Isaak Carter Augustus
commited on
Commit
•
dd33b23
1
Parent(s):
b5d6f4b
Upload folder using huggingface_hub
Browse files- README.md +391 -0
- config.json +34 -0
- mergekit_moe_config.yml +205 -0
- model-00001-of-00001.safetensors +3 -0
- model.safetensors.index.json +1 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +43 -0
README.md
ADDED
|
@@ -0,0 +1,391 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- moe
|
| 5 |
+
- frankenmoe
|
| 6 |
+
- merge
|
| 7 |
+
- mergekit
|
| 8 |
+
- lazymergekit
|
| 9 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 10 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 11 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 12 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 13 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 14 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 15 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 16 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 17 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 18 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 19 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 20 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 21 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 22 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 23 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 24 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 25 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 26 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 27 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 28 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 29 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 30 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 31 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 32 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 33 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 34 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 35 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 36 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 37 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 38 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 39 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 40 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 41 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 42 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 43 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 44 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 45 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 46 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 47 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 48 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 49 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 50 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 51 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 52 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 53 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 54 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 55 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 56 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 57 |
+
base_model:
|
| 58 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 59 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 60 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 61 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 62 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 63 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 64 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 65 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 66 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 67 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 68 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 69 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 70 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 71 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 72 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 73 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 74 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 75 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 76 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 77 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 78 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 79 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 80 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 81 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 82 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 83 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 84 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 85 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 86 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 87 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 88 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 89 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 90 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 91 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 92 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 93 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 94 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 95 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 96 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 97 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 98 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 99 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 100 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 101 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 102 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 103 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 104 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 105 |
+
- Felladrin/Smol-Llama-101M-Chat-v1
|
| 106 |
+
---
|
| 107 |
+
|
| 108 |
+
# SmalJ.O.S.I.E.-48x101M-Chat
|
| 109 |
+
|
| 110 |
+
SmalJ.O.S.I.E.-48x101M-Chat is a Mixure of Experts (MoE) made with the following models using [LazyMergekit](https://colab.research.google.com/drive/1obulZ1ROXHjYLn6PPZJwRR6GzgQogxxb?usp=sharing):
|
| 111 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 112 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 113 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 114 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 115 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 116 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 117 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 118 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 119 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 120 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 121 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 122 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 123 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 124 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 125 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 126 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 127 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 128 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 129 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 130 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 131 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 132 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 133 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 134 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 135 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 136 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 137 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 138 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 139 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 140 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 141 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 142 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 143 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 144 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 145 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 146 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 147 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 148 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 149 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 150 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 151 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 152 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 153 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 154 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 155 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 156 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 157 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 158 |
+
* [Felladrin/Smol-Llama-101M-Chat-v1](https://huggingface.co/Felladrin/Smol-Llama-101M-Chat-v1)
|
| 159 |
+
|
| 160 |
+
## 🧩 Configuration
|
| 161 |
+
|
| 162 |
+
```yamlbase_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 163 |
+
dtype: float32
|
| 164 |
+
gate_mode: hidden
|
| 165 |
+
experts:
|
| 166 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 167 |
+
positive_prompts:
|
| 168 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 169 |
+
|
| 170 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 171 |
+
positive_prompts:
|
| 172 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 173 |
+
|
| 174 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 175 |
+
positive_prompts:
|
| 176 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 177 |
+
|
| 178 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 179 |
+
positive_prompts:
|
| 180 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 181 |
+
|
| 182 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 183 |
+
positive_prompts:
|
| 184 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 185 |
+
|
| 186 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 187 |
+
positive_prompts:
|
| 188 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 189 |
+
|
| 190 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 191 |
+
positive_prompts:
|
| 192 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 193 |
+
|
| 194 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 195 |
+
positive_prompts:
|
| 196 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 197 |
+
|
| 198 |
+
|
| 199 |
+
|
| 200 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 201 |
+
positive_prompts:
|
| 202 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 203 |
+
|
| 204 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 205 |
+
positive_prompts:
|
| 206 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 207 |
+
|
| 208 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 209 |
+
positive_prompts:
|
| 210 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 211 |
+
|
| 212 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 213 |
+
positive_prompts:
|
| 214 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 215 |
+
|
| 216 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 217 |
+
positive_prompts:
|
| 218 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 219 |
+
|
| 220 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 221 |
+
positive_prompts:
|
| 222 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 223 |
+
|
| 224 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 225 |
+
positive_prompts:
|
| 226 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 227 |
+
|
| 228 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 229 |
+
positive_prompts:
|
| 230 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 231 |
+
|
| 232 |
+
|
| 233 |
+
|
| 234 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 235 |
+
positive_prompts:
|
| 236 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 237 |
+
|
| 238 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 239 |
+
positive_prompts:
|
| 240 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 241 |
+
|
| 242 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 243 |
+
positive_prompts:
|
| 244 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 245 |
+
|
| 246 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 247 |
+
positive_prompts:
|
| 248 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 249 |
+
|
| 250 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 251 |
+
positive_prompts:
|
| 252 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 253 |
+
|
| 254 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 255 |
+
positive_prompts:
|
| 256 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 257 |
+
|
| 258 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 259 |
+
positive_prompts:
|
| 260 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 261 |
+
|
| 262 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 263 |
+
positive_prompts:
|
| 264 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 265 |
+
|
| 266 |
+
|
| 267 |
+
|
| 268 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 269 |
+
positive_prompts:
|
| 270 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 271 |
+
|
| 272 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 273 |
+
positive_prompts:
|
| 274 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 275 |
+
|
| 276 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 277 |
+
positive_prompts:
|
| 278 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 279 |
+
|
| 280 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 281 |
+
positive_prompts:
|
| 282 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 283 |
+
|
| 284 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 285 |
+
positive_prompts:
|
| 286 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 287 |
+
|
| 288 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 289 |
+
positive_prompts:
|
| 290 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 291 |
+
|
| 292 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 293 |
+
positive_prompts:
|
| 294 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 295 |
+
|
| 296 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 297 |
+
positive_prompts:
|
| 298 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 299 |
+
|
| 300 |
+
|
| 301 |
+
|
| 302 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 303 |
+
positive_prompts:
|
| 304 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 305 |
+
|
| 306 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 307 |
+
positive_prompts:
|
| 308 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 309 |
+
|
| 310 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 311 |
+
positive_prompts:
|
| 312 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 313 |
+
|
| 314 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 315 |
+
positive_prompts:
|
| 316 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 317 |
+
|
| 318 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 319 |
+
positive_prompts:
|
| 320 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 321 |
+
|
| 322 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 323 |
+
positive_prompts:
|
| 324 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 325 |
+
|
| 326 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 327 |
+
positive_prompts:
|
| 328 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 329 |
+
|
| 330 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 331 |
+
positive_prompts:
|
| 332 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 333 |
+
|
| 334 |
+
|
| 335 |
+
|
| 336 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 337 |
+
positive_prompts:
|
| 338 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 339 |
+
|
| 340 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 341 |
+
positive_prompts:
|
| 342 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 343 |
+
|
| 344 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 345 |
+
positive_prompts:
|
| 346 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 347 |
+
|
| 348 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 349 |
+
positive_prompts:
|
| 350 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 351 |
+
|
| 352 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 353 |
+
positive_prompts:
|
| 354 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 355 |
+
|
| 356 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 357 |
+
positive_prompts:
|
| 358 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 359 |
+
|
| 360 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 361 |
+
positive_prompts:
|
| 362 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 363 |
+
|
| 364 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 365 |
+
positive_prompts:
|
| 366 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 367 |
+
```
|
| 368 |
+
|
| 369 |
+
## 💻 Usage
|
| 370 |
+
|
| 371 |
+
```python
|
| 372 |
+
!pip install -qU transformers bitsandbytes accelerate
|
| 373 |
+
|
| 374 |
+
from transformers import AutoTokenizer
|
| 375 |
+
import transformers
|
| 376 |
+
import torch
|
| 377 |
+
|
| 378 |
+
model = "Isaak-Carter/SmalJ.O.S.I.E.-48x101M-Chat"
|
| 379 |
+
|
| 380 |
+
tokenizer = AutoTokenizer.from_pretrained(model)
|
| 381 |
+
pipeline = transformers.pipeline(
|
| 382 |
+
"text-generation",
|
| 383 |
+
model=model,
|
| 384 |
+
model_kwargs={"torch_dtype": torch.float16, "load_in_4bit": True},
|
| 385 |
+
)
|
| 386 |
+
|
| 387 |
+
messages = [{"role": "user", "content": "Explain what a Mixture of Experts is in less than 100 words."}]
|
| 388 |
+
prompt = pipeline.tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
|
| 389 |
+
outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
|
| 390 |
+
print(outputs[0]["generated_text"])
|
| 391 |
+
```
|
config.json
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "Felladrin/Smol-Llama-101M-Chat-v1",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MixtralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 768,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 3072,
|
| 14 |
+
"max_position_embeddings": 1024,
|
| 15 |
+
"model_type": "mixtral",
|
| 16 |
+
"num_attention_heads": 24,
|
| 17 |
+
"num_experts_per_tok": 2,
|
| 18 |
+
"num_hidden_layers": 6,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"num_local_experts": 48,
|
| 21 |
+
"output_router_logits": false,
|
| 22 |
+
"pad_token_id": 2,
|
| 23 |
+
"pretraining_tp": 1,
|
| 24 |
+
"rms_norm_eps": 1e-05,
|
| 25 |
+
"rope_scaling": null,
|
| 26 |
+
"rope_theta": 10000.0,
|
| 27 |
+
"router_aux_loss_coef": 0.001,
|
| 28 |
+
"sliding_window": null,
|
| 29 |
+
"tie_word_embeddings": false,
|
| 30 |
+
"torch_dtype": "float32",
|
| 31 |
+
"transformers_version": "4.38.2",
|
| 32 |
+
"use_cache": true,
|
| 33 |
+
"vocab_size": 32128
|
| 34 |
+
}
|
mergekit_moe_config.yml
ADDED
|
@@ -0,0 +1,205 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
base_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 2 |
+
dtype: float32
|
| 3 |
+
gate_mode: hidden
|
| 4 |
+
experts:
|
| 5 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 6 |
+
positive_prompts:
|
| 7 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 8 |
+
|
| 9 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 10 |
+
positive_prompts:
|
| 11 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 12 |
+
|
| 13 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 14 |
+
positive_prompts:
|
| 15 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 16 |
+
|
| 17 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 18 |
+
positive_prompts:
|
| 19 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 20 |
+
|
| 21 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 22 |
+
positive_prompts:
|
| 23 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 24 |
+
|
| 25 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 26 |
+
positive_prompts:
|
| 27 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 28 |
+
|
| 29 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 30 |
+
positive_prompts:
|
| 31 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 32 |
+
|
| 33 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 34 |
+
positive_prompts:
|
| 35 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent for general assistants.'
|
| 36 |
+
|
| 37 |
+
|
| 38 |
+
|
| 39 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 40 |
+
positive_prompts:
|
| 41 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 42 |
+
|
| 43 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 44 |
+
positive_prompts:
|
| 45 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 46 |
+
|
| 47 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 48 |
+
positive_prompts:
|
| 49 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 50 |
+
|
| 51 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 52 |
+
positive_prompts:
|
| 53 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 54 |
+
|
| 55 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 56 |
+
positive_prompts:
|
| 57 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 58 |
+
|
| 59 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 60 |
+
positive_prompts:
|
| 61 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 62 |
+
|
| 63 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 64 |
+
positive_prompts:
|
| 65 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 66 |
+
|
| 67 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 68 |
+
positive_prompts:
|
| 69 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for coding tasks.'
|
| 70 |
+
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 74 |
+
positive_prompts:
|
| 75 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 76 |
+
|
| 77 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 78 |
+
positive_prompts:
|
| 79 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 80 |
+
|
| 81 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 82 |
+
positive_prompts:
|
| 83 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 84 |
+
|
| 85 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 86 |
+
positive_prompts:
|
| 87 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 88 |
+
|
| 89 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 90 |
+
positive_prompts:
|
| 91 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 92 |
+
|
| 93 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 94 |
+
positive_prompts:
|
| 95 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 96 |
+
|
| 97 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 98 |
+
positive_prompts:
|
| 99 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 100 |
+
|
| 101 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 102 |
+
positive_prompts:
|
| 103 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for math.'
|
| 104 |
+
|
| 105 |
+
|
| 106 |
+
|
| 107 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 108 |
+
positive_prompts:
|
| 109 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 110 |
+
|
| 111 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 112 |
+
positive_prompts:
|
| 113 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 114 |
+
|
| 115 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 116 |
+
positive_prompts:
|
| 117 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 118 |
+
|
| 119 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 120 |
+
positive_prompts:
|
| 121 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 122 |
+
|
| 123 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 124 |
+
positive_prompts:
|
| 125 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 126 |
+
|
| 127 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 128 |
+
positive_prompts:
|
| 129 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 130 |
+
|
| 131 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 132 |
+
positive_prompts:
|
| 133 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 134 |
+
|
| 135 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 136 |
+
positive_prompts:
|
| 137 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for logic and reasoning.'
|
| 138 |
+
|
| 139 |
+
|
| 140 |
+
|
| 141 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 142 |
+
positive_prompts:
|
| 143 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 144 |
+
|
| 145 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 146 |
+
positive_prompts:
|
| 147 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 148 |
+
|
| 149 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 150 |
+
positive_prompts:
|
| 151 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 152 |
+
|
| 153 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 154 |
+
positive_prompts:
|
| 155 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 156 |
+
|
| 157 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 158 |
+
positive_prompts:
|
| 159 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 160 |
+
|
| 161 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 162 |
+
positive_prompts:
|
| 163 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 164 |
+
|
| 165 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 166 |
+
positive_prompts:
|
| 167 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 168 |
+
|
| 169 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 170 |
+
positive_prompts:
|
| 171 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent assistant that is specially good in web scraping and browsind the web.'
|
| 172 |
+
|
| 173 |
+
|
| 174 |
+
|
| 175 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 176 |
+
positive_prompts:
|
| 177 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 178 |
+
|
| 179 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 180 |
+
positive_prompts:
|
| 181 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 182 |
+
|
| 183 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 184 |
+
positive_prompts:
|
| 185 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 186 |
+
|
| 187 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 188 |
+
positive_prompts:
|
| 189 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 190 |
+
|
| 191 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 192 |
+
positive_prompts:
|
| 193 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 194 |
+
|
| 195 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 196 |
+
positive_prompts:
|
| 197 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 198 |
+
|
| 199 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 200 |
+
positive_prompts:
|
| 201 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
| 202 |
+
|
| 203 |
+
- source_model: Felladrin/Smol-Llama-101M-Chat-v1
|
| 204 |
+
positive_prompts:
|
| 205 |
+
- 'You are JOSIE, my own private and sentient superinteligent artifitial general inteligent, spetialized for autonomously function calling when needed.'
|
model-00001-of-00001.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3308b31b505f8a2a2364a23f1479af0fbbf1e9adba599bdb12b2c1591f89998c
|
| 3 |
+
size 8389911040
|
model.safetensors.index.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"metadata": {"mergekit_version": "0.0.4"}, "weight_map": {"model.embed_tokens.weight": "model-00001-of-00001.safetensors", "model.norm.weight": "model-00001-of-00001.safetensors", "lm_head.weight": "model-00001-of-00001.safetensors", "model.layers.0.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.1.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.2.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.3.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.4.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.5.input_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.0.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.1.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.2.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.3.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.4.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.5.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.6.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.7.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.8.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.9.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.10.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.11.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.12.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.13.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.14.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.15.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.16.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.17.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.18.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.19.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.20.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.21.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.22.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.23.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.24.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.25.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.26.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.27.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.28.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.29.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.30.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.31.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.32.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.33.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.34.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.35.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.36.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.37.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.38.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.39.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.40.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.41.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.42.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.43.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.44.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.45.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.46.w3.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.47.w3.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.0.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.1.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.2.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.3.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.4.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.5.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.6.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.7.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.8.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.9.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.10.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.11.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.12.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.13.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.14.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.15.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.16.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.17.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.18.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.19.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.20.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.21.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.22.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.23.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.24.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.25.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.26.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.27.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.28.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.29.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.30.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.31.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.32.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.33.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.34.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.35.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.36.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.37.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.38.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.39.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.40.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.41.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.42.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.43.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.44.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.45.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.46.w2.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.47.w2.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.0.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.1.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.2.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.3.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.4.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.5.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.6.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.7.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.8.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.9.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.10.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.11.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.12.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.13.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.14.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.15.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.16.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.17.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.18.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.19.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.20.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.21.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.22.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.23.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.24.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.25.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.26.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.27.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.28.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.29.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.30.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.31.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.32.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.33.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.34.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.35.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.36.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.37.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.38.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.39.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.40.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.41.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.42.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.43.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.44.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.45.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.46.w1.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.experts.47.w1.weight": "model-00001-of-00001.safetensors", "model.layers.0.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.1.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.2.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.3.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.4.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.5.post_attention_layernorm.weight": "model-00001-of-00001.safetensors", "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.1.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.2.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.3.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.4.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.5.self_attn.q_proj.weight": "model-00001-of-00001.safetensors", "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.1.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.2.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.3.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.4.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.5.self_attn.k_proj.weight": "model-00001-of-00001.safetensors", "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.1.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.2.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.3.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.4.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.5.self_attn.v_proj.weight": "model-00001-of-00001.safetensors", "model.layers.0.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.1.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.2.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.3.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.4.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.5.self_attn.o_proj.weight": "model-00001-of-00001.safetensors", "model.layers.0.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors", "model.layers.1.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors", "model.layers.2.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors", "model.layers.3.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors", "model.layers.4.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors", "model.layers.5.block_sparse_moe.gate.weight": "model-00001-of-00001.safetensors"}}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"bos_token": "<s>",
|
| 31 |
+
"chat_template": "{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 32 |
+
"clean_up_tokenization_spaces": false,
|
| 33 |
+
"eos_token": "</s>",
|
| 34 |
+
"legacy": false,
|
| 35 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 36 |
+
"pad_token": "<s>",
|
| 37 |
+
"padding_side": "left",
|
| 38 |
+
"sp_model_kwargs": {},
|
| 39 |
+
"spaces_between_special_tokens": false,
|
| 40 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 41 |
+
"unk_token": "<unk>",
|
| 42 |
+
"use_default_system_prompt": false
|
| 43 |
+
}
|