"coupled_vpm.vpm1.blocks.13.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 70 |
+
"coupled_vpm.vpm1.blocks.13.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 71 |
+
"coupled_vpm.vpm1.blocks.13.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 72 |
+
"coupled_vpm.vpm1.blocks.13.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 73 |
+
"coupled_vpm.vpm1.blocks.13.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 74 |
+
"coupled_vpm.vpm1.blocks.13.norm1.bias": "model-00002-of-00002.safetensors",
|
| 75 |
+
"coupled_vpm.vpm1.blocks.13.norm1.weight": "model-00002-of-00002.safetensors",
|
| 76 |
+
"coupled_vpm.vpm1.blocks.13.norm2.bias": "model-00002-of-00002.safetensors",
|
| 77 |
+
"coupled_vpm.vpm1.blocks.13.norm2.weight": "model-00002-of-00002.safetensors",
|
| 78 |
+
"coupled_vpm.vpm1.blocks.14.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 79 |
+
"coupled_vpm.vpm1.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 80 |
+
"coupled_vpm.vpm1.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 81 |
+
"coupled_vpm.vpm1.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 82 |
+
"coupled_vpm.vpm1.blocks.14.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 83 |
+
"coupled_vpm.vpm1.blocks.14.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 84 |
+
"coupled_vpm.vpm1.blocks.14.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 85 |
+
"coupled_vpm.vpm1.blocks.14.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 86 |
+
"coupled_vpm.vpm1.blocks.14.norm1.bias": "model-00002-of-00002.safetensors",
|
| 87 |
+
"coupled_vpm.vpm1.blocks.14.norm1.weight": "model-00002-of-00002.safetensors",
|
| 88 |
+
"coupled_vpm.vpm1.blocks.14.norm2.bias": "model-00002-of-00002.safetensors",
|
| 89 |
+
"coupled_vpm.vpm1.blocks.14.norm2.weight": "model-00002-of-00002.safetensors",
|
| 90 |
+
"coupled_vpm.vpm1.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 91 |
+
"coupled_vpm.vpm1.blocks.15.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 92 |
+
"coupled_vpm.vpm1.blocks.15.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 93 |
+
"coupled_vpm.vpm1.blocks.15.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 94 |
+
"coupled_vpm.vpm1.blocks.15.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 95 |
+
"coupled_vpm.vpm1.blocks.15.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 96 |
+
"coupled_vpm.vpm1.blocks.15.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 97 |
+
"coupled_vpm.vpm1.blocks.15.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 98 |
+
"coupled_vpm.vpm1.blocks.15.norm1.bias": "model-00002-of-00002.safetensors",
|
| 99 |
+
"coupled_vpm.vpm1.blocks.15.norm1.weight": "model-00002-of-00002.safetensors",
|
| 100 |
+
"coupled_vpm.vpm1.blocks.15.norm2.bias": "model-00002-of-00002.safetensors",
|
| 101 |
+
"coupled_vpm.vpm1.blocks.15.norm2.weight": "model-00002-of-00002.safetensors",
|
| 102 |
+
"coupled_vpm.vpm1.blocks.16.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 103 |
+
"coupled_vpm.vpm1.blocks.16.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 104 |
+
"coupled_vpm.vpm1.blocks.16.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 105 |
+
"coupled_vpm.vpm1.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 106 |
+
"coupled_vpm.vpm1.blocks.16.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 107 |
+
"coupled_vpm.vpm1.blocks.16.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 108 |
+
"coupled_vpm.vpm1.blocks.16.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 109 |
+
"coupled_vpm.vpm1.blocks.16.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 110 |
+
"coupled_vpm.vpm1.blocks.16.norm1.bias": "model-00002-of-00002.safetensors",
|
| 111 |
+
"coupled_vpm.vpm1.blocks.16.norm1.weight": "model-00002-of-00002.safetensors",
|
| 112 |
+
"coupled_vpm.vpm1.blocks.16.norm2.bias": "model-00002-of-00002.safetensors",
|
| 113 |
+
"coupled_vpm.vpm1.blocks.16.norm2.weight": "model-00002-of-00002.safetensors",
|
| 114 |
+
"coupled_vpm.vpm1.blocks.17.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 115 |
+
"coupled_vpm.vpm1.blocks.17.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 116 |
+
"coupled_vpm.vpm1.blocks.17.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 117 |
+
"coupled_vpm.vpm1.blocks.17.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 118 |
+
"coupled_vpm.vpm1.blocks.17.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 119 |
+
"coupled_vpm.vpm1.blocks.17.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 120 |
+
"coupled_vpm.vpm1.blocks.17.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 121 |
+
"coupled_vpm.vpm1.blocks.17.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 122 |
+
"coupled_vpm.vpm1.blocks.17.norm1.bias": "model-00002-of-00002.safetensors",
|
| 123 |
+
"coupled_vpm.vpm1.blocks.17.norm1.weight": "model-00002-of-00002.safetensors",
|
| 124 |
+
"coupled_vpm.vpm1.blocks.17.norm2.bias": "model-00002-of-00002.safetensors",
|
| 125 |
+
"coupled_vpm.vpm1.blocks.17.norm2.weight": "model-00002-of-00002.safetensors",
|
| 126 |
+
"coupled_vpm.vpm1.blocks.18.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 127 |
+
"coupled_vpm.vpm1.blocks.18.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 128 |
+
"coupled_vpm.vpm1.blocks.18.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 129 |
+
"coupled_vpm.vpm1.blocks.18.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 130 |
+
"coupled_vpm.vpm1.blocks.18.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 131 |
+
"coupled_vpm.vpm1.blocks.18.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 132 |
+
"coupled_vpm.vpm1.blocks.18.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 133 |
+
"coupled_vpm.vpm1.blocks.18.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 134 |
+
"coupled_vpm.vpm1.blocks.18.norm1.bias": "model-00002-of-00002.safetensors",
|
| 135 |
+
"coupled_vpm.vpm1.blocks.18.norm1.weight": "model-00002-of-00002.safetensors",
|
| 136 |
+
"coupled_vpm.vpm1.blocks.18.norm2.bias": "model-00002-of-00002.safetensors",
|
| 137 |
+
"coupled_vpm.vpm1.blocks.18.norm2.weight": "model-00002-of-00002.safetensors",
|
| 138 |
+
"coupled_vpm.vpm1.blocks.19.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 139 |
+
"coupled_vpm.vpm1.blocks.19.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 140 |
+
"coupled_vpm.vpm1.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 141 |
+
"coupled_vpm.vpm1.blocks.19.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 142 |
+
"coupled_vpm.vpm1.blocks.19.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 143 |
+
"coupled_vpm.vpm1.blocks.19.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 144 |
+
"coupled_vpm.vpm1.blocks.19.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 145 |
+
"coupled_vpm.vpm1.blocks.19.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 146 |
+
"coupled_vpm.vpm1.blocks.19.norm1.bias": "model-00002-of-00002.safetensors",
|
| 147 |
+
"coupled_vpm.vpm1.blocks.19.norm1.weight": "model-00002-of-00002.safetensors",
|
| 148 |
+
"coupled_vpm.vpm1.blocks.19.norm2.bias": "model-00002-of-00002.safetensors",
|
| 149 |
+
"coupled_vpm.vpm1.blocks.19.norm2.weight": "model-00002-of-00002.safetensors",
|
| 150 |
+
"coupled_vpm.vpm1.blocks.2.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 151 |
+
"coupled_vpm.vpm1.blocks.2.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 152 |
+
"coupled_vpm.vpm1.blocks.2.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 153 |
+
"coupled_vpm.vpm1.blocks.2.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 154 |
+
"coupled_vpm.vpm1.blocks.2.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 155 |
+
"coupled_vpm.vpm1.blocks.2.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 156 |
+
"coupled_vpm.vpm1.blocks.2.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 157 |
+
"coupled_vpm.vpm1.blocks.2.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 158 |
+
"coupled_vpm.vpm1.blocks.2.norm1.bias": "model-00002-of-00002.safetensors",
|
| 159 |
+
"coupled_vpm.vpm1.blocks.2.norm1.weight": "model-00002-of-00002.safetensors",
|
| 160 |
+
"coupled_vpm.vpm1.blocks.2.norm2.bias": "model-00002-of-00002.safetensors",
|
| 161 |
+
"coupled_vpm.vpm1.blocks.2.norm2.weight": "model-00002-of-00002.safetensors",
|
| 162 |
+
"coupled_vpm.vpm1.blocks.20.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 163 |
+
"coupled_vpm.vpm1.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 164 |
+
"coupled_vpm.vpm1.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 165 |
+
"coupled_vpm.vpm1.blocks.20.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 166 |
+
"coupled_vpm.vpm1.blocks.20.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 167 |
+
"coupled_vpm.vpm1.blocks.20.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 168 |
+
"coupled_vpm.vpm1.blocks.20.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 169 |
+
"coupled_vpm.vpm1.blocks.20.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 170 |
+
"coupled_vpm.vpm1.blocks.20.norm1.bias": "model-00002-of-00002.safetensors",
|
| 171 |
+
"coupled_vpm.vpm1.blocks.20.norm1.weight": "model-00002-of-00002.safetensors",
|
| 172 |
+
"coupled_vpm.vpm1.blocks.20.norm2.bias": "model-00002-of-00002.safetensors",
|
| 173 |
+
"coupled_vpm.vpm1.blocks.20.norm2.weight": "model-00002-of-00002.safetensors",
|
| 174 |
+
"coupled_vpm.vpm1.blocks.21.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 175 |
+
"coupled_vpm.vpm1.blocks.21.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 176 |
+
"coupled_vpm.vpm1.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 177 |
+
"coupled_vpm.vpm1.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 178 |
+
"coupled_vpm.vpm1.blocks.21.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 179 |
+
"coupled_vpm.vpm1.blocks.21.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 180 |
+
"coupled_vpm.vpm1.blocks.21.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 181 |
+
"coupled_vpm.vpm1.blocks.21.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 182 |
+
"coupled_vpm.vpm1.blocks.21.norm1.bias": "model-00002-of-00002.safetensors",
|
| 183 |
+
"coupled_vpm.vpm1.blocks.21.norm1.weight": "model-00002-of-00002.safetensors",
|
| 184 |
+
"coupled_vpm.vpm1.blocks.21.norm2.bias": "model-00002-of-00002.safetensors",
|
| 185 |
+
"coupled_vpm.vpm1.blocks.21.norm2.weight": "model-00002-of-00002.safetensors",
|
| 186 |
+
"coupled_vpm.vpm1.blocks.22.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 187 |
+
"coupled_vpm.vpm1.blocks.22.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 188 |
+
"coupled_vpm.vpm1.blocks.22.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 189 |
+
"coupled_vpm.vpm1.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 190 |
+
"coupled_vpm.vpm1.blocks.22.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 191 |
+
"coupled_vpm.vpm1.blocks.22.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 192 |
+
"coupled_vpm.vpm1.blocks.22.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 193 |
+
"coupled_vpm.vpm1.blocks.22.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 194 |
+
"coupled_vpm.vpm1.blocks.22.norm1.bias": "model-00002-of-00002.safetensors",
|
| 195 |
+
"coupled_vpm.vpm1.blocks.22.norm1.weight": "model-00002-of-00002.safetensors",
|
| 196 |
+
"coupled_vpm.vpm1.blocks.22.norm2.bias": "model-00002-of-00002.safetensors",
|
| 197 |
+
"coupled_vpm.vpm1.blocks.22.norm2.weight": "model-00002-of-00002.safetensors",
|
| 198 |
+
"coupled_vpm.vpm1.blocks.23.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 199 |
+
"coupled_vpm.vpm1.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 200 |
+
"coupled_vpm.vpm1.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 201 |
+
"coupled_vpm.vpm1.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 202 |
+
"coupled_vpm.vpm1.blocks.23.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 203 |
+
"coupled_vpm.vpm1.blocks.23.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 204 |
+
"coupled_vpm.vpm1.blocks.23.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 205 |
+
"coupled_vpm.vpm1.blocks.23.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 206 |
+
"coupled_vpm.vpm1.blocks.23.norm1.bias": "model-00002-of-00002.safetensors",
|
| 207 |
+
"coupled_vpm.vpm1.blocks.23.norm1.weight": "model-00002-of-00002.safetensors",
|
| 208 |
+
"coupled_vpm.vpm1.blocks.23.norm2.bias": "model-00002-of-00002.safetensors",
|
| 209 |
+
"coupled_vpm.vpm1.blocks.23.norm2.weight": "model-00002-of-00002.safetensors",
|
| 210 |
+
"coupled_vpm.vpm1.blocks.24.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 211 |
+
"coupled_vpm.vpm1.blocks.24.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 212 |
+
"coupled_vpm.vpm1.blocks.24.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 213 |
+
"coupled_vpm.vpm1.blocks.24.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 214 |
+
"coupled_vpm.vpm1.blocks.24.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 215 |
+
"coupled_vpm.vpm1.blocks.24.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 216 |
+
"coupled_vpm.vpm1.blocks.24.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 217 |
+
"coupled_vpm.vpm1.blocks.24.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 218 |
+
"coupled_vpm.vpm1.blocks.24.norm1.bias": "model-00002-of-00002.safetensors",
|
| 219 |
+
"coupled_vpm.vpm1.blocks.24.norm1.weight": "model-00002-of-00002.safetensors",
|
| 220 |
+
"coupled_vpm.vpm1.blocks.24.norm2.bias": "model-00002-of-00002.safetensors",
|
| 221 |
+
"coupled_vpm.vpm1.blocks.24.norm2.weight": "model-00002-of-00002.safetensors",
|
| 222 |
+
"coupled_vpm.vpm1.blocks.25.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 223 |
+
"coupled_vpm.vpm1.blocks.25.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 224 |
+
"coupled_vpm.vpm1.blocks.25.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 225 |
+
"coupled_vpm.vpm1.blocks.25.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 226 |
+
"coupled_vpm.vpm1.blocks.25.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 227 |
+
"coupled_vpm.vpm1.blocks.25.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 228 |
+
"coupled_vpm.vpm1.blocks.25.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 229 |
+
"coupled_vpm.vpm1.blocks.25.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 230 |
+
"coupled_vpm.vpm1.blocks.25.norm1.bias": "model-00002-of-00002.safetensors",
|
| 231 |
+
"coupled_vpm.vpm1.blocks.25.norm1.weight": "model-00002-of-00002.safetensors",
|
| 232 |
+
"coupled_vpm.vpm1.blocks.25.norm2.bias": "model-00002-of-00002.safetensors",
|
| 233 |
+
"coupled_vpm.vpm1.blocks.25.norm2.weight": "model-00002-of-00002.safetensors",
|
| 234 |
+
"coupled_vpm.vpm1.blocks.3.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 235 |
+
"coupled_vpm.vpm1.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 236 |
+
"coupled_vpm.vpm1.blocks.3.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 237 |
+
"coupled_vpm.vpm1.blocks.3.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 238 |
+
"coupled_vpm.vpm1.blocks.3.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 239 |
+
"coupled_vpm.vpm1.blocks.3.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 240 |
+
"coupled_vpm.vpm1.blocks.3.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 241 |
+
"coupled_vpm.vpm1.blocks.3.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 242 |
+
"coupled_vpm.vpm1.blocks.3.norm1.bias": "model-00002-of-00002.safetensors",
|
| 243 |
+
"coupled_vpm.vpm1.blocks.3.norm1.weight": "model-00002-of-00002.safetensors",
|
| 244 |
+
"coupled_vpm.vpm1.blocks.3.norm2.bias": "model-00002-of-00002.safetensors",
|
| 245 |
+
"coupled_vpm.vpm1.blocks.3.norm2.weight": "model-00002-of-00002.safetensors",
|
| 246 |
+
"coupled_vpm.vpm1.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 247 |
+
"coupled_vpm.vpm1.blocks.4.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 248 |
+
"coupled_vpm.vpm1.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 249 |
+
"coupled_vpm.vpm1.blocks.4.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 250 |
+
"coupled_vpm.vpm1.blocks.4.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 251 |
+
"coupled_vpm.vpm1.blocks.4.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 252 |
+
"coupled_vpm.vpm1.blocks.4.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 253 |
+
"coupled_vpm.vpm1.blocks.4.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 254 |
+
"coupled_vpm.vpm1.blocks.4.norm1.bias": "model-00002-of-00002.safetensors",
|
| 255 |
+
"coupled_vpm.vpm1.blocks.4.norm1.weight": "model-00002-of-00002.safetensors",
|
| 256 |
+
"coupled_vpm.vpm1.blocks.4.norm2.bias": "model-00002-of-00002.safetensors",
|
| 257 |
+
"coupled_vpm.vpm1.blocks.4.norm2.weight": "model-00002-of-00002.safetensors",
|
| 258 |
+
"coupled_vpm.vpm1.blocks.5.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 259 |
+
"coupled_vpm.vpm1.blocks.5.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 260 |
+
"coupled_vpm.vpm1.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 261 |
+
"coupled_vpm.vpm1.blocks.5.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 262 |
+
"coupled_vpm.vpm1.blocks.5.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 263 |
+
"coupled_vpm.vpm1.blocks.5.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 264 |
+
"coupled_vpm.vpm1.blocks.5.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 265 |
+
"coupled_vpm.vpm1.blocks.5.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 266 |
+
"coupled_vpm.vpm1.blocks.5.norm1.bias": "model-00002-of-00002.safetensors",
|
| 267 |
+
"coupled_vpm.vpm1.blocks.5.norm1.weight": "model-00002-of-00002.safetensors",
|
| 268 |
+
"coupled_vpm.vpm1.blocks.5.norm2.bias": "model-00002-of-00002.safetensors",
|
| 269 |
+
"coupled_vpm.vpm1.blocks.5.norm2.weight": "model-00002-of-00002.safetensors",
|
| 270 |
+
"coupled_vpm.vpm1.blocks.6.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 271 |
+
"coupled_vpm.vpm1.blocks.6.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 272 |
+
"coupled_vpm.vpm1.blocks.6.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 273 |
+
"coupled_vpm.vpm1.blocks.6.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 274 |
+
"coupled_vpm.vpm1.blocks.6.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 275 |
+
"coupled_vpm.vpm1.blocks.6.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 276 |
+
"coupled_vpm.vpm1.blocks.6.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 277 |
+
"coupled_vpm.vpm1.blocks.6.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 278 |
+
"coupled_vpm.vpm1.blocks.6.norm1.bias": "model-00002-of-00002.safetensors",
|
| 279 |
+
"coupled_vpm.vpm1.blocks.6.norm1.weight": "model-00002-of-00002.safetensors",
|
| 280 |
+
"coupled_vpm.vpm1.blocks.6.norm2.bias": "model-00002-of-00002.safetensors",
|
| 281 |
+
"coupled_vpm.vpm1.blocks.6.norm2.weight": "model-00002-of-00002.safetensors",
|
| 282 |
+
"coupled_vpm.vpm1.blocks.7.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 283 |
+
"coupled_vpm.vpm1.blocks.7.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 284 |
+
"coupled_vpm.vpm1.blocks.7.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 285 |
+
"coupled_vpm.vpm1.blocks.7.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 286 |
+
"coupled_vpm.vpm1.blocks.7.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 287 |
+
"coupled_vpm.vpm1.blocks.7.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 288 |
+
"coupled_vpm.vpm1.blocks.7.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 289 |
+
"coupled_vpm.vpm1.blocks.7.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 290 |
+
"coupled_vpm.vpm1.blocks.7.norm1.bias": "model-00002-of-00002.safetensors",
|
| 291 |
+
"coupled_vpm.vpm1.blocks.7.norm1.weight": "model-00002-of-00002.safetensors",
|
| 292 |
+
"coupled_vpm.vpm1.blocks.7.norm2.bias": "model-00002-of-00002.safetensors",
|
| 293 |
+
"coupled_vpm.vpm1.blocks.7.norm2.weight": "model-00002-of-00002.safetensors",
|
| 294 |
+
"coupled_vpm.vpm1.blocks.8.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 295 |
+
"coupled_vpm.vpm1.blocks.8.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 296 |
+
"coupled_vpm.vpm1.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 297 |
+
"coupled_vpm.vpm1.blocks.8.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 298 |
+
"coupled_vpm.vpm1.blocks.8.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 299 |
+
"coupled_vpm.vpm1.blocks.8.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 300 |
+
"coupled_vpm.vpm1.blocks.8.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 301 |
+
"coupled_vpm.vpm1.blocks.8.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 302 |
+
"coupled_vpm.vpm1.blocks.8.norm1.bias": "model-00002-of-00002.safetensors",
|
| 303 |
+
"coupled_vpm.vpm1.blocks.8.norm1.weight": "model-00002-of-00002.safetensors",
|
| 304 |
+
"coupled_vpm.vpm1.blocks.8.norm2.bias": "model-00002-of-00002.safetensors",
|
| 305 |
+
"coupled_vpm.vpm1.blocks.8.norm2.weight": "model-00002-of-00002.safetensors",
|
| 306 |
+
"coupled_vpm.vpm1.blocks.9.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 307 |
+
"coupled_vpm.vpm1.blocks.9.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 308 |
+
"coupled_vpm.vpm1.blocks.9.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 309 |
+
"coupled_vpm.vpm1.blocks.9.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 310 |
+
"coupled_vpm.vpm1.blocks.9.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 311 |
+
"coupled_vpm.vpm1.blocks.9.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 312 |
+
"coupled_vpm.vpm1.blocks.9.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 313 |
+
"coupled_vpm.vpm1.blocks.9.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 314 |
+
"coupled_vpm.vpm1.blocks.9.norm1.bias": "model-00002-of-00002.safetensors",
|
| 315 |
+
"coupled_vpm.vpm1.blocks.9.norm1.weight": "model-00002-of-00002.safetensors",
|
| 316 |
+
"coupled_vpm.vpm1.blocks.9.norm2.bias": "model-00002-of-00002.safetensors",
|
| 317 |
+
"coupled_vpm.vpm1.blocks.9.norm2.weight": "model-00002-of-00002.safetensors",
|
| 318 |
+
"coupled_vpm.vpm1.norm.bias": "model-00002-of-00002.safetensors",
|
| 319 |
+
"coupled_vpm.vpm1.norm.weight": "model-00002-of-00002.safetensors",
|
| 320 |
+
"coupled_vpm.vpm1.patch_embed.proj.bias": "model-00002-of-00002.safetensors",
|
| 321 |
+
"coupled_vpm.vpm1.patch_embed.proj.weight": "model-00002-of-00002.safetensors",
|
| 322 |
+
"coupled_vpm.vpm1.pos_embed": "model-00002-of-00002.safetensors",
|
| 323 |
+
"coupled_vpm.vpm2.blocks.0.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 324 |
+
"coupled_vpm.vpm2.blocks.0.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 325 |
+
"coupled_vpm.vpm2.blocks.0.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 326 |
+
"coupled_vpm.vpm2.blocks.0.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 327 |
+
"coupled_vpm.vpm2.blocks.0.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 328 |
+
"coupled_vpm.vpm2.blocks.0.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 329 |
+
"coupled_vpm.vpm2.blocks.0.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 330 |
+
"coupled_vpm.vpm2.blocks.0.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 331 |
+
"coupled_vpm.vpm2.blocks.0.norm1.bias": "model-00002-of-00002.safetensors",
|
| 332 |
+
"coupled_vpm.vpm2.blocks.0.norm1.weight": "model-00002-of-00002.safetensors",
|
| 333 |
+
"coupled_vpm.vpm2.blocks.0.norm2.bias": "model-00002-of-00002.safetensors",
|
| 334 |
+
"coupled_vpm.vpm2.blocks.0.norm2.weight": "model-00002-of-00002.safetensors",
|
| 335 |
+
"coupled_vpm.vpm2.blocks.1.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 336 |
+
"coupled_vpm.vpm2.blocks.1.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 337 |
+
"coupled_vpm.vpm2.blocks.1.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 338 |
+
"coupled_vpm.vpm2.blocks.1.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 339 |
+
"coupled_vpm.vpm2.blocks.1.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 340 |
+
"coupled_vpm.vpm2.blocks.1.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 341 |
+
"coupled_vpm.vpm2.blocks.1.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 342 |
+
"coupled_vpm.vpm2.blocks.1.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 343 |
+
"coupled_vpm.vpm2.blocks.1.norm1.bias": "model-00002-of-00002.safetensors",
|
| 344 |
+
"coupled_vpm.vpm2.blocks.1.norm1.weight": "model-00002-of-00002.safetensors",
|
| 345 |
+
"coupled_vpm.vpm2.blocks.1.norm2.bias": "model-00002-of-00002.safetensors",
|
| 346 |
+
"coupled_vpm.vpm2.blocks.1.norm2.weight": "model-00002-of-00002.safetensors",
|
| 347 |
+
"coupled_vpm.vpm2.blocks.10.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 348 |
+
"coupled_vpm.vpm2.blocks.10.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 349 |
+
"coupled_vpm.vpm2.blocks.10.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 350 |
+
"coupled_vpm.vpm2.blocks.10.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 351 |
+
"coupled_vpm.vpm2.blocks.10.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 352 |
+
"coupled_vpm.vpm2.blocks.10.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 353 |
+
"coupled_vpm.vpm2.blocks.10.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 354 |
+
"coupled_vpm.vpm2.blocks.10.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 355 |
+
"coupled_vpm.vpm2.blocks.10.norm1.bias": "model-00002-of-00002.safetensors",
|
| 356 |
+
"coupled_vpm.vpm2.blocks.10.norm1.weight": "model-00002-of-00002.safetensors",
|
| 357 |
+
"coupled_vpm.vpm2.blocks.10.norm2.bias": "model-00002-of-00002.safetensors",
|
| 358 |
+
"coupled_vpm.vpm2.blocks.10.norm2.weight": "model-00002-of-00002.safetensors",
|
| 359 |
+
"coupled_vpm.vpm2.blocks.11.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 360 |
+
"coupled_vpm.vpm2.blocks.11.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 361 |
+
"coupled_vpm.vpm2.blocks.11.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 362 |
+
"coupled_vpm.vpm2.blocks.11.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 363 |
+
"coupled_vpm.vpm2.blocks.11.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 364 |
+
"coupled_vpm.vpm2.blocks.11.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 365 |
+
"coupled_vpm.vpm2.blocks.11.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 366 |
+
"coupled_vpm.vpm2.blocks.11.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 367 |
+
"coupled_vpm.vpm2.blocks.11.norm1.bias": "model-00002-of-00002.safetensors",
|
| 368 |
+
"coupled_vpm.vpm2.blocks.11.norm1.weight": "model-00002-of-00002.safetensors",
|
| 369 |
+
"coupled_vpm.vpm2.blocks.11.norm2.bias": "model-00002-of-00002.safetensors",
|
| 370 |
+
"coupled_vpm.vpm2.blocks.11.norm2.weight": "model-00002-of-00002.safetensors",
|
| 371 |
+
"coupled_vpm.vpm2.blocks.12.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 372 |
+
"coupled_vpm.vpm2.blocks.12.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 373 |
+
"coupled_vpm.vpm2.blocks.12.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 374 |
+
"coupled_vpm.vpm2.blocks.12.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 375 |
+
"coupled_vpm.vpm2.blocks.12.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 376 |
+
"coupled_vpm.vpm2.blocks.12.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 377 |
+
"coupled_vpm.vpm2.blocks.12.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 378 |
+
"coupled_vpm.vpm2.blocks.12.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 379 |
+
"coupled_vpm.vpm2.blocks.12.norm1.bias": "model-00002-of-00002.safetensors",
|
| 380 |
+
"coupled_vpm.vpm2.blocks.12.norm1.weight": "model-00002-of-00002.safetensors",
|
| 381 |
+
"coupled_vpm.vpm2.blocks.12.norm2.bias": "model-00002-of-00002.safetensors",
|
| 382 |
+
"coupled_vpm.vpm2.blocks.12.norm2.weight": "model-00002-of-00002.safetensors",
|
| 383 |
+
"coupled_vpm.vpm2.blocks.13.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 384 |
+
"coupled_vpm.vpm2.blocks.13.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 385 |
+
"coupled_vpm.vpm2.blocks.13.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 386 |
+
"coupled_vpm.vpm2.blocks.13.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 387 |
+
"coupled_vpm.vpm2.blocks.13.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 388 |
+
"coupled_vpm.vpm2.blocks.13.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 389 |
+
"coupled_vpm.vpm2.blocks.13.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 390 |
+
"coupled_vpm.vpm2.blocks.13.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 391 |
+
"coupled_vpm.vpm2.blocks.13.norm1.bias": "model-00002-of-00002.safetensors",
|
| 392 |
+
"coupled_vpm.vpm2.blocks.13.norm1.weight": "model-00002-of-00002.safetensors",
|
| 393 |
+
"coupled_vpm.vpm2.blocks.13.norm2.bias": "model-00002-of-00002.safetensors",
|
| 394 |
+
"coupled_vpm.vpm2.blocks.13.norm2.weight": "model-00002-of-00002.safetensors",
|
| 395 |
+
"coupled_vpm.vpm2.blocks.14.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 396 |
+
"coupled_vpm.vpm2.blocks.14.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 397 |
+
"coupled_vpm.vpm2.blocks.14.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 398 |
+
"coupled_vpm.vpm2.blocks.14.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 399 |
+
"coupled_vpm.vpm2.blocks.14.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 400 |
+
"coupled_vpm.vpm2.blocks.14.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 401 |
+
"coupled_vpm.vpm2.blocks.14.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 402 |
+
"coupled_vpm.vpm2.blocks.14.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 403 |
+
"coupled_vpm.vpm2.blocks.14.norm1.bias": "model-00002-of-00002.safetensors",
|
| 404 |
+
"coupled_vpm.vpm2.blocks.14.norm1.weight": "model-00002-of-00002.safetensors",
|
| 405 |
+
"coupled_vpm.vpm2.blocks.14.norm2.bias": "model-00002-of-00002.safetensors",
|
| 406 |
+
"coupled_vpm.vpm2.blocks.14.norm2.weight": "model-00002-of-00002.safetensors",
|
| 407 |
+
"coupled_vpm.vpm2.blocks.15.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 408 |
+
"coupled_vpm.vpm2.blocks.15.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 409 |
+
"coupled_vpm.vpm2.blocks.15.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 410 |
+
"coupled_vpm.vpm2.blocks.15.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 411 |
+
"coupled_vpm.vpm2.blocks.15.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 412 |
+
"coupled_vpm.vpm2.blocks.15.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 413 |
+
"coupled_vpm.vpm2.blocks.15.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 414 |
+
"coupled_vpm.vpm2.blocks.15.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 415 |
+
"coupled_vpm.vpm2.blocks.15.norm1.bias": "model-00002-of-00002.safetensors",
|
| 416 |
+
"coupled_vpm.vpm2.blocks.15.norm1.weight": "model-00002-of-00002.safetensors",
|
| 417 |
+
"coupled_vpm.vpm2.blocks.15.norm2.bias": "model-00002-of-00002.safetensors",
|
| 418 |
+
"coupled_vpm.vpm2.blocks.15.norm2.weight": "model-00002-of-00002.safetensors",
|
| 419 |
+
"coupled_vpm.vpm2.blocks.16.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 420 |
+
"coupled_vpm.vpm2.blocks.16.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 421 |
+
"coupled_vpm.vpm2.blocks.16.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 422 |
+
"coupled_vpm.vpm2.blocks.16.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 423 |
+
"coupled_vpm.vpm2.blocks.16.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 424 |
+
"coupled_vpm.vpm2.blocks.16.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 425 |
+
"coupled_vpm.vpm2.blocks.16.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 426 |
+
"coupled_vpm.vpm2.blocks.16.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 427 |
+
"coupled_vpm.vpm2.blocks.16.norm1.bias": "model-00002-of-00002.safetensors",
|
| 428 |
+
"coupled_vpm.vpm2.blocks.16.norm1.weight": "model-00002-of-00002.safetensors",
|
| 429 |
+
"coupled_vpm.vpm2.blocks.16.norm2.bias": "model-00002-of-00002.safetensors",
|
| 430 |
+
"coupled_vpm.vpm2.blocks.16.norm2.weight": "model-00002-of-00002.safetensors",
|
| 431 |
+
"coupled_vpm.vpm2.blocks.17.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 432 |
+
"coupled_vpm.vpm2.blocks.17.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 433 |
+
"coupled_vpm.vpm2.blocks.17.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 434 |
+
"coupled_vpm.vpm2.blocks.17.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 435 |
+
"coupled_vpm.vpm2.blocks.17.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 436 |
+
"coupled_vpm.vpm2.blocks.17.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 437 |
+
"coupled_vpm.vpm2.blocks.17.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 438 |
+
"coupled_vpm.vpm2.blocks.17.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 439 |
+
"coupled_vpm.vpm2.blocks.17.norm1.bias": "model-00002-of-00002.safetensors",
|
| 440 |
+
"coupled_vpm.vpm2.blocks.17.norm1.weight": "model-00002-of-00002.safetensors",
|
| 441 |
+
"coupled_vpm.vpm2.blocks.17.norm2.bias": "model-00002-of-00002.safetensors",
|
| 442 |
+
"coupled_vpm.vpm2.blocks.17.norm2.weight": "model-00002-of-00002.safetensors",
|
| 443 |
+
"coupled_vpm.vpm2.blocks.18.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 444 |
+
"coupled_vpm.vpm2.blocks.18.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 445 |
+
"coupled_vpm.vpm2.blocks.18.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 446 |
+
"coupled_vpm.vpm2.blocks.18.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 447 |
+
"coupled_vpm.vpm2.blocks.18.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 448 |
+
"coupled_vpm.vpm2.blocks.18.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 449 |
+
"coupled_vpm.vpm2.blocks.18.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 450 |
+
"coupled_vpm.vpm2.blocks.18.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 451 |
+
"coupled_vpm.vpm2.blocks.18.norm1.bias": "model-00002-of-00002.safetensors",
|
| 452 |
+
"coupled_vpm.vpm2.blocks.18.norm1.weight": "model-00002-of-00002.safetensors",
|
| 453 |
+
"coupled_vpm.vpm2.blocks.18.norm2.bias": "model-00002-of-00002.safetensors",
|
| 454 |
+
"coupled_vpm.vpm2.blocks.18.norm2.weight": "model-00002-of-00002.safetensors",
|
| 455 |
+
"coupled_vpm.vpm2.blocks.19.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 456 |
+
"coupled_vpm.vpm2.blocks.19.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 457 |
+
"coupled_vpm.vpm2.blocks.19.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 458 |
+
"coupled_vpm.vpm2.blocks.19.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 459 |
+
"coupled_vpm.vpm2.blocks.19.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 460 |
+
"coupled_vpm.vpm2.blocks.19.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 461 |
+
"coupled_vpm.vpm2.blocks.19.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 462 |
+
"coupled_vpm.vpm2.blocks.19.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 463 |
+
"coupled_vpm.vpm2.blocks.19.norm1.bias": "model-00002-of-00002.safetensors",
|
| 464 |
+
"coupled_vpm.vpm2.blocks.19.norm1.weight": "model-00002-of-00002.safetensors",
|
| 465 |
+
"coupled_vpm.vpm2.blocks.19.norm2.bias": "model-00002-of-00002.safetensors",
|
| 466 |
+
"coupled_vpm.vpm2.blocks.19.norm2.weight": "model-00002-of-00002.safetensors",
|
| 467 |
+
"coupled_vpm.vpm2.blocks.2.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 468 |
+
"coupled_vpm.vpm2.blocks.2.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 469 |
+
"coupled_vpm.vpm2.blocks.2.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 470 |
+
"coupled_vpm.vpm2.blocks.2.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 471 |
+
"coupled_vpm.vpm2.blocks.2.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 472 |
+
"coupled_vpm.vpm2.blocks.2.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 473 |
+
"coupled_vpm.vpm2.blocks.2.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 474 |
+
"coupled_vpm.vpm2.blocks.2.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 475 |
+
"coupled_vpm.vpm2.blocks.2.norm1.bias": "model-00002-of-00002.safetensors",
|
| 476 |
+
"coupled_vpm.vpm2.blocks.2.norm1.weight": "model-00002-of-00002.safetensors",
|
| 477 |
+
"coupled_vpm.vpm2.blocks.2.norm2.bias": "model-00002-of-00002.safetensors",
|
| 478 |
+
"coupled_vpm.vpm2.blocks.2.norm2.weight": "model-00002-of-00002.safetensors",
|
| 479 |
+
"coupled_vpm.vpm2.blocks.20.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 480 |
+
"coupled_vpm.vpm2.blocks.20.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 481 |
+
"coupled_vpm.vpm2.blocks.20.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 482 |
+
"coupled_vpm.vpm2.blocks.20.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 483 |
+
"coupled_vpm.vpm2.blocks.20.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 484 |
+
"coupled_vpm.vpm2.blocks.20.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 485 |
+
"coupled_vpm.vpm2.blocks.20.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 486 |
+
"coupled_vpm.vpm2.blocks.20.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 487 |
+
"coupled_vpm.vpm2.blocks.20.norm1.bias": "model-00002-of-00002.safetensors",
|
| 488 |
+
"coupled_vpm.vpm2.blocks.20.norm1.weight": "model-00002-of-00002.safetensors",
|
| 489 |
+
"coupled_vpm.vpm2.blocks.20.norm2.bias": "model-00002-of-00002.safetensors",
|
| 490 |
+
"coupled_vpm.vpm2.blocks.20.norm2.weight": "model-00002-of-00002.safetensors",
|
| 491 |
+
"coupled_vpm.vpm2.blocks.21.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 492 |
+
"coupled_vpm.vpm2.blocks.21.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 493 |
+
"coupled_vpm.vpm2.blocks.21.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 494 |
+
"coupled_vpm.vpm2.blocks.21.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 495 |
+
"coupled_vpm.vpm2.blocks.21.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 496 |
+
"coupled_vpm.vpm2.blocks.21.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 497 |
+
"coupled_vpm.vpm2.blocks.21.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 498 |
+
"coupled_vpm.vpm2.blocks.21.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 499 |
+
"coupled_vpm.vpm2.blocks.21.norm1.bias": "model-00002-of-00002.safetensors",
|
| 500 |
+
"coupled_vpm.vpm2.blocks.21.norm1.weight": "model-00002-of-00002.safetensors",
|
| 501 |
+
"coupled_vpm.vpm2.blocks.21.norm2.bias": "model-00002-of-00002.safetensors",
|
| 502 |
+
"coupled_vpm.vpm2.blocks.21.norm2.weight": "model-00002-of-00002.safetensors",
|
| 503 |
+
"coupled_vpm.vpm2.blocks.22.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 504 |
+
"coupled_vpm.vpm2.blocks.22.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 505 |
+
"coupled_vpm.vpm2.blocks.22.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 506 |
+
"coupled_vpm.vpm2.blocks.22.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 507 |
+
"coupled_vpm.vpm2.blocks.22.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 508 |
+
"coupled_vpm.vpm2.blocks.22.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 509 |
+
"coupled_vpm.vpm2.blocks.22.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 510 |
+
"coupled_vpm.vpm2.blocks.22.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 511 |
+
"coupled_vpm.vpm2.blocks.22.norm1.bias": "model-00002-of-00002.safetensors",
|
| 512 |
+
"coupled_vpm.vpm2.blocks.22.norm1.weight": "model-00002-of-00002.safetensors",
|
| 513 |
+
"coupled_vpm.vpm2.blocks.22.norm2.bias": "model-00002-of-00002.safetensors",
|
| 514 |
+
"coupled_vpm.vpm2.blocks.22.norm2.weight": "model-00002-of-00002.safetensors",
|
| 515 |
+
"coupled_vpm.vpm2.blocks.23.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 516 |
+
"coupled_vpm.vpm2.blocks.23.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 517 |
+
"coupled_vpm.vpm2.blocks.23.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 518 |
+
"coupled_vpm.vpm2.blocks.23.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 519 |
+
"coupled_vpm.vpm2.blocks.23.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 520 |
+
"coupled_vpm.vpm2.blocks.23.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 521 |
+
"coupled_vpm.vpm2.blocks.23.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 522 |
+
"coupled_vpm.vpm2.blocks.23.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 523 |
+
"coupled_vpm.vpm2.blocks.23.norm1.bias": "model-00002-of-00002.safetensors",
|
| 524 |
+
"coupled_vpm.vpm2.blocks.23.norm1.weight": "model-00002-of-00002.safetensors",
|
| 525 |
+
"coupled_vpm.vpm2.blocks.23.norm2.bias": "model-00002-of-00002.safetensors",
|
| 526 |
+
"coupled_vpm.vpm2.blocks.23.norm2.weight": "model-00002-of-00002.safetensors",
|
| 527 |
+
"coupled_vpm.vpm2.blocks.24.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 528 |
+
"coupled_vpm.vpm2.blocks.24.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 529 |
+
"coupled_vpm.vpm2.blocks.24.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 530 |
+
"coupled_vpm.vpm2.blocks.24.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 531 |
+
"coupled_vpm.vpm2.blocks.24.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 532 |
+
"coupled_vpm.vpm2.blocks.24.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 533 |
+
"coupled_vpm.vpm2.blocks.24.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 534 |
+
"coupled_vpm.vpm2.blocks.24.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 535 |
+
"coupled_vpm.vpm2.blocks.24.norm1.bias": "model-00002-of-00002.safetensors",
|
| 536 |
+
"coupled_vpm.vpm2.blocks.24.norm1.weight": "model-00002-of-00002.safetensors",
|
| 537 |
+
"coupled_vpm.vpm2.blocks.24.norm2.bias": "model-00002-of-00002.safetensors",
|
| 538 |
+
"coupled_vpm.vpm2.blocks.24.norm2.weight": "model-00002-of-00002.safetensors",
|
| 539 |
+
"coupled_vpm.vpm2.blocks.25.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 540 |
+
"coupled_vpm.vpm2.blocks.25.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 541 |
+
"coupled_vpm.vpm2.blocks.25.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 542 |
+
"coupled_vpm.vpm2.blocks.25.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 543 |
+
"coupled_vpm.vpm2.blocks.25.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 544 |
+
"coupled_vpm.vpm2.blocks.25.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 545 |
+
"coupled_vpm.vpm2.blocks.25.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 546 |
+
"coupled_vpm.vpm2.blocks.25.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 547 |
+
"coupled_vpm.vpm2.blocks.25.norm1.bias": "model-00002-of-00002.safetensors",
|
| 548 |
+
"coupled_vpm.vpm2.blocks.25.norm1.weight": "model-00002-of-00002.safetensors",
|
| 549 |
+
"coupled_vpm.vpm2.blocks.25.norm2.bias": "model-00002-of-00002.safetensors",
|
| 550 |
+
"coupled_vpm.vpm2.blocks.25.norm2.weight": "model-00002-of-00002.safetensors",
|
| 551 |
+
"coupled_vpm.vpm2.blocks.3.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 552 |
+
"coupled_vpm.vpm2.blocks.3.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 553 |
+
"coupled_vpm.vpm2.blocks.3.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 554 |
+
"coupled_vpm.vpm2.blocks.3.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 555 |
+
"coupled_vpm.vpm2.blocks.3.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 556 |
+
"coupled_vpm.vpm2.blocks.3.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 557 |
+
"coupled_vpm.vpm2.blocks.3.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 558 |
+
"coupled_vpm.vpm2.blocks.3.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 559 |
+
"coupled_vpm.vpm2.blocks.3.norm1.bias": "model-00002-of-00002.safetensors",
|
| 560 |
+
"coupled_vpm.vpm2.blocks.3.norm1.weight": "model-00002-of-00002.safetensors",
|
| 561 |
+
"coupled_vpm.vpm2.blocks.3.norm2.bias": "model-00002-of-00002.safetensors",
|
| 562 |
+
"coupled_vpm.vpm2.blocks.3.norm2.weight": "model-00002-of-00002.safetensors",
|
| 563 |
+
"coupled_vpm.vpm2.blocks.4.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 564 |
+
"coupled_vpm.vpm2.blocks.4.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 565 |
+
"coupled_vpm.vpm2.blocks.4.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 566 |
+
"coupled_vpm.vpm2.blocks.4.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 567 |
+
"coupled_vpm.vpm2.blocks.4.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 568 |
+
"coupled_vpm.vpm2.blocks.4.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 569 |
+
"coupled_vpm.vpm2.blocks.4.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 570 |
+
"coupled_vpm.vpm2.blocks.4.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 571 |
+
"coupled_vpm.vpm2.blocks.4.norm1.bias": "model-00002-of-00002.safetensors",
|
| 572 |
+
"coupled_vpm.vpm2.blocks.4.norm1.weight": "model-00002-of-00002.safetensors",
|
| 573 |
+
"coupled_vpm.vpm2.blocks.4.norm2.bias": "model-00002-of-00002.safetensors",
|
| 574 |
+
"coupled_vpm.vpm2.blocks.4.norm2.weight": "model-00002-of-00002.safetensors",
|
| 575 |
+
"coupled_vpm.vpm2.blocks.5.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 576 |
+
"coupled_vpm.vpm2.blocks.5.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 577 |
+
"coupled_vpm.vpm2.blocks.5.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 578 |
+
"coupled_vpm.vpm2.blocks.5.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 579 |
+
"coupled_vpm.vpm2.blocks.5.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 580 |
+
"coupled_vpm.vpm2.blocks.5.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 581 |
+
"coupled_vpm.vpm2.blocks.5.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 582 |
+
"coupled_vpm.vpm2.blocks.5.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 583 |
+
"coupled_vpm.vpm2.blocks.5.norm1.bias": "model-00002-of-00002.safetensors",
|
| 584 |
+
"coupled_vpm.vpm2.blocks.5.norm1.weight": "model-00002-of-00002.safetensors",
|
| 585 |
+
"coupled_vpm.vpm2.blocks.5.norm2.bias": "model-00002-of-00002.safetensors",
|
| 586 |
+
"coupled_vpm.vpm2.blocks.5.norm2.weight": "model-00002-of-00002.safetensors",
|
| 587 |
+
"coupled_vpm.vpm2.blocks.6.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 588 |
+
"coupled_vpm.vpm2.blocks.6.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 589 |
+
"coupled_vpm.vpm2.blocks.6.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 590 |
+
"coupled_vpm.vpm2.blocks.6.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 591 |
+
"coupled_vpm.vpm2.blocks.6.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 592 |
+
"coupled_vpm.vpm2.blocks.6.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 593 |
+
"coupled_vpm.vpm2.blocks.6.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 594 |
+
"coupled_vpm.vpm2.blocks.6.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 595 |
+
"coupled_vpm.vpm2.blocks.6.norm1.bias": "model-00002-of-00002.safetensors",
|
| 596 |
+
"coupled_vpm.vpm2.blocks.6.norm1.weight": "model-00002-of-00002.safetensors",
|
| 597 |
+
"coupled_vpm.vpm2.blocks.6.norm2.bias": "model-00002-of-00002.safetensors",
|
| 598 |
+
"coupled_vpm.vpm2.blocks.6.norm2.weight": "model-00002-of-00002.safetensors",
|
| 599 |
+
"coupled_vpm.vpm2.blocks.7.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 600 |
+
"coupled_vpm.vpm2.blocks.7.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 601 |
+
"coupled_vpm.vpm2.blocks.7.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 602 |
+
"coupled_vpm.vpm2.blocks.7.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 603 |
+
"coupled_vpm.vpm2.blocks.7.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 604 |
+
"coupled_vpm.vpm2.blocks.7.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 605 |
+
"coupled_vpm.vpm2.blocks.7.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 606 |
+
"coupled_vpm.vpm2.blocks.7.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 607 |
+
"coupled_vpm.vpm2.blocks.7.norm1.bias": "model-00002-of-00002.safetensors",
|
| 608 |
+
"coupled_vpm.vpm2.blocks.7.norm1.weight": "model-00002-of-00002.safetensors",
|
| 609 |
+
"coupled_vpm.vpm2.blocks.7.norm2.bias": "model-00002-of-00002.safetensors",
|
| 610 |
+
"coupled_vpm.vpm2.blocks.7.norm2.weight": "model-00002-of-00002.safetensors",
|
| 611 |
+
"coupled_vpm.vpm2.blocks.8.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 612 |
+
"coupled_vpm.vpm2.blocks.8.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 613 |
+
"coupled_vpm.vpm2.blocks.8.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 614 |
+
"coupled_vpm.vpm2.blocks.8.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 615 |
+
"coupled_vpm.vpm2.blocks.8.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 616 |
+
"coupled_vpm.vpm2.blocks.8.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 617 |
+
"coupled_vpm.vpm2.blocks.8.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 618 |
+
"coupled_vpm.vpm2.blocks.8.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 619 |
+
"coupled_vpm.vpm2.blocks.8.norm1.bias": "model-00002-of-00002.safetensors",
|
| 620 |
+
"coupled_vpm.vpm2.blocks.8.norm1.weight": "model-00002-of-00002.safetensors",
|
| 621 |
+
"coupled_vpm.vpm2.blocks.8.norm2.bias": "model-00002-of-00002.safetensors",
|
| 622 |
+
"coupled_vpm.vpm2.blocks.8.norm2.weight": "model-00002-of-00002.safetensors",
|
| 623 |
+
"coupled_vpm.vpm2.blocks.9.attn.proj.bias": "model-00002-of-00002.safetensors",
|
| 624 |
+
"coupled_vpm.vpm2.blocks.9.attn.proj.weight": "model-00002-of-00002.safetensors",
|
| 625 |
+
"coupled_vpm.vpm2.blocks.9.attn.qkv.bias": "model-00002-of-00002.safetensors",
|
| 626 |
+
"coupled_vpm.vpm2.blocks.9.attn.qkv.weight": "model-00002-of-00002.safetensors",
|
| 627 |
+
"coupled_vpm.vpm2.blocks.9.mlp.fc1.bias": "model-00002-of-00002.safetensors",
|
| 628 |
+
"coupled_vpm.vpm2.blocks.9.mlp.fc1.weight": "model-00002-of-00002.safetensors",
|
| 629 |
+
"coupled_vpm.vpm2.blocks.9.mlp.fc2.bias": "model-00002-of-00002.safetensors",
|
| 630 |
+
"coupled_vpm.vpm2.blocks.9.mlp.fc2.weight": "model-00002-of-00002.safetensors",
|
| 631 |
+
"coupled_vpm.vpm2.blocks.9.norm1.bias": "model-00002-of-00002.safetensors",
|
| 632 |
+
"coupled_vpm.vpm2.blocks.9.norm1.weight": "model-00002-of-00002.safetensors",
|
| 633 |
+
"coupled_vpm.vpm2.blocks.9.norm2.bias": "model-00002-of-00002.safetensors",
|
| 634 |
+
"coupled_vpm.vpm2.blocks.9.norm2.weight": "model-00002-of-00002.safetensors",
|
| 635 |
+
"coupled_vpm.vpm2.norm.bias": "model-00002-of-00002.safetensors",
|
| 636 |
+
"coupled_vpm.vpm2.norm.weight": "model-00002-of-00002.safetensors",
|
| 637 |
+
"coupled_vpm.vpm2.patch_embed.proj.bias": "model-00002-of-00002.safetensors",
|
| 638 |
+
"coupled_vpm.vpm2.patch_embed.proj.weight": "model-00002-of-00002.safetensors",
|
| 639 |
+
"coupled_vpm.vpm2.pos_embed": "model-00002-of-00002.safetensors",
|
| 640 |
+
"coupled_vpm.zero_linears.0.bias": "model-00002-of-00002.safetensors",
|
| 641 |
+
"coupled_vpm.zero_linears.0.weight": "model-00002-of-00002.safetensors",
|
| 642 |
+
"coupled_vpm.zero_linears.1.bias": "model-00002-of-00002.safetensors",
|
| 643 |
+
"coupled_vpm.zero_linears.1.weight": "model-00002-of-00002.safetensors",
|
| 644 |
+
"coupled_vpm.zero_linears.10.bias": "model-00002-of-00002.safetensors",
|
| 645 |
+
"coupled_vpm.zero_linears.10.weight": "model-00002-of-00002.safetensors",
|
| 646 |
+
"coupled_vpm.zero_linears.11.bias": "model-00002-of-00002.safetensors",
|
| 647 |
+
"coupled_vpm.zero_linears.11.weight": "model-00002-of-00002.safetensors",
|
| 648 |
+
"coupled_vpm.zero_linears.12.bias": "model-00002-of-00002.safetensors",
|
| 649 |
+
"coupled_vpm.zero_linears.12.weight": "model-00002-of-00002.safetensors",
|
| 650 |
+
"coupled_vpm.zero_linears.13.bias": "model-00002-of-00002.safetensors",
|
| 651 |
+
"coupled_vpm.zero_linears.13.weight": "model-00002-of-00002.safetensors",
|
| 652 |
+
"coupled_vpm.zero_linears.14.bias": "model-00002-of-00002.safetensors",
|
| 653 |
+
"coupled_vpm.zero_linears.14.weight": "model-00002-of-00002.safetensors",
|
| 654 |
+
"coupled_vpm.zero_linears.15.bias": "model-00002-of-00002.safetensors",
|
| 655 |
+
"coupled_vpm.zero_linears.15.weight": "model-00002-of-00002.safetensors",
|
| 656 |
+
"coupled_vpm.zero_linears.16.bias": "model-00002-of-00002.safetensors",
|
| 657 |
+
"coupled_vpm.zero_linears.16.weight": "model-00002-of-00002.safetensors",
|
| 658 |
+
"coupled_vpm.zero_linears.17.bias": "model-00002-of-00002.safetensors",
|
| 659 |
+
"coupled_vpm.zero_linears.17.weight": "model-00002-of-00002.safetensors",
|
| 660 |
+
"coupled_vpm.zero_linears.18.bias": "model-00002-of-00002.safetensors",
|
| 661 |
+
"coupled_vpm.zero_linears.18.weight": "model-00002-of-00002.safetensors",
|
| 662 |
+
"coupled_vpm.zero_linears.19.bias": "model-00002-of-00002.safetensors",
|
| 663 |
+
"coupled_vpm.zero_linears.19.weight": "model-00002-of-00002.safetensors",
|
| 664 |
+
"coupled_vpm.zero_linears.2.bias": "model-00002-of-00002.safetensors",
|
| 665 |
+
"coupled_vpm.zero_linears.2.weight": "model-00002-of-00002.safetensors",
|
| 666 |
+
"coupled_vpm.zero_linears.20.bias": "model-00002-of-00002.safetensors",
|
| 667 |
+
"coupled_vpm.zero_linears.20.weight": "model-00002-of-00002.safetensors",
|
| 668 |
+
"coupled_vpm.zero_linears.21.bias": "model-00002-of-00002.safetensors",
|
| 669 |
+
"coupled_vpm.zero_linears.21.weight": "model-00002-of-00002.safetensors",
|
| 670 |
+
"coupled_vpm.zero_linears.22.bias": "model-00002-of-00002.safetensors",
|
| 671 |
+
"coupled_vpm.zero_linears.22.weight": "model-00002-of-00002.safetensors",
|
| 672 |
+
"coupled_vpm.zero_linears.23.bias": "model-00002-of-00002.safetensors",
|
| 673 |
+
"coupled_vpm.zero_linears.23.weight": "model-00002-of-00002.safetensors",
|
| 674 |
+
"coupled_vpm.zero_linears.24.bias": "model-00002-of-00002.safetensors",
|
| 675 |
+
"coupled_vpm.zero_linears.24.weight": "model-00002-of-00002.safetensors",
|
| 676 |
+
"coupled_vpm.zero_linears.25.bias": "model-00002-of-00002.safetensors",
|
| 677 |
+
"coupled_vpm.zero_linears.25.weight": "model-00002-of-00002.safetensors",
|
| 678 |
+
"coupled_vpm.zero_linears.3.bias": "model-00002-of-00002.safetensors",
|
| 679 |
+
"coupled_vpm.zero_linears.3.weight": "model-00002-of-00002.safetensors",
|
| 680 |
+
"coupled_vpm.zero_linears.4.bias": "model-00002-of-00002.safetensors",
|
| 681 |
+
"coupled_vpm.zero_linears.4.weight": "model-00002-of-00002.safetensors",
|
| 682 |
+
"coupled_vpm.zero_linears.5.bias": "model-00002-of-00002.safetensors",
|
| 683 |
+
"coupled_vpm.zero_linears.5.weight": "model-00002-of-00002.safetensors",
|
| 684 |
+
"coupled_vpm.zero_linears.6.bias": "model-00002-of-00002.safetensors",
|
| 685 |
+
"coupled_vpm.zero_linears.6.weight": "model-00002-of-00002.safetensors",
|
| 686 |
+
"coupled_vpm.zero_linears.7.bias": "model-00002-of-00002.safetensors",
|
| 687 |
+
"coupled_vpm.zero_linears.7.weight": "model-00002-of-00002.safetensors",
|
| 688 |
+
"coupled_vpm.zero_linears.8.bias": "model-00002-of-00002.safetensors",
|
| 689 |
+
"coupled_vpm.zero_linears.8.weight": "model-00002-of-00002.safetensors",
|
| 690 |
+
"coupled_vpm.zero_linears.9.bias": "model-00002-of-00002.safetensors",
|
| 691 |
+
"coupled_vpm.zero_linears.9.weight": "model-00002-of-00002.safetensors",
|
| 692 |
+
"llm.lm_head.weight": "model-00002-of-00002.safetensors",
|
| 693 |
+
"llm.model.embed_tokens.weight": "model-00001-of-00002.safetensors",
|
| 694 |
+
"llm.model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 695 |
+
"llm.model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 696 |
+
"llm.model.layers.0.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 697 |
+
"llm.model.layers.0.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 698 |
+
"llm.model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 699 |
+
"llm.model.layers.0.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 700 |
+
"llm.model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 701 |
+
"llm.model.layers.0.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 702 |
+
"llm.model.layers.0.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 703 |
+
"llm.model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 704 |
+
"llm.model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 705 |
+
"llm.model.layers.1.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 706 |
+
"llm.model.layers.1.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 707 |
+
"llm.model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 708 |
+
"llm.model.layers.1.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 709 |
+
"llm.model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 710 |
+
"llm.model.layers.1.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 711 |
+
"llm.model.layers.1.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 712 |
+
"llm.model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 713 |
+
"llm.model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 714 |
+
"llm.model.layers.10.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 715 |
+
"llm.model.layers.10.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 716 |
+
"llm.model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 717 |
+
"llm.model.layers.10.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 718 |
+
"llm.model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 719 |
+
"llm.model.layers.10.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 720 |
+
"llm.model.layers.10.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 721 |
+
"llm.model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 722 |
+
"llm.model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 723 |
+
"llm.model.layers.11.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 724 |
+
"llm.model.layers.11.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 725 |
+
"llm.model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 726 |
+
"llm.model.layers.11.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 727 |
+
"llm.model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 728 |
+
"llm.model.layers.11.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 729 |
+
"llm.model.layers.11.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 730 |
+
"llm.model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 731 |
+
"llm.model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 732 |
+
"llm.model.layers.12.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 733 |
+
"llm.model.layers.12.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 734 |
+
"llm.model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 735 |
+
"llm.model.layers.12.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 736 |
+
"llm.model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 737 |
+
"llm.model.layers.12.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 738 |
+
"llm.model.layers.12.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 739 |
+
"llm.model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 740 |
+
"llm.model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 741 |
+
"llm.model.layers.13.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 742 |
+
"llm.model.layers.13.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 743 |
+
"llm.model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 744 |
+
"llm.model.layers.13.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 745 |
+
"llm.model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 746 |
+
"llm.model.layers.13.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 747 |
+
"llm.model.layers.13.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 748 |
+
"llm.model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 749 |
+
"llm.model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 750 |
+
"llm.model.layers.14.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 751 |
+
"llm.model.layers.14.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 752 |
+
"llm.model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 753 |
+
"llm.model.layers.14.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 754 |
+
"llm.model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 755 |
+
"llm.model.layers.14.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 756 |
+
"llm.model.layers.14.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 757 |
+
"llm.model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 758 |
+
"llm.model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 759 |
+
"llm.model.layers.15.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 760 |
+
"llm.model.layers.15.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 761 |
+
"llm.model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 762 |
+
"llm.model.layers.15.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 763 |
+
"llm.model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 764 |
+
"llm.model.layers.15.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 765 |
+
"llm.model.layers.15.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 766 |
+
"llm.model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 767 |
+
"llm.model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 768 |
+
"llm.model.layers.16.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 769 |
+
"llm.model.layers.16.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 770 |
+
"llm.model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 771 |
+
"llm.model.layers.16.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 772 |
+
"llm.model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 773 |
+
"llm.model.layers.16.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 774 |
+
"llm.model.layers.16.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 775 |
+
"llm.model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 776 |
+
"llm.model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 777 |
+
"llm.model.layers.17.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 778 |
+
"llm.model.layers.17.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 779 |
+
"llm.model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 780 |
+
"llm.model.layers.17.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 781 |
+
"llm.model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 782 |
+
"llm.model.layers.17.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 783 |
+
"llm.model.layers.17.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 784 |
+
"llm.model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 785 |
+
"llm.model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 786 |
+
"llm.model.layers.18.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 787 |
+
"llm.model.layers.18.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 788 |
+
"llm.model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 789 |
+
"llm.model.layers.18.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 790 |
+
"llm.model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 791 |
+
"llm.model.layers.18.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 792 |
+
"llm.model.layers.18.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 793 |
+
"llm.model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 794 |
+
"llm.model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 795 |
+
"llm.model.layers.19.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 796 |
+
"llm.model.layers.19.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 797 |
+
"llm.model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 798 |
+
"llm.model.layers.19.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 799 |
+
"llm.model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 800 |
+
"llm.model.layers.19.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 801 |
+
"llm.model.layers.19.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 802 |
+
"llm.model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 803 |
+
"llm.model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 804 |
+
"llm.model.layers.2.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 805 |
+
"llm.model.layers.2.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 806 |
+
"llm.model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 807 |
+
"llm.model.layers.2.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 808 |
+
"llm.model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 809 |
+
"llm.model.layers.2.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 810 |
+
"llm.model.layers.2.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 811 |
+
"llm.model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 812 |
+
"llm.model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 813 |
+
"llm.model.layers.20.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 814 |
+
"llm.model.layers.20.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 815 |
+
"llm.model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 816 |
+
"llm.model.layers.20.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 817 |
+
"llm.model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 818 |
+
"llm.model.layers.20.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 819 |
+
"llm.model.layers.20.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 820 |
+
"llm.model.layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 821 |
+
"llm.model.layers.21.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 822 |
+
"llm.model.layers.21.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 823 |
+
"llm.model.layers.21.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 824 |
+
"llm.model.layers.21.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 825 |
+
"llm.model.layers.21.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 826 |
+
"llm.model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 827 |
+
"llm.model.layers.21.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 828 |
+
"llm.model.layers.21.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 829 |
+
"llm.model.layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 830 |
+
"llm.model.layers.22.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 831 |
+
"llm.model.layers.22.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 832 |
+
"llm.model.layers.22.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 833 |
+
"llm.model.layers.22.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 834 |
+
"llm.model.layers.22.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 835 |
+
"llm.model.layers.22.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 836 |
+
"llm.model.layers.22.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 837 |
+
"llm.model.layers.22.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 838 |
+
"llm.model.layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 839 |
+
"llm.model.layers.23.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 840 |
+
"llm.model.layers.23.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 841 |
+
"llm.model.layers.23.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 842 |
+
"llm.model.layers.23.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 843 |
+
"llm.model.layers.23.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 844 |
+
"llm.model.layers.23.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 845 |
+
"llm.model.layers.23.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 846 |
+
"llm.model.layers.23.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 847 |
+
"llm.model.layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 848 |
+
"llm.model.layers.24.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 849 |
+
"llm.model.layers.24.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 850 |
+
"llm.model.layers.24.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 851 |
+
"llm.model.layers.24.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 852 |
+
"llm.model.layers.24.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 853 |
+
"llm.model.layers.24.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 854 |
+
"llm.model.layers.24.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 855 |
+
"llm.model.layers.24.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 856 |
+
"llm.model.layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 857 |
+
"llm.model.layers.25.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 858 |
+
"llm.model.layers.25.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 859 |
+
"llm.model.layers.25.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 860 |
+
"llm.model.layers.25.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 861 |
+
"llm.model.layers.25.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 862 |
+
"llm.model.layers.25.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 863 |
+
"llm.model.layers.25.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 864 |
+
"llm.model.layers.25.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 865 |
+
"llm.model.layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 866 |
+
"llm.model.layers.26.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 867 |
+
"llm.model.layers.26.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 868 |
+
"llm.model.layers.26.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 869 |
+
"llm.model.layers.26.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 870 |
+
"llm.model.layers.26.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 871 |
+
"llm.model.layers.26.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 872 |
+
"llm.model.layers.26.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 873 |
+
"llm.model.layers.26.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 874 |
+
"llm.model.layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 875 |
+
"llm.model.layers.27.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 876 |
+
"llm.model.layers.27.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 877 |
+
"llm.model.layers.27.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 878 |
+
"llm.model.layers.27.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 879 |
+
"llm.model.layers.27.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 880 |
+
"llm.model.layers.27.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 881 |
+
"llm.model.layers.27.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 882 |
+
"llm.model.layers.27.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 883 |
+
"llm.model.layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 884 |
+
"llm.model.layers.28.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 885 |
+
"llm.model.layers.28.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 886 |
+
"llm.model.layers.28.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 887 |
+
"llm.model.layers.28.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 888 |
+
"llm.model.layers.28.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 889 |
+
"llm.model.layers.28.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 890 |
+
"llm.model.layers.28.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 891 |
+
"llm.model.layers.28.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 892 |
+
"llm.model.layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 893 |
+
"llm.model.layers.29.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 894 |
+
"llm.model.layers.29.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 895 |
+
"llm.model.layers.29.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 896 |
+
"llm.model.layers.29.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 897 |
+
"llm.model.layers.29.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 898 |
+
"llm.model.layers.29.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 899 |
+
"llm.model.layers.29.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 900 |
+
"llm.model.layers.29.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 901 |
+
"llm.model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 902 |
+
"llm.model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 903 |
+
"llm.model.layers.3.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 904 |
+
"llm.model.layers.3.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 905 |
+
"llm.model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 906 |
+
"llm.model.layers.3.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 907 |
+
"llm.model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 908 |
+
"llm.model.layers.3.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 909 |
+
"llm.model.layers.3.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 910 |
+
"llm.model.layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 911 |
+
"llm.model.layers.30.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 912 |
+
"llm.model.layers.30.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 913 |
+
"llm.model.layers.30.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 914 |
+
"llm.model.layers.30.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 915 |
+
"llm.model.layers.30.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 916 |
+
"llm.model.layers.30.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 917 |
+
"llm.model.layers.30.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 918 |
+
"llm.model.layers.30.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 919 |
+
"llm.model.layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 920 |
+
"llm.model.layers.31.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 921 |
+
"llm.model.layers.31.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 922 |
+
"llm.model.layers.31.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 923 |
+
"llm.model.layers.31.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 924 |
+
"llm.model.layers.31.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 925 |
+
"llm.model.layers.31.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 926 |
+
"llm.model.layers.31.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 927 |
+
"llm.model.layers.31.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 928 |
+
"llm.model.layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 929 |
+
"llm.model.layers.32.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 930 |
+
"llm.model.layers.32.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 931 |
+
"llm.model.layers.32.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 932 |
+
"llm.model.layers.32.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 933 |
+
"llm.model.layers.32.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 934 |
+
"llm.model.layers.32.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 935 |
+
"llm.model.layers.32.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 936 |
+
"llm.model.layers.32.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 937 |
+
"llm.model.layers.33.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 938 |
+
"llm.model.layers.33.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 939 |
+
"llm.model.layers.33.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 940 |
+
"llm.model.layers.33.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 941 |
+
"llm.model.layers.33.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 942 |
+
"llm.model.layers.33.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 943 |
+
"llm.model.layers.33.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 944 |
+
"llm.model.layers.33.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 945 |
+
"llm.model.layers.33.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 946 |
+
"llm.model.layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 947 |
+
"llm.model.layers.34.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 948 |
+
"llm.model.layers.34.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 949 |
+
"llm.model.layers.34.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 950 |
+
"llm.model.layers.34.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 951 |
+
"llm.model.layers.34.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 952 |
+
"llm.model.layers.34.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 953 |
+
"llm.model.layers.34.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 954 |
+
"llm.model.layers.34.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 955 |
+
"llm.model.layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 956 |
+
"llm.model.layers.35.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 957 |
+
"llm.model.layers.35.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 958 |
+
"llm.model.layers.35.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 959 |
+
"llm.model.layers.35.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 960 |
+
"llm.model.layers.35.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 961 |
+
"llm.model.layers.35.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 962 |
+
"llm.model.layers.35.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 963 |
+
"llm.model.layers.35.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 964 |
+
"llm.model.layers.36.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 965 |
+
"llm.model.layers.36.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 966 |
+
"llm.model.layers.36.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 967 |
+
"llm.model.layers.36.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 968 |
+
"llm.model.layers.36.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 969 |
+
"llm.model.layers.36.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 970 |
+
"llm.model.layers.36.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 971 |
+
"llm.model.layers.36.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 972 |
+
"llm.model.layers.36.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 973 |
+
"llm.model.layers.37.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 974 |
+
"llm.model.layers.37.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 975 |
+
"llm.model.layers.37.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 976 |
+
"llm.model.layers.37.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 977 |
+
"llm.model.layers.37.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 978 |
+
"llm.model.layers.37.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 979 |
+
"llm.model.layers.37.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 980 |
+
"llm.model.layers.37.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 981 |
+
"llm.model.layers.37.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 982 |
+
"llm.model.layers.38.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 983 |
+
"llm.model.layers.38.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 984 |
+
"llm.model.layers.38.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 985 |
+
"llm.model.layers.38.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 986 |
+
"llm.model.layers.38.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 987 |
+
"llm.model.layers.38.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 988 |
+
"llm.model.layers.38.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 989 |
+
"llm.model.layers.38.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 990 |
+
"llm.model.layers.38.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 991 |
+
"llm.model.layers.39.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 992 |
+
"llm.model.layers.39.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
| 993 |
+
"llm.model.layers.39.mlp.gate_proj.weight": "model-00002-of-00002.safetensors",
|
| 994 |
+
"llm.model.layers.39.mlp.up_proj.weight": "model-00002-of-00002.safetensors",
|
| 995 |
+
"llm.model.layers.39.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
| 996 |
+
"llm.model.layers.39.self_attn.k_proj.weight": "model-00002-of-00002.safetensors",
|
| 997 |
+
"llm.model.layers.39.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
| 998 |
+
"llm.model.layers.39.self_attn.q_proj.weight": "model-00002-of-00002.safetensors",
|
| 999 |
+
"llm.model.layers.39.self_attn.v_proj.weight": "model-00002-of-00002.safetensors",
|
| 1000 |
+
"llm.model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1001 |
+
"llm.model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1002 |
+
"llm.model.layers.4.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1003 |
+
"llm.model.layers.4.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1004 |
+
"llm.model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1005 |
+
"llm.model.layers.4.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1006 |
+
"llm.model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1007 |
+
"llm.model.layers.4.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1008 |
+
"llm.model.layers.4.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1009 |
+
"llm.model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1010 |
+
"llm.model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1011 |
+
"llm.model.layers.5.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1012 |
+
"llm.model.layers.5.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1013 |
+
"llm.model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1014 |
+
"llm.model.layers.5.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1015 |
+
"llm.model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1016 |
+
"llm.model.layers.5.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1017 |
+
"llm.model.layers.5.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1018 |
+
"llm.model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1019 |
+
"llm.model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1020 |
+
"llm.model.layers.6.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1021 |
+
"llm.model.layers.6.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1022 |
+
"llm.model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1023 |
+
"llm.model.layers.6.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1024 |
+
"llm.model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1025 |
+
"llm.model.layers.6.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1026 |
+
"llm.model.layers.6.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1027 |
+
"llm.model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1028 |
+
"llm.model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1029 |
+
"llm.model.layers.7.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1030 |
+
"llm.model.layers.7.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1031 |
+
"llm.model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1032 |
+
"llm.model.layers.7.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1033 |
+
"llm.model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1034 |
+
"llm.model.layers.7.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1035 |
+
"llm.model.layers.7.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1036 |
+
"llm.model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1037 |
+
"llm.model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1038 |
+
"llm.model.layers.8.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1039 |
+
"llm.model.layers.8.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1040 |
+
"llm.model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1041 |
+
"llm.model.layers.8.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1042 |
+
"llm.model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1043 |
+
"llm.model.layers.8.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1044 |
+
"llm.model.layers.8.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1045 |
+
"llm.model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1046 |
+
"llm.model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
| 1047 |
+
"llm.model.layers.9.mlp.gate_proj.weight": "model-00001-of-00002.safetensors",
|
| 1048 |
+
"llm.model.layers.9.mlp.up_proj.weight": "model-00001-of-00002.safetensors",
|
| 1049 |
+
"llm.model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
| 1050 |
+
"llm.model.layers.9.self_attn.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 1051 |
+
"llm.model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
| 1052 |
+
"llm.model.layers.9.self_attn.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 1053 |
+
"llm.model.layers.9.self_attn.v_proj.weight": "model-00001-of-00002.safetensors",
|
| 1054 |
+
"llm.model.norm.weight": "model-00002-of-00002.safetensors",
|
| 1055 |
+
"resampler.attn.in_proj_bias": "model-00002-of-00002.safetensors",
|
| 1056 |
+
"resampler.attn.in_proj_weight": "model-00002-of-00002.safetensors",
|
| 1057 |
+
"resampler.attn.out_proj.bias": "model-00002-of-00002.safetensors",
|
| 1058 |
+
"resampler.attn.out_proj.weight": "model-00002-of-00002.safetensors",
|
| 1059 |
+
"resampler.kv_proj.weight": "model-00002-of-00002.safetensors",
|
| 1060 |
+
"resampler.ln_kv.bias": "model-00002-of-00002.safetensors",
|
| 1061 |
+
"resampler.ln_kv.weight": "model-00002-of-00002.safetensors",
|
| 1062 |
+
"resampler.ln_post.bias": "model-00002-of-00002.safetensors",
|
| 1063 |
+
"resampler.ln_post.weight": "model-00002-of-00002.safetensors",
|
| 1064 |
+
"resampler.ln_q.bias": "model-00002-of-00002.safetensors",
|
| 1065 |
+
"resampler.ln_q.weight": "model-00002-of-00002.safetensors",
|
| 1066 |
+
"resampler.pos_embed": "model-00002-of-00002.safetensors",
|
| 1067 |
+
"resampler.proj": "model-00002-of-00002.safetensors",
|
| 1068 |
+
"resampler.query": "model-00002-of-00002.safetensors"
|
| 1069 |
+
}
|
| 1070 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<image>",
|
| 4 |
+
"</image>",
|
| 5 |
+
"<ref>",
|
| 6 |
+
"</ref>",
|
| 7 |
+
"<box>",
|
| 8 |
+
"</box>",
|
| 9 |
+
"<quad>",
|
| 10 |
+
"</quad>",
|
| 11 |
+
"<point>",
|
| 12 |
+
"</point>",
|
| 13 |
+
"<slice>",
|
| 14 |
+
"</slice>"
|
| 15 |
+
],
|
| 16 |
+
"bos_token": {
|
| 17 |
+
"content": "<s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"eos_token": {
|
| 24 |
+
"content": "</s>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"pad_token": {
|
| 31 |
+
"content": "<unk>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"unk_token": {
|
| 38 |
+
"content": "<unk>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false
|
| 43 |
+
}
|
| 44 |
+
}
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b046f377b2f09a90623a46611bd5d23cb76ac89ca7804a372408f37aff7b96a0
|
| 3 |
+
size 1994817
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,161 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": true,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
},
|
| 30 |
+
"101": {
|
| 31 |
+
"content": "<image>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
},
|
| 38 |
+
"102": {
|
| 39 |
+
"content": "</image>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": true
|
| 45 |
+
},
|
| 46 |
+
"103": {
|
| 47 |
+
"content": "<ref>",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": false,
|
| 50 |
+
"rstrip": false,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": true
|
| 53 |
+
},
|
| 54 |
+
"104": {
|
| 55 |
+
"content": "</ref>",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": false,
|
| 58 |
+
"rstrip": false,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": true
|
| 61 |
+
},
|
| 62 |
+
"105": {
|
| 63 |
+
"content": "<box>",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": false,
|
| 66 |
+
"rstrip": false,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": true
|
| 69 |
+
},
|
| 70 |
+
"106": {
|
| 71 |
+
"content": "</box>",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": false,
|
| 74 |
+
"rstrip": false,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": true
|
| 77 |
+
},
|
| 78 |
+
"107": {
|
| 79 |
+
"content": "<quad>",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": false,
|
| 82 |
+
"rstrip": false,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": true
|
| 85 |
+
},
|
| 86 |
+
"108": {
|
| 87 |
+
"content": "</quad>",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": false,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": true
|
| 93 |
+
},
|
| 94 |
+
"109": {
|
| 95 |
+
"content": "<point>",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": false,
|
| 98 |
+
"rstrip": false,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": true
|
| 101 |
+
},
|
| 102 |
+
"110": {
|
| 103 |
+
"content": "</point>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": true
|
| 109 |
+
},
|
| 110 |
+
"111": {
|
| 111 |
+
"content": "<slice>",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": false,
|
| 114 |
+
"rstrip": false,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": true
|
| 117 |
+
},
|
| 118 |
+
"112": {
|
| 119 |
+
"content": "</slice>",
|
| 120 |
+
"lstrip": false,
|
| 121 |
+
"normalized": false,
|
| 122 |
+
"rstrip": false,
|
| 123 |
+
"single_word": false,
|
| 124 |
+
"special": true
|
| 125 |
+
}
|
| 126 |
+
},
|
| 127 |
+
"additional_special_tokens": [
|
| 128 |
+
"<image>",
|
| 129 |
+
"</image>",
|
| 130 |
+
"<ref>",
|
| 131 |
+
"</ref>",
|
| 132 |
+
"<box>",
|
| 133 |
+
"</box>",
|
| 134 |
+
"<quad>",
|
| 135 |
+
"</quad>",
|
| 136 |
+
"<point>",
|
| 137 |
+
"</point>",
|
| 138 |
+
"<slice>",
|
| 139 |
+
"</slice>"
|
| 140 |
+
],
|
| 141 |
+
"auto_map": {
|
| 142 |
+
"AutoTokenizer": [
|
| 143 |
+
"tokenizer.LlamaTokenizerWrapper",
|
| 144 |
+
null
|
| 145 |
+
]
|
| 146 |
+
},
|
| 147 |
+
"bos_token": "<s>",
|
| 148 |
+
"clean_up_tokenization_spaces": false,
|
| 149 |
+
"eos_token": "</s>",
|
| 150 |
+
"legacy": true,
|
| 151 |
+
"model_max_length": 2048,
|
| 152 |
+
"pad_token": "<unk>",
|
| 153 |
+
"padding_side": "right",
|
| 154 |
+
"sp_model_kwargs": {},
|
| 155 |
+
"spaces_between_special_tokens": false,
|
| 156 |
+
"tokenizer_class": "LlamaTokenizerWrapper",
|
| 157 |
+
"truncation_side": "right",
|
| 158 |
+
"unk_token": "<unk>",
|
| 159 |
+
"use_default_system_prompt": false,
|
| 160 |
+
"use_fast": false
|
| 161 |
+
}
|