chinoll commited on
Commit
3be9c40
·
0 Parent(s):

Release MOSS Image Edit LoRA with showcase and usage guides

Browse files
.gitattributes ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ assets/moss-image-edit-showcase.jpg filter=lfs diff=lfs merge=lfs -text
37
+ assets/moss-image-edit-banner-cases.png filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ pipeline_tag: image-to-image
3
+ base_model: OpenMOSS-Team/MOSS-Image-SFT
4
+ language:
5
+ - en
6
+ - zh
7
+ tags:
8
+ - lora
9
+ - image-editing
10
+ - moss-image
11
+ - comfyui
12
+ ---
13
+
14
+ <p align="center">
15
+ <img src="assets/moss-image-edit-banner-cases.png" alt="MOSS Image Edit" width="1280">
16
+ </p>
17
+
18
+ <p align="center">
19
+ <strong>Give your images a new expression.</strong><br>
20
+ MOSS Image Edit is an <strong>image-editing LoRA for MOSS-Image</strong>. Edit colors, outfits, lighting, seasons, and visual style with text instructions.<br>
21
+ This repository provides LoRA weights and configuration. Use them with the MOSS-Image Base or SFT model.
22
+ </p>
23
+
24
+ <p align="center">
25
+ <a href="https://github.com/Chenkin-x/Moss-Image-Edit">GitHub · 中文介绍</a>
26
+ &nbsp; · &nbsp;
27
+ <a href="https://github.com/Chenkin-x/Moss-Image-Edit/blob/main/README.en.md">English</a>
28
+ &nbsp; · &nbsp;
29
+ <a href="https://github.com/Chenkin-x/ComfyUI-Moss-Image-Edit/blob/main/README.en.md">ComfyUI guide</a>
30
+ </p>
31
+
32
+ ## Editing showcase
33
+
34
+ Explore changes in color, outfits, lighting, seasons, and visual style. Each group starts with the original image, followed by its edits.
35
+
36
+ ![MOSS Image Edit showcase](assets/moss-image-edit-showcase.jpg)
37
+
38
+ ## Get started
39
+
40
+ Download this editing LoRA together with either [MOSS-Image-Base](https://huggingface.co/OpenMOSS-Team/MOSS-Image-Base) or [MOSS-Image-SFT](https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT), then open a ComfyUI workflow. **Both base models are supported; choose either one.**
41
+
42
+ | Resource | Link |
43
+ |:---|:---|
44
+ | Installation and model placement | [English guide](https://github.com/Chenkin-x/ComfyUI-Moss-Image-Edit/blob/main/README.en.md) · [中文教程](https://github.com/Chenkin-x/ComfyUI-Moss-Image-Edit) |
45
+ | Edit with one reference | [Single-reference workflow](https://github.com/Chenkin-x/ComfyUI-Moss-Image-Edit/blob/main/workflows/moss_edit_single_reference.json) |
46
+ | Edit with multiple references | [Multi-reference workflow](https://github.com/Chenkin-x/ComfyUI-Moss-Image-Edit/blob/main/workflows/moss_edit_multi_reference.json) |
47
+
48
+ Upload your reference image, describe the change, and generate the result. With multiple references, explain the role of each image in your instruction; up to four images can be connected.
49
+
50
+ To download the LoRA and base-model files from a terminal, using SFT as the example:
51
+
52
+ ```bash
53
+ hf download chinoll/Moss-Image-Edit adapter_config.json adapter_model.safetensors --local-dir models/Moss-Image-Edit
54
+ hf download OpenMOSS-Team/MOSS-Image-SFT --local-dir models/MOSS-Image-SFT
55
+ ```
56
+
57
+ For Base, use this base-model download command instead:
58
+
59
+ ```bash
60
+ hf download OpenMOSS-Team/MOSS-Image-Base --local-dir models/MOSS-Image-Base
61
+ ```
62
+
63
+ Set the ComfyUI loader's `snapshot_dir` to the directory containing your chosen base model.
64
+
65
+ ## Credits
66
+
67
+ - Thanks to [@niangao2331](https://github.com/niangao2331) for contributing part of the image-editing data used in this project.
68
+ - Thanks to the [ConceptEdit team](https://huggingface.co/inclusionAI) for releasing the [ConceptEdit-12M dataset](https://huggingface.co/datasets/inclusionAI/ConceptEdit-12M) for single-image editing.
69
+ - Thanks to [@onglon114514](https://github.com/onglon114514) for generating the showcase images.
70
+
71
+ Built on [MOSS-Image](https://huggingface.co/OpenMOSS-Team/MOSS-Image-SFT). Project code is available under [GPLv3](https://github.com/Chenkin-x/Moss-Image-Edit/blob/main/LICENSE).
adapter_config.json ADDED
@@ -0,0 +1,2294 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "format": "moss-edit-lora",
3
+ "format_version": 1,
4
+ "metadata": {
5
+ "attention_backend": "fa4-varlen-mask_mod",
6
+ "dit_quantization": "none",
7
+ "edit_contract": "ref-prefix-only-t0__target-text-original-moss-v1",
8
+ "model_config": {
9
+ "dim": 3072,
10
+ "ffn_dim": 8192,
11
+ "freq_dim": 256,
12
+ "latent_channels": 128,
13
+ "num_heads": 24,
14
+ "num_layers": 42,
15
+ "rope_base": 10000.0,
16
+ "text_dim": 4096
17
+ },
18
+ "snapshot_dir": "/data/comfyui/models/huggingface/OpenMOSS-Team/MOSS-Image-SFT",
19
+ "text_quantization": "quanto_int8"
20
+ },
21
+ "modules": {
22
+ "layers.0.proj.ffn_down": {
23
+ "alpha": 64.0,
24
+ "dropout": 0.0,
25
+ "in_features": 8192,
26
+ "out_features": 3072,
27
+ "param_dtype": "bfloat16",
28
+ "rank": 64
29
+ },
30
+ "layers.0.proj.ffn_up_gate": {
31
+ "alpha": 64.0,
32
+ "dropout": 0.0,
33
+ "in_features": 3072,
34
+ "out_features": 16384,
35
+ "param_dtype": "bfloat16",
36
+ "rank": 64
37
+ },
38
+ "layers.0.proj.k": {
39
+ "alpha": 64.0,
40
+ "dropout": 0.0,
41
+ "in_features": 3072,
42
+ "out_features": 3072,
43
+ "param_dtype": "bfloat16",
44
+ "rank": 64
45
+ },
46
+ "layers.0.proj.o": {
47
+ "alpha": 64.0,
48
+ "dropout": 0.0,
49
+ "in_features": 3072,
50
+ "out_features": 3072,
51
+ "param_dtype": "bfloat16",
52
+ "rank": 64
53
+ },
54
+ "layers.0.proj.q": {
55
+ "alpha": 64.0,
56
+ "dropout": 0.0,
57
+ "in_features": 3072,
58
+ "out_features": 3072,
59
+ "param_dtype": "bfloat16",
60
+ "rank": 64
61
+ },
62
+ "layers.0.proj.v": {
63
+ "alpha": 64.0,
64
+ "dropout": 0.0,
65
+ "in_features": 3072,
66
+ "out_features": 3072,
67
+ "param_dtype": "bfloat16",
68
+ "rank": 64
69
+ },
70
+ "layers.1.proj.ffn_down": {
71
+ "alpha": 64.0,
72
+ "dropout": 0.0,
73
+ "in_features": 8192,
74
+ "out_features": 3072,
75
+ "param_dtype": "bfloat16",
76
+ "rank": 64
77
+ },
78
+ "layers.1.proj.ffn_up_gate": {
79
+ "alpha": 64.0,
80
+ "dropout": 0.0,
81
+ "in_features": 3072,
82
+ "out_features": 16384,
83
+ "param_dtype": "bfloat16",
84
+ "rank": 64
85
+ },
86
+ "layers.1.proj.k": {
87
+ "alpha": 64.0,
88
+ "dropout": 0.0,
89
+ "in_features": 3072,
90
+ "out_features": 3072,
91
+ "param_dtype": "bfloat16",
92
+ "rank": 64
93
+ },
94
+ "layers.1.proj.o": {
95
+ "alpha": 64.0,
96
+ "dropout": 0.0,
97
+ "in_features": 3072,
98
+ "out_features": 3072,
99
+ "param_dtype": "bfloat16",
100
+ "rank": 64
101
+ },
102
+ "layers.1.proj.q": {
103
+ "alpha": 64.0,
104
+ "dropout": 0.0,
105
+ "in_features": 3072,
106
+ "out_features": 3072,
107
+ "param_dtype": "bfloat16",
108
+ "rank": 64
109
+ },
110
+ "layers.1.proj.v": {
111
+ "alpha": 64.0,
112
+ "dropout": 0.0,
113
+ "in_features": 3072,
114
+ "out_features": 3072,
115
+ "param_dtype": "bfloat16",
116
+ "rank": 64
117
+ },
118
+ "layers.10.proj.ffn_down": {
119
+ "alpha": 64.0,
120
+ "dropout": 0.0,
121
+ "in_features": 8192,
122
+ "out_features": 3072,
123
+ "param_dtype": "bfloat16",
124
+ "rank": 64
125
+ },
126
+ "layers.10.proj.ffn_up_gate": {
127
+ "alpha": 64.0,
128
+ "dropout": 0.0,
129
+ "in_features": 3072,
130
+ "out_features": 16384,
131
+ "param_dtype": "bfloat16",
132
+ "rank": 64
133
+ },
134
+ "layers.10.proj.k": {
135
+ "alpha": 64.0,
136
+ "dropout": 0.0,
137
+ "in_features": 3072,
138
+ "out_features": 3072,
139
+ "param_dtype": "bfloat16",
140
+ "rank": 64
141
+ },
142
+ "layers.10.proj.o": {
143
+ "alpha": 64.0,
144
+ "dropout": 0.0,
145
+ "in_features": 3072,
146
+ "out_features": 3072,
147
+ "param_dtype": "bfloat16",
148
+ "rank": 64
149
+ },
150
+ "layers.10.proj.q": {
151
+ "alpha": 64.0,
152
+ "dropout": 0.0,
153
+ "in_features": 3072,
154
+ "out_features": 3072,
155
+ "param_dtype": "bfloat16",
156
+ "rank": 64
157
+ },
158
+ "layers.10.proj.v": {
159
+ "alpha": 64.0,
160
+ "dropout": 0.0,
161
+ "in_features": 3072,
162
+ "out_features": 3072,
163
+ "param_dtype": "bfloat16",
164
+ "rank": 64
165
+ },
166
+ "layers.11.proj.ffn_down": {
167
+ "alpha": 64.0,
168
+ "dropout": 0.0,
169
+ "in_features": 8192,
170
+ "out_features": 3072,
171
+ "param_dtype": "bfloat16",
172
+ "rank": 64
173
+ },
174
+ "layers.11.proj.ffn_up_gate": {
175
+ "alpha": 64.0,
176
+ "dropout": 0.0,
177
+ "in_features": 3072,
178
+ "out_features": 16384,
179
+ "param_dtype": "bfloat16",
180
+ "rank": 64
181
+ },
182
+ "layers.11.proj.k": {
183
+ "alpha": 64.0,
184
+ "dropout": 0.0,
185
+ "in_features": 3072,
186
+ "out_features": 3072,
187
+ "param_dtype": "bfloat16",
188
+ "rank": 64
189
+ },
190
+ "layers.11.proj.o": {
191
+ "alpha": 64.0,
192
+ "dropout": 0.0,
193
+ "in_features": 3072,
194
+ "out_features": 3072,
195
+ "param_dtype": "bfloat16",
196
+ "rank": 64
197
+ },
198
+ "layers.11.proj.q": {
199
+ "alpha": 64.0,
200
+ "dropout": 0.0,
201
+ "in_features": 3072,
202
+ "out_features": 3072,
203
+ "param_dtype": "bfloat16",
204
+ "rank": 64
205
+ },
206
+ "layers.11.proj.v": {
207
+ "alpha": 64.0,
208
+ "dropout": 0.0,
209
+ "in_features": 3072,
210
+ "out_features": 3072,
211
+ "param_dtype": "bfloat16",
212
+ "rank": 64
213
+ },
214
+ "layers.12.proj.ffn_down": {
215
+ "alpha": 64.0,
216
+ "dropout": 0.0,
217
+ "in_features": 8192,
218
+ "out_features": 3072,
219
+ "param_dtype": "bfloat16",
220
+ "rank": 64
221
+ },
222
+ "layers.12.proj.ffn_up_gate": {
223
+ "alpha": 64.0,
224
+ "dropout": 0.0,
225
+ "in_features": 3072,
226
+ "out_features": 16384,
227
+ "param_dtype": "bfloat16",
228
+ "rank": 64
229
+ },
230
+ "layers.12.proj.k": {
231
+ "alpha": 64.0,
232
+ "dropout": 0.0,
233
+ "in_features": 3072,
234
+ "out_features": 3072,
235
+ "param_dtype": "bfloat16",
236
+ "rank": 64
237
+ },
238
+ "layers.12.proj.o": {
239
+ "alpha": 64.0,
240
+ "dropout": 0.0,
241
+ "in_features": 3072,
242
+ "out_features": 3072,
243
+ "param_dtype": "bfloat16",
244
+ "rank": 64
245
+ },
246
+ "layers.12.proj.q": {
247
+ "alpha": 64.0,
248
+ "dropout": 0.0,
249
+ "in_features": 3072,
250
+ "out_features": 3072,
251
+ "param_dtype": "bfloat16",
252
+ "rank": 64
253
+ },
254
+ "layers.12.proj.v": {
255
+ "alpha": 64.0,
256
+ "dropout": 0.0,
257
+ "in_features": 3072,
258
+ "out_features": 3072,
259
+ "param_dtype": "bfloat16",
260
+ "rank": 64
261
+ },
262
+ "layers.13.proj.ffn_down": {
263
+ "alpha": 64.0,
264
+ "dropout": 0.0,
265
+ "in_features": 8192,
266
+ "out_features": 3072,
267
+ "param_dtype": "bfloat16",
268
+ "rank": 64
269
+ },
270
+ "layers.13.proj.ffn_up_gate": {
271
+ "alpha": 64.0,
272
+ "dropout": 0.0,
273
+ "in_features": 3072,
274
+ "out_features": 16384,
275
+ "param_dtype": "bfloat16",
276
+ "rank": 64
277
+ },
278
+ "layers.13.proj.k": {
279
+ "alpha": 64.0,
280
+ "dropout": 0.0,
281
+ "in_features": 3072,
282
+ "out_features": 3072,
283
+ "param_dtype": "bfloat16",
284
+ "rank": 64
285
+ },
286
+ "layers.13.proj.o": {
287
+ "alpha": 64.0,
288
+ "dropout": 0.0,
289
+ "in_features": 3072,
290
+ "out_features": 3072,
291
+ "param_dtype": "bfloat16",
292
+ "rank": 64
293
+ },
294
+ "layers.13.proj.q": {
295
+ "alpha": 64.0,
296
+ "dropout": 0.0,
297
+ "in_features": 3072,
298
+ "out_features": 3072,
299
+ "param_dtype": "bfloat16",
300
+ "rank": 64
301
+ },
302
+ "layers.13.proj.v": {
303
+ "alpha": 64.0,
304
+ "dropout": 0.0,
305
+ "in_features": 3072,
306
+ "out_features": 3072,
307
+ "param_dtype": "bfloat16",
308
+ "rank": 64
309
+ },
310
+ "layers.14.proj.ffn_down": {
311
+ "alpha": 64.0,
312
+ "dropout": 0.0,
313
+ "in_features": 8192,
314
+ "out_features": 3072,
315
+ "param_dtype": "bfloat16",
316
+ "rank": 64
317
+ },
318
+ "layers.14.proj.ffn_up_gate": {
319
+ "alpha": 64.0,
320
+ "dropout": 0.0,
321
+ "in_features": 3072,
322
+ "out_features": 16384,
323
+ "param_dtype": "bfloat16",
324
+ "rank": 64
325
+ },
326
+ "layers.14.proj.k": {
327
+ "alpha": 64.0,
328
+ "dropout": 0.0,
329
+ "in_features": 3072,
330
+ "out_features": 3072,
331
+ "param_dtype": "bfloat16",
332
+ "rank": 64
333
+ },
334
+ "layers.14.proj.o": {
335
+ "alpha": 64.0,
336
+ "dropout": 0.0,
337
+ "in_features": 3072,
338
+ "out_features": 3072,
339
+ "param_dtype": "bfloat16",
340
+ "rank": 64
341
+ },
342
+ "layers.14.proj.q": {
343
+ "alpha": 64.0,
344
+ "dropout": 0.0,
345
+ "in_features": 3072,
346
+ "out_features": 3072,
347
+ "param_dtype": "bfloat16",
348
+ "rank": 64
349
+ },
350
+ "layers.14.proj.v": {
351
+ "alpha": 64.0,
352
+ "dropout": 0.0,
353
+ "in_features": 3072,
354
+ "out_features": 3072,
355
+ "param_dtype": "bfloat16",
356
+ "rank": 64
357
+ },
358
+ "layers.15.proj.ffn_down": {
359
+ "alpha": 64.0,
360
+ "dropout": 0.0,
361
+ "in_features": 8192,
362
+ "out_features": 3072,
363
+ "param_dtype": "bfloat16",
364
+ "rank": 64
365
+ },
366
+ "layers.15.proj.ffn_up_gate": {
367
+ "alpha": 64.0,
368
+ "dropout": 0.0,
369
+ "in_features": 3072,
370
+ "out_features": 16384,
371
+ "param_dtype": "bfloat16",
372
+ "rank": 64
373
+ },
374
+ "layers.15.proj.k": {
375
+ "alpha": 64.0,
376
+ "dropout": 0.0,
377
+ "in_features": 3072,
378
+ "out_features": 3072,
379
+ "param_dtype": "bfloat16",
380
+ "rank": 64
381
+ },
382
+ "layers.15.proj.o": {
383
+ "alpha": 64.0,
384
+ "dropout": 0.0,
385
+ "in_features": 3072,
386
+ "out_features": 3072,
387
+ "param_dtype": "bfloat16",
388
+ "rank": 64
389
+ },
390
+ "layers.15.proj.q": {
391
+ "alpha": 64.0,
392
+ "dropout": 0.0,
393
+ "in_features": 3072,
394
+ "out_features": 3072,
395
+ "param_dtype": "bfloat16",
396
+ "rank": 64
397
+ },
398
+ "layers.15.proj.v": {
399
+ "alpha": 64.0,
400
+ "dropout": 0.0,
401
+ "in_features": 3072,
402
+ "out_features": 3072,
403
+ "param_dtype": "bfloat16",
404
+ "rank": 64
405
+ },
406
+ "layers.16.proj.ffn_down": {
407
+ "alpha": 64.0,
408
+ "dropout": 0.0,
409
+ "in_features": 8192,
410
+ "out_features": 3072,
411
+ "param_dtype": "bfloat16",
412
+ "rank": 64
413
+ },
414
+ "layers.16.proj.ffn_up_gate": {
415
+ "alpha": 64.0,
416
+ "dropout": 0.0,
417
+ "in_features": 3072,
418
+ "out_features": 16384,
419
+ "param_dtype": "bfloat16",
420
+ "rank": 64
421
+ },
422
+ "layers.16.proj.k": {
423
+ "alpha": 64.0,
424
+ "dropout": 0.0,
425
+ "in_features": 3072,
426
+ "out_features": 3072,
427
+ "param_dtype": "bfloat16",
428
+ "rank": 64
429
+ },
430
+ "layers.16.proj.o": {
431
+ "alpha": 64.0,
432
+ "dropout": 0.0,
433
+ "in_features": 3072,
434
+ "out_features": 3072,
435
+ "param_dtype": "bfloat16",
436
+ "rank": 64
437
+ },
438
+ "layers.16.proj.q": {
439
+ "alpha": 64.0,
440
+ "dropout": 0.0,
441
+ "in_features": 3072,
442
+ "out_features": 3072,
443
+ "param_dtype": "bfloat16",
444
+ "rank": 64
445
+ },
446
+ "layers.16.proj.v": {
447
+ "alpha": 64.0,
448
+ "dropout": 0.0,
449
+ "in_features": 3072,
450
+ "out_features": 3072,
451
+ "param_dtype": "bfloat16",
452
+ "rank": 64
453
+ },
454
+ "layers.17.proj.ffn_down": {
455
+ "alpha": 64.0,
456
+ "dropout": 0.0,
457
+ "in_features": 8192,
458
+ "out_features": 3072,
459
+ "param_dtype": "bfloat16",
460
+ "rank": 64
461
+ },
462
+ "layers.17.proj.ffn_up_gate": {
463
+ "alpha": 64.0,
464
+ "dropout": 0.0,
465
+ "in_features": 3072,
466
+ "out_features": 16384,
467
+ "param_dtype": "bfloat16",
468
+ "rank": 64
469
+ },
470
+ "layers.17.proj.k": {
471
+ "alpha": 64.0,
472
+ "dropout": 0.0,
473
+ "in_features": 3072,
474
+ "out_features": 3072,
475
+ "param_dtype": "bfloat16",
476
+ "rank": 64
477
+ },
478
+ "layers.17.proj.o": {
479
+ "alpha": 64.0,
480
+ "dropout": 0.0,
481
+ "in_features": 3072,
482
+ "out_features": 3072,
483
+ "param_dtype": "bfloat16",
484
+ "rank": 64
485
+ },
486
+ "layers.17.proj.q": {
487
+ "alpha": 64.0,
488
+ "dropout": 0.0,
489
+ "in_features": 3072,
490
+ "out_features": 3072,
491
+ "param_dtype": "bfloat16",
492
+ "rank": 64
493
+ },
494
+ "layers.17.proj.v": {
495
+ "alpha": 64.0,
496
+ "dropout": 0.0,
497
+ "in_features": 3072,
498
+ "out_features": 3072,
499
+ "param_dtype": "bfloat16",
500
+ "rank": 64
501
+ },
502
+ "layers.18.proj.ffn_down": {
503
+ "alpha": 64.0,
504
+ "dropout": 0.0,
505
+ "in_features": 8192,
506
+ "out_features": 3072,
507
+ "param_dtype": "bfloat16",
508
+ "rank": 64
509
+ },
510
+ "layers.18.proj.ffn_up_gate": {
511
+ "alpha": 64.0,
512
+ "dropout": 0.0,
513
+ "in_features": 3072,
514
+ "out_features": 16384,
515
+ "param_dtype": "bfloat16",
516
+ "rank": 64
517
+ },
518
+ "layers.18.proj.k": {
519
+ "alpha": 64.0,
520
+ "dropout": 0.0,
521
+ "in_features": 3072,
522
+ "out_features": 3072,
523
+ "param_dtype": "bfloat16",
524
+ "rank": 64
525
+ },
526
+ "layers.18.proj.o": {
527
+ "alpha": 64.0,
528
+ "dropout": 0.0,
529
+ "in_features": 3072,
530
+ "out_features": 3072,
531
+ "param_dtype": "bfloat16",
532
+ "rank": 64
533
+ },
534
+ "layers.18.proj.q": {
535
+ "alpha": 64.0,
536
+ "dropout": 0.0,
537
+ "in_features": 3072,
538
+ "out_features": 3072,
539
+ "param_dtype": "bfloat16",
540
+ "rank": 64
541
+ },
542
+ "layers.18.proj.v": {
543
+ "alpha": 64.0,
544
+ "dropout": 0.0,
545
+ "in_features": 3072,
546
+ "out_features": 3072,
547
+ "param_dtype": "bfloat16",
548
+ "rank": 64
549
+ },
550
+ "layers.19.proj.ffn_down": {
551
+ "alpha": 64.0,
552
+ "dropout": 0.0,
553
+ "in_features": 8192,
554
+ "out_features": 3072,
555
+ "param_dtype": "bfloat16",
556
+ "rank": 64
557
+ },
558
+ "layers.19.proj.ffn_up_gate": {
559
+ "alpha": 64.0,
560
+ "dropout": 0.0,
561
+ "in_features": 3072,
562
+ "out_features": 16384,
563
+ "param_dtype": "bfloat16",
564
+ "rank": 64
565
+ },
566
+ "layers.19.proj.k": {
567
+ "alpha": 64.0,
568
+ "dropout": 0.0,
569
+ "in_features": 3072,
570
+ "out_features": 3072,
571
+ "param_dtype": "bfloat16",
572
+ "rank": 64
573
+ },
574
+ "layers.19.proj.o": {
575
+ "alpha": 64.0,
576
+ "dropout": 0.0,
577
+ "in_features": 3072,
578
+ "out_features": 3072,
579
+ "param_dtype": "bfloat16",
580
+ "rank": 64
581
+ },
582
+ "layers.19.proj.q": {
583
+ "alpha": 64.0,
584
+ "dropout": 0.0,
585
+ "in_features": 3072,
586
+ "out_features": 3072,
587
+ "param_dtype": "bfloat16",
588
+ "rank": 64
589
+ },
590
+ "layers.19.proj.v": {
591
+ "alpha": 64.0,
592
+ "dropout": 0.0,
593
+ "in_features": 3072,
594
+ "out_features": 3072,
595
+ "param_dtype": "bfloat16",
596
+ "rank": 64
597
+ },
598
+ "layers.2.proj.ffn_down": {
599
+ "alpha": 64.0,
600
+ "dropout": 0.0,
601
+ "in_features": 8192,
602
+ "out_features": 3072,
603
+ "param_dtype": "bfloat16",
604
+ "rank": 64
605
+ },
606
+ "layers.2.proj.ffn_up_gate": {
607
+ "alpha": 64.0,
608
+ "dropout": 0.0,
609
+ "in_features": 3072,
610
+ "out_features": 16384,
611
+ "param_dtype": "bfloat16",
612
+ "rank": 64
613
+ },
614
+ "layers.2.proj.k": {
615
+ "alpha": 64.0,
616
+ "dropout": 0.0,
617
+ "in_features": 3072,
618
+ "out_features": 3072,
619
+ "param_dtype": "bfloat16",
620
+ "rank": 64
621
+ },
622
+ "layers.2.proj.o": {
623
+ "alpha": 64.0,
624
+ "dropout": 0.0,
625
+ "in_features": 3072,
626
+ "out_features": 3072,
627
+ "param_dtype": "bfloat16",
628
+ "rank": 64
629
+ },
630
+ "layers.2.proj.q": {
631
+ "alpha": 64.0,
632
+ "dropout": 0.0,
633
+ "in_features": 3072,
634
+ "out_features": 3072,
635
+ "param_dtype": "bfloat16",
636
+ "rank": 64
637
+ },
638
+ "layers.2.proj.v": {
639
+ "alpha": 64.0,
640
+ "dropout": 0.0,
641
+ "in_features": 3072,
642
+ "out_features": 3072,
643
+ "param_dtype": "bfloat16",
644
+ "rank": 64
645
+ },
646
+ "layers.20.proj.ffn_down": {
647
+ "alpha": 64.0,
648
+ "dropout": 0.0,
649
+ "in_features": 8192,
650
+ "out_features": 3072,
651
+ "param_dtype": "bfloat16",
652
+ "rank": 64
653
+ },
654
+ "layers.20.proj.ffn_up_gate": {
655
+ "alpha": 64.0,
656
+ "dropout": 0.0,
657
+ "in_features": 3072,
658
+ "out_features": 16384,
659
+ "param_dtype": "bfloat16",
660
+ "rank": 64
661
+ },
662
+ "layers.20.proj.k": {
663
+ "alpha": 64.0,
664
+ "dropout": 0.0,
665
+ "in_features": 3072,
666
+ "out_features": 3072,
667
+ "param_dtype": "bfloat16",
668
+ "rank": 64
669
+ },
670
+ "layers.20.proj.o": {
671
+ "alpha": 64.0,
672
+ "dropout": 0.0,
673
+ "in_features": 3072,
674
+ "out_features": 3072,
675
+ "param_dtype": "bfloat16",
676
+ "rank": 64
677
+ },
678
+ "layers.20.proj.q": {
679
+ "alpha": 64.0,
680
+ "dropout": 0.0,
681
+ "in_features": 3072,
682
+ "out_features": 3072,
683
+ "param_dtype": "bfloat16",
684
+ "rank": 64
685
+ },
686
+ "layers.20.proj.v": {
687
+ "alpha": 64.0,
688
+ "dropout": 0.0,
689
+ "in_features": 3072,
690
+ "out_features": 3072,
691
+ "param_dtype": "bfloat16",
692
+ "rank": 64
693
+ },
694
+ "layers.21.proj.ffn_down": {
695
+ "alpha": 64.0,
696
+ "dropout": 0.0,
697
+ "in_features": 8192,
698
+ "out_features": 3072,
699
+ "param_dtype": "bfloat16",
700
+ "rank": 64
701
+ },
702
+ "layers.21.proj.ffn_up_gate": {
703
+ "alpha": 64.0,
704
+ "dropout": 0.0,
705
+ "in_features": 3072,
706
+ "out_features": 16384,
707
+ "param_dtype": "bfloat16",
708
+ "rank": 64
709
+ },
710
+ "layers.21.proj.k": {
711
+ "alpha": 64.0,
712
+ "dropout": 0.0,
713
+ "in_features": 3072,
714
+ "out_features": 3072,
715
+ "param_dtype": "bfloat16",
716
+ "rank": 64
717
+ },
718
+ "layers.21.proj.o": {
719
+ "alpha": 64.0,
720
+ "dropout": 0.0,
721
+ "in_features": 3072,
722
+ "out_features": 3072,
723
+ "param_dtype": "bfloat16",
724
+ "rank": 64
725
+ },
726
+ "layers.21.proj.q": {
727
+ "alpha": 64.0,
728
+ "dropout": 0.0,
729
+ "in_features": 3072,
730
+ "out_features": 3072,
731
+ "param_dtype": "bfloat16",
732
+ "rank": 64
733
+ },
734
+ "layers.21.proj.v": {
735
+ "alpha": 64.0,
736
+ "dropout": 0.0,
737
+ "in_features": 3072,
738
+ "out_features": 3072,
739
+ "param_dtype": "bfloat16",
740
+ "rank": 64
741
+ },
742
+ "layers.22.proj.ffn_down": {
743
+ "alpha": 64.0,
744
+ "dropout": 0.0,
745
+ "in_features": 8192,
746
+ "out_features": 3072,
747
+ "param_dtype": "bfloat16",
748
+ "rank": 64
749
+ },
750
+ "layers.22.proj.ffn_up_gate": {
751
+ "alpha": 64.0,
752
+ "dropout": 0.0,
753
+ "in_features": 3072,
754
+ "out_features": 16384,
755
+ "param_dtype": "bfloat16",
756
+ "rank": 64
757
+ },
758
+ "layers.22.proj.k": {
759
+ "alpha": 64.0,
760
+ "dropout": 0.0,
761
+ "in_features": 3072,
762
+ "out_features": 3072,
763
+ "param_dtype": "bfloat16",
764
+ "rank": 64
765
+ },
766
+ "layers.22.proj.o": {
767
+ "alpha": 64.0,
768
+ "dropout": 0.0,
769
+ "in_features": 3072,
770
+ "out_features": 3072,
771
+ "param_dtype": "bfloat16",
772
+ "rank": 64
773
+ },
774
+ "layers.22.proj.q": {
775
+ "alpha": 64.0,
776
+ "dropout": 0.0,
777
+ "in_features": 3072,
778
+ "out_features": 3072,
779
+ "param_dtype": "bfloat16",
780
+ "rank": 64
781
+ },
782
+ "layers.22.proj.v": {
783
+ "alpha": 64.0,
784
+ "dropout": 0.0,
785
+ "in_features": 3072,
786
+ "out_features": 3072,
787
+ "param_dtype": "bfloat16",
788
+ "rank": 64
789
+ },
790
+ "layers.23.proj.ffn_down": {
791
+ "alpha": 64.0,
792
+ "dropout": 0.0,
793
+ "in_features": 8192,
794
+ "out_features": 3072,
795
+ "param_dtype": "bfloat16",
796
+ "rank": 64
797
+ },
798
+ "layers.23.proj.ffn_up_gate": {
799
+ "alpha": 64.0,
800
+ "dropout": 0.0,
801
+ "in_features": 3072,
802
+ "out_features": 16384,
803
+ "param_dtype": "bfloat16",
804
+ "rank": 64
805
+ },
806
+ "layers.23.proj.k": {
807
+ "alpha": 64.0,
808
+ "dropout": 0.0,
809
+ "in_features": 3072,
810
+ "out_features": 3072,
811
+ "param_dtype": "bfloat16",
812
+ "rank": 64
813
+ },
814
+ "layers.23.proj.o": {
815
+ "alpha": 64.0,
816
+ "dropout": 0.0,
817
+ "in_features": 3072,
818
+ "out_features": 3072,
819
+ "param_dtype": "bfloat16",
820
+ "rank": 64
821
+ },
822
+ "layers.23.proj.q": {
823
+ "alpha": 64.0,
824
+ "dropout": 0.0,
825
+ "in_features": 3072,
826
+ "out_features": 3072,
827
+ "param_dtype": "bfloat16",
828
+ "rank": 64
829
+ },
830
+ "layers.23.proj.v": {
831
+ "alpha": 64.0,
832
+ "dropout": 0.0,
833
+ "in_features": 3072,
834
+ "out_features": 3072,
835
+ "param_dtype": "bfloat16",
836
+ "rank": 64
837
+ },
838
+ "layers.24.proj.ffn_down": {
839
+ "alpha": 64.0,
840
+ "dropout": 0.0,
841
+ "in_features": 8192,
842
+ "out_features": 3072,
843
+ "param_dtype": "bfloat16",
844
+ "rank": 64
845
+ },
846
+ "layers.24.proj.ffn_up_gate": {
847
+ "alpha": 64.0,
848
+ "dropout": 0.0,
849
+ "in_features": 3072,
850
+ "out_features": 16384,
851
+ "param_dtype": "bfloat16",
852
+ "rank": 64
853
+ },
854
+ "layers.24.proj.k": {
855
+ "alpha": 64.0,
856
+ "dropout": 0.0,
857
+ "in_features": 3072,
858
+ "out_features": 3072,
859
+ "param_dtype": "bfloat16",
860
+ "rank": 64
861
+ },
862
+ "layers.24.proj.o": {
863
+ "alpha": 64.0,
864
+ "dropout": 0.0,
865
+ "in_features": 3072,
866
+ "out_features": 3072,
867
+ "param_dtype": "bfloat16",
868
+ "rank": 64
869
+ },
870
+ "layers.24.proj.q": {
871
+ "alpha": 64.0,
872
+ "dropout": 0.0,
873
+ "in_features": 3072,
874
+ "out_features": 3072,
875
+ "param_dtype": "bfloat16",
876
+ "rank": 64
877
+ },
878
+ "layers.24.proj.v": {
879
+ "alpha": 64.0,
880
+ "dropout": 0.0,
881
+ "in_features": 3072,
882
+ "out_features": 3072,
883
+ "param_dtype": "bfloat16",
884
+ "rank": 64
885
+ },
886
+ "layers.25.proj.ffn_down": {
887
+ "alpha": 64.0,
888
+ "dropout": 0.0,
889
+ "in_features": 8192,
890
+ "out_features": 3072,
891
+ "param_dtype": "bfloat16",
892
+ "rank": 64
893
+ },
894
+ "layers.25.proj.ffn_up_gate": {
895
+ "alpha": 64.0,
896
+ "dropout": 0.0,
897
+ "in_features": 3072,
898
+ "out_features": 16384,
899
+ "param_dtype": "bfloat16",
900
+ "rank": 64
901
+ },
902
+ "layers.25.proj.k": {
903
+ "alpha": 64.0,
904
+ "dropout": 0.0,
905
+ "in_features": 3072,
906
+ "out_features": 3072,
907
+ "param_dtype": "bfloat16",
908
+ "rank": 64
909
+ },
910
+ "layers.25.proj.o": {
911
+ "alpha": 64.0,
912
+ "dropout": 0.0,
913
+ "in_features": 3072,
914
+ "out_features": 3072,
915
+ "param_dtype": "bfloat16",
916
+ "rank": 64
917
+ },
918
+ "layers.25.proj.q": {
919
+ "alpha": 64.0,
920
+ "dropout": 0.0,
921
+ "in_features": 3072,
922
+ "out_features": 3072,
923
+ "param_dtype": "bfloat16",
924
+ "rank": 64
925
+ },
926
+ "layers.25.proj.v": {
927
+ "alpha": 64.0,
928
+ "dropout": 0.0,
929
+ "in_features": 3072,
930
+ "out_features": 3072,
931
+ "param_dtype": "bfloat16",
932
+ "rank": 64
933
+ },
934
+ "layers.26.proj.ffn_down": {
935
+ "alpha": 64.0,
936
+ "dropout": 0.0,
937
+ "in_features": 8192,
938
+ "out_features": 3072,
939
+ "param_dtype": "bfloat16",
940
+ "rank": 64
941
+ },
942
+ "layers.26.proj.ffn_up_gate": {
943
+ "alpha": 64.0,
944
+ "dropout": 0.0,
945
+ "in_features": 3072,
946
+ "out_features": 16384,
947
+ "param_dtype": "bfloat16",
948
+ "rank": 64
949
+ },
950
+ "layers.26.proj.k": {
951
+ "alpha": 64.0,
952
+ "dropout": 0.0,
953
+ "in_features": 3072,
954
+ "out_features": 3072,
955
+ "param_dtype": "bfloat16",
956
+ "rank": 64
957
+ },
958
+ "layers.26.proj.o": {
959
+ "alpha": 64.0,
960
+ "dropout": 0.0,
961
+ "in_features": 3072,
962
+ "out_features": 3072,
963
+ "param_dtype": "bfloat16",
964
+ "rank": 64
965
+ },
966
+ "layers.26.proj.q": {
967
+ "alpha": 64.0,
968
+ "dropout": 0.0,
969
+ "in_features": 3072,
970
+ "out_features": 3072,
971
+ "param_dtype": "bfloat16",
972
+ "rank": 64
973
+ },
974
+ "layers.26.proj.v": {
975
+ "alpha": 64.0,
976
+ "dropout": 0.0,
977
+ "in_features": 3072,
978
+ "out_features": 3072,
979
+ "param_dtype": "bfloat16",
980
+ "rank": 64
981
+ },
982
+ "layers.27.proj.ffn_down": {
983
+ "alpha": 64.0,
984
+ "dropout": 0.0,
985
+ "in_features": 8192,
986
+ "out_features": 3072,
987
+ "param_dtype": "bfloat16",
988
+ "rank": 64
989
+ },
990
+ "layers.27.proj.ffn_up_gate": {
991
+ "alpha": 64.0,
992
+ "dropout": 0.0,
993
+ "in_features": 3072,
994
+ "out_features": 16384,
995
+ "param_dtype": "bfloat16",
996
+ "rank": 64
997
+ },
998
+ "layers.27.proj.k": {
999
+ "alpha": 64.0,
1000
+ "dropout": 0.0,
1001
+ "in_features": 3072,
1002
+ "out_features": 3072,
1003
+ "param_dtype": "bfloat16",
1004
+ "rank": 64
1005
+ },
1006
+ "layers.27.proj.o": {
1007
+ "alpha": 64.0,
1008
+ "dropout": 0.0,
1009
+ "in_features": 3072,
1010
+ "out_features": 3072,
1011
+ "param_dtype": "bfloat16",
1012
+ "rank": 64
1013
+ },
1014
+ "layers.27.proj.q": {
1015
+ "alpha": 64.0,
1016
+ "dropout": 0.0,
1017
+ "in_features": 3072,
1018
+ "out_features": 3072,
1019
+ "param_dtype": "bfloat16",
1020
+ "rank": 64
1021
+ },
1022
+ "layers.27.proj.v": {
1023
+ "alpha": 64.0,
1024
+ "dropout": 0.0,
1025
+ "in_features": 3072,
1026
+ "out_features": 3072,
1027
+ "param_dtype": "bfloat16",
1028
+ "rank": 64
1029
+ },
1030
+ "layers.28.proj.ffn_down": {
1031
+ "alpha": 64.0,
1032
+ "dropout": 0.0,
1033
+ "in_features": 8192,
1034
+ "out_features": 3072,
1035
+ "param_dtype": "bfloat16",
1036
+ "rank": 64
1037
+ },
1038
+ "layers.28.proj.ffn_up_gate": {
1039
+ "alpha": 64.0,
1040
+ "dropout": 0.0,
1041
+ "in_features": 3072,
1042
+ "out_features": 16384,
1043
+ "param_dtype": "bfloat16",
1044
+ "rank": 64
1045
+ },
1046
+ "layers.28.proj.k": {
1047
+ "alpha": 64.0,
1048
+ "dropout": 0.0,
1049
+ "in_features": 3072,
1050
+ "out_features": 3072,
1051
+ "param_dtype": "bfloat16",
1052
+ "rank": 64
1053
+ },
1054
+ "layers.28.proj.o": {
1055
+ "alpha": 64.0,
1056
+ "dropout": 0.0,
1057
+ "in_features": 3072,
1058
+ "out_features": 3072,
1059
+ "param_dtype": "bfloat16",
1060
+ "rank": 64
1061
+ },
1062
+ "layers.28.proj.q": {
1063
+ "alpha": 64.0,
1064
+ "dropout": 0.0,
1065
+ "in_features": 3072,
1066
+ "out_features": 3072,
1067
+ "param_dtype": "bfloat16",
1068
+ "rank": 64
1069
+ },
1070
+ "layers.28.proj.v": {
1071
+ "alpha": 64.0,
1072
+ "dropout": 0.0,
1073
+ "in_features": 3072,
1074
+ "out_features": 3072,
1075
+ "param_dtype": "bfloat16",
1076
+ "rank": 64
1077
+ },
1078
+ "layers.29.proj.ffn_down": {
1079
+ "alpha": 64.0,
1080
+ "dropout": 0.0,
1081
+ "in_features": 8192,
1082
+ "out_features": 3072,
1083
+ "param_dtype": "bfloat16",
1084
+ "rank": 64
1085
+ },
1086
+ "layers.29.proj.ffn_up_gate": {
1087
+ "alpha": 64.0,
1088
+ "dropout": 0.0,
1089
+ "in_features": 3072,
1090
+ "out_features": 16384,
1091
+ "param_dtype": "bfloat16",
1092
+ "rank": 64
1093
+ },
1094
+ "layers.29.proj.k": {
1095
+ "alpha": 64.0,
1096
+ "dropout": 0.0,
1097
+ "in_features": 3072,
1098
+ "out_features": 3072,
1099
+ "param_dtype": "bfloat16",
1100
+ "rank": 64
1101
+ },
1102
+ "layers.29.proj.o": {
1103
+ "alpha": 64.0,
1104
+ "dropout": 0.0,
1105
+ "in_features": 3072,
1106
+ "out_features": 3072,
1107
+ "param_dtype": "bfloat16",
1108
+ "rank": 64
1109
+ },
1110
+ "layers.29.proj.q": {
1111
+ "alpha": 64.0,
1112
+ "dropout": 0.0,
1113
+ "in_features": 3072,
1114
+ "out_features": 3072,
1115
+ "param_dtype": "bfloat16",
1116
+ "rank": 64
1117
+ },
1118
+ "layers.29.proj.v": {
1119
+ "alpha": 64.0,
1120
+ "dropout": 0.0,
1121
+ "in_features": 3072,
1122
+ "out_features": 3072,
1123
+ "param_dtype": "bfloat16",
1124
+ "rank": 64
1125
+ },
1126
+ "layers.3.proj.ffn_down": {
1127
+ "alpha": 64.0,
1128
+ "dropout": 0.0,
1129
+ "in_features": 8192,
1130
+ "out_features": 3072,
1131
+ "param_dtype": "bfloat16",
1132
+ "rank": 64
1133
+ },
1134
+ "layers.3.proj.ffn_up_gate": {
1135
+ "alpha": 64.0,
1136
+ "dropout": 0.0,
1137
+ "in_features": 3072,
1138
+ "out_features": 16384,
1139
+ "param_dtype": "bfloat16",
1140
+ "rank": 64
1141
+ },
1142
+ "layers.3.proj.k": {
1143
+ "alpha": 64.0,
1144
+ "dropout": 0.0,
1145
+ "in_features": 3072,
1146
+ "out_features": 3072,
1147
+ "param_dtype": "bfloat16",
1148
+ "rank": 64
1149
+ },
1150
+ "layers.3.proj.o": {
1151
+ "alpha": 64.0,
1152
+ "dropout": 0.0,
1153
+ "in_features": 3072,
1154
+ "out_features": 3072,
1155
+ "param_dtype": "bfloat16",
1156
+ "rank": 64
1157
+ },
1158
+ "layers.3.proj.q": {
1159
+ "alpha": 64.0,
1160
+ "dropout": 0.0,
1161
+ "in_features": 3072,
1162
+ "out_features": 3072,
1163
+ "param_dtype": "bfloat16",
1164
+ "rank": 64
1165
+ },
1166
+ "layers.3.proj.v": {
1167
+ "alpha": 64.0,
1168
+ "dropout": 0.0,
1169
+ "in_features": 3072,
1170
+ "out_features": 3072,
1171
+ "param_dtype": "bfloat16",
1172
+ "rank": 64
1173
+ },
1174
+ "layers.30.proj.ffn_down": {
1175
+ "alpha": 64.0,
1176
+ "dropout": 0.0,
1177
+ "in_features": 8192,
1178
+ "out_features": 3072,
1179
+ "param_dtype": "bfloat16",
1180
+ "rank": 64
1181
+ },
1182
+ "layers.30.proj.ffn_up_gate": {
1183
+ "alpha": 64.0,
1184
+ "dropout": 0.0,
1185
+ "in_features": 3072,
1186
+ "out_features": 16384,
1187
+ "param_dtype": "bfloat16",
1188
+ "rank": 64
1189
+ },
1190
+ "layers.30.proj.k": {
1191
+ "alpha": 64.0,
1192
+ "dropout": 0.0,
1193
+ "in_features": 3072,
1194
+ "out_features": 3072,
1195
+ "param_dtype": "bfloat16",
1196
+ "rank": 64
1197
+ },
1198
+ "layers.30.proj.o": {
1199
+ "alpha": 64.0,
1200
+ "dropout": 0.0,
1201
+ "in_features": 3072,
1202
+ "out_features": 3072,
1203
+ "param_dtype": "bfloat16",
1204
+ "rank": 64
1205
+ },
1206
+ "layers.30.proj.q": {
1207
+ "alpha": 64.0,
1208
+ "dropout": 0.0,
1209
+ "in_features": 3072,
1210
+ "out_features": 3072,
1211
+ "param_dtype": "bfloat16",
1212
+ "rank": 64
1213
+ },
1214
+ "layers.30.proj.v": {
1215
+ "alpha": 64.0,
1216
+ "dropout": 0.0,
1217
+ "in_features": 3072,
1218
+ "out_features": 3072,
1219
+ "param_dtype": "bfloat16",
1220
+ "rank": 64
1221
+ },
1222
+ "layers.31.proj.ffn_down": {
1223
+ "alpha": 64.0,
1224
+ "dropout": 0.0,
1225
+ "in_features": 8192,
1226
+ "out_features": 3072,
1227
+ "param_dtype": "bfloat16",
1228
+ "rank": 64
1229
+ },
1230
+ "layers.31.proj.ffn_up_gate": {
1231
+ "alpha": 64.0,
1232
+ "dropout": 0.0,
1233
+ "in_features": 3072,
1234
+ "out_features": 16384,
1235
+ "param_dtype": "bfloat16",
1236
+ "rank": 64
1237
+ },
1238
+ "layers.31.proj.k": {
1239
+ "alpha": 64.0,
1240
+ "dropout": 0.0,
1241
+ "in_features": 3072,
1242
+ "out_features": 3072,
1243
+ "param_dtype": "bfloat16",
1244
+ "rank": 64
1245
+ },
1246
+ "layers.31.proj.o": {
1247
+ "alpha": 64.0,
1248
+ "dropout": 0.0,
1249
+ "in_features": 3072,
1250
+ "out_features": 3072,
1251
+ "param_dtype": "bfloat16",
1252
+ "rank": 64
1253
+ },
1254
+ "layers.31.proj.q": {
1255
+ "alpha": 64.0,
1256
+ "dropout": 0.0,
1257
+ "in_features": 3072,
1258
+ "out_features": 3072,
1259
+ "param_dtype": "bfloat16",
1260
+ "rank": 64
1261
+ },
1262
+ "layers.31.proj.v": {
1263
+ "alpha": 64.0,
1264
+ "dropout": 0.0,
1265
+ "in_features": 3072,
1266
+ "out_features": 3072,
1267
+ "param_dtype": "bfloat16",
1268
+ "rank": 64
1269
+ },
1270
+ "layers.32.proj.ffn_down": {
1271
+ "alpha": 64.0,
1272
+ "dropout": 0.0,
1273
+ "in_features": 8192,
1274
+ "out_features": 3072,
1275
+ "param_dtype": "bfloat16",
1276
+ "rank": 64
1277
+ },
1278
+ "layers.32.proj.ffn_up_gate": {
1279
+ "alpha": 64.0,
1280
+ "dropout": 0.0,
1281
+ "in_features": 3072,
1282
+ "out_features": 16384,
1283
+ "param_dtype": "bfloat16",
1284
+ "rank": 64
1285
+ },
1286
+ "layers.32.proj.k": {
1287
+ "alpha": 64.0,
1288
+ "dropout": 0.0,
1289
+ "in_features": 3072,
1290
+ "out_features": 3072,
1291
+ "param_dtype": "bfloat16",
1292
+ "rank": 64
1293
+ },
1294
+ "layers.32.proj.o": {
1295
+ "alpha": 64.0,
1296
+ "dropout": 0.0,
1297
+ "in_features": 3072,
1298
+ "out_features": 3072,
1299
+ "param_dtype": "bfloat16",
1300
+ "rank": 64
1301
+ },
1302
+ "layers.32.proj.q": {
1303
+ "alpha": 64.0,
1304
+ "dropout": 0.0,
1305
+ "in_features": 3072,
1306
+ "out_features": 3072,
1307
+ "param_dtype": "bfloat16",
1308
+ "rank": 64
1309
+ },
1310
+ "layers.32.proj.v": {
1311
+ "alpha": 64.0,
1312
+ "dropout": 0.0,
1313
+ "in_features": 3072,
1314
+ "out_features": 3072,
1315
+ "param_dtype": "bfloat16",
1316
+ "rank": 64
1317
+ },
1318
+ "layers.33.proj.ffn_down": {
1319
+ "alpha": 64.0,
1320
+ "dropout": 0.0,
1321
+ "in_features": 8192,
1322
+ "out_features": 3072,
1323
+ "param_dtype": "bfloat16",
1324
+ "rank": 64
1325
+ },
1326
+ "layers.33.proj.ffn_up_gate": {
1327
+ "alpha": 64.0,
1328
+ "dropout": 0.0,
1329
+ "in_features": 3072,
1330
+ "out_features": 16384,
1331
+ "param_dtype": "bfloat16",
1332
+ "rank": 64
1333
+ },
1334
+ "layers.33.proj.k": {
1335
+ "alpha": 64.0,
1336
+ "dropout": 0.0,
1337
+ "in_features": 3072,
1338
+ "out_features": 3072,
1339
+ "param_dtype": "bfloat16",
1340
+ "rank": 64
1341
+ },
1342
+ "layers.33.proj.o": {
1343
+ "alpha": 64.0,
1344
+ "dropout": 0.0,
1345
+ "in_features": 3072,
1346
+ "out_features": 3072,
1347
+ "param_dtype": "bfloat16",
1348
+ "rank": 64
1349
+ },
1350
+ "layers.33.proj.q": {
1351
+ "alpha": 64.0,
1352
+ "dropout": 0.0,
1353
+ "in_features": 3072,
1354
+ "out_features": 3072,
1355
+ "param_dtype": "bfloat16",
1356
+ "rank": 64
1357
+ },
1358
+ "layers.33.proj.v": {
1359
+ "alpha": 64.0,
1360
+ "dropout": 0.0,
1361
+ "in_features": 3072,
1362
+ "out_features": 3072,
1363
+ "param_dtype": "bfloat16",
1364
+ "rank": 64
1365
+ },
1366
+ "layers.34.proj.ffn_down": {
1367
+ "alpha": 64.0,
1368
+ "dropout": 0.0,
1369
+ "in_features": 8192,
1370
+ "out_features": 3072,
1371
+ "param_dtype": "bfloat16",
1372
+ "rank": 64
1373
+ },
1374
+ "layers.34.proj.ffn_up_gate": {
1375
+ "alpha": 64.0,
1376
+ "dropout": 0.0,
1377
+ "in_features": 3072,
1378
+ "out_features": 16384,
1379
+ "param_dtype": "bfloat16",
1380
+ "rank": 64
1381
+ },
1382
+ "layers.34.proj.k": {
1383
+ "alpha": 64.0,
1384
+ "dropout": 0.0,
1385
+ "in_features": 3072,
1386
+ "out_features": 3072,
1387
+ "param_dtype": "bfloat16",
1388
+ "rank": 64
1389
+ },
1390
+ "layers.34.proj.o": {
1391
+ "alpha": 64.0,
1392
+ "dropout": 0.0,
1393
+ "in_features": 3072,
1394
+ "out_features": 3072,
1395
+ "param_dtype": "bfloat16",
1396
+ "rank": 64
1397
+ },
1398
+ "layers.34.proj.q": {
1399
+ "alpha": 64.0,
1400
+ "dropout": 0.0,
1401
+ "in_features": 3072,
1402
+ "out_features": 3072,
1403
+ "param_dtype": "bfloat16",
1404
+ "rank": 64
1405
+ },
1406
+ "layers.34.proj.v": {
1407
+ "alpha": 64.0,
1408
+ "dropout": 0.0,
1409
+ "in_features": 3072,
1410
+ "out_features": 3072,
1411
+ "param_dtype": "bfloat16",
1412
+ "rank": 64
1413
+ },
1414
+ "layers.35.proj.ffn_down": {
1415
+ "alpha": 64.0,
1416
+ "dropout": 0.0,
1417
+ "in_features": 8192,
1418
+ "out_features": 3072,
1419
+ "param_dtype": "bfloat16",
1420
+ "rank": 64
1421
+ },
1422
+ "layers.35.proj.ffn_up_gate": {
1423
+ "alpha": 64.0,
1424
+ "dropout": 0.0,
1425
+ "in_features": 3072,
1426
+ "out_features": 16384,
1427
+ "param_dtype": "bfloat16",
1428
+ "rank": 64
1429
+ },
1430
+ "layers.35.proj.k": {
1431
+ "alpha": 64.0,
1432
+ "dropout": 0.0,
1433
+ "in_features": 3072,
1434
+ "out_features": 3072,
1435
+ "param_dtype": "bfloat16",
1436
+ "rank": 64
1437
+ },
1438
+ "layers.35.proj.o": {
1439
+ "alpha": 64.0,
1440
+ "dropout": 0.0,
1441
+ "in_features": 3072,
1442
+ "out_features": 3072,
1443
+ "param_dtype": "bfloat16",
1444
+ "rank": 64
1445
+ },
1446
+ "layers.35.proj.q": {
1447
+ "alpha": 64.0,
1448
+ "dropout": 0.0,
1449
+ "in_features": 3072,
1450
+ "out_features": 3072,
1451
+ "param_dtype": "bfloat16",
1452
+ "rank": 64
1453
+ },
1454
+ "layers.35.proj.v": {
1455
+ "alpha": 64.0,
1456
+ "dropout": 0.0,
1457
+ "in_features": 3072,
1458
+ "out_features": 3072,
1459
+ "param_dtype": "bfloat16",
1460
+ "rank": 64
1461
+ },
1462
+ "layers.36.proj.ffn_down": {
1463
+ "alpha": 64.0,
1464
+ "dropout": 0.0,
1465
+ "in_features": 8192,
1466
+ "out_features": 3072,
1467
+ "param_dtype": "bfloat16",
1468
+ "rank": 64
1469
+ },
1470
+ "layers.36.proj.ffn_up_gate": {
1471
+ "alpha": 64.0,
1472
+ "dropout": 0.0,
1473
+ "in_features": 3072,
1474
+ "out_features": 16384,
1475
+ "param_dtype": "bfloat16",
1476
+ "rank": 64
1477
+ },
1478
+ "layers.36.proj.k": {
1479
+ "alpha": 64.0,
1480
+ "dropout": 0.0,
1481
+ "in_features": 3072,
1482
+ "out_features": 3072,
1483
+ "param_dtype": "bfloat16",
1484
+ "rank": 64
1485
+ },
1486
+ "layers.36.proj.o": {
1487
+ "alpha": 64.0,
1488
+ "dropout": 0.0,
1489
+ "in_features": 3072,
1490
+ "out_features": 3072,
1491
+ "param_dtype": "bfloat16",
1492
+ "rank": 64
1493
+ },
1494
+ "layers.36.proj.q": {
1495
+ "alpha": 64.0,
1496
+ "dropout": 0.0,
1497
+ "in_features": 3072,
1498
+ "out_features": 3072,
1499
+ "param_dtype": "bfloat16",
1500
+ "rank": 64
1501
+ },
1502
+ "layers.36.proj.v": {
1503
+ "alpha": 64.0,
1504
+ "dropout": 0.0,
1505
+ "in_features": 3072,
1506
+ "out_features": 3072,
1507
+ "param_dtype": "bfloat16",
1508
+ "rank": 64
1509
+ },
1510
+ "layers.37.proj.ffn_down": {
1511
+ "alpha": 64.0,
1512
+ "dropout": 0.0,
1513
+ "in_features": 8192,
1514
+ "out_features": 3072,
1515
+ "param_dtype": "bfloat16",
1516
+ "rank": 64
1517
+ },
1518
+ "layers.37.proj.ffn_up_gate": {
1519
+ "alpha": 64.0,
1520
+ "dropout": 0.0,
1521
+ "in_features": 3072,
1522
+ "out_features": 16384,
1523
+ "param_dtype": "bfloat16",
1524
+ "rank": 64
1525
+ },
1526
+ "layers.37.proj.k": {
1527
+ "alpha": 64.0,
1528
+ "dropout": 0.0,
1529
+ "in_features": 3072,
1530
+ "out_features": 3072,
1531
+ "param_dtype": "bfloat16",
1532
+ "rank": 64
1533
+ },
1534
+ "layers.37.proj.o": {
1535
+ "alpha": 64.0,
1536
+ "dropout": 0.0,
1537
+ "in_features": 3072,
1538
+ "out_features": 3072,
1539
+ "param_dtype": "bfloat16",
1540
+ "rank": 64
1541
+ },
1542
+ "layers.37.proj.q": {
1543
+ "alpha": 64.0,
1544
+ "dropout": 0.0,
1545
+ "in_features": 3072,
1546
+ "out_features": 3072,
1547
+ "param_dtype": "bfloat16",
1548
+ "rank": 64
1549
+ },
1550
+ "layers.37.proj.v": {
1551
+ "alpha": 64.0,
1552
+ "dropout": 0.0,
1553
+ "in_features": 3072,
1554
+ "out_features": 3072,
1555
+ "param_dtype": "bfloat16",
1556
+ "rank": 64
1557
+ },
1558
+ "layers.38.proj.ffn_down": {
1559
+ "alpha": 64.0,
1560
+ "dropout": 0.0,
1561
+ "in_features": 8192,
1562
+ "out_features": 3072,
1563
+ "param_dtype": "bfloat16",
1564
+ "rank": 64
1565
+ },
1566
+ "layers.38.proj.ffn_up_gate": {
1567
+ "alpha": 64.0,
1568
+ "dropout": 0.0,
1569
+ "in_features": 3072,
1570
+ "out_features": 16384,
1571
+ "param_dtype": "bfloat16",
1572
+ "rank": 64
1573
+ },
1574
+ "layers.38.proj.k": {
1575
+ "alpha": 64.0,
1576
+ "dropout": 0.0,
1577
+ "in_features": 3072,
1578
+ "out_features": 3072,
1579
+ "param_dtype": "bfloat16",
1580
+ "rank": 64
1581
+ },
1582
+ "layers.38.proj.o": {
1583
+ "alpha": 64.0,
1584
+ "dropout": 0.0,
1585
+ "in_features": 3072,
1586
+ "out_features": 3072,
1587
+ "param_dtype": "bfloat16",
1588
+ "rank": 64
1589
+ },
1590
+ "layers.38.proj.q": {
1591
+ "alpha": 64.0,
1592
+ "dropout": 0.0,
1593
+ "in_features": 3072,
1594
+ "out_features": 3072,
1595
+ "param_dtype": "bfloat16",
1596
+ "rank": 64
1597
+ },
1598
+ "layers.38.proj.v": {
1599
+ "alpha": 64.0,
1600
+ "dropout": 0.0,
1601
+ "in_features": 3072,
1602
+ "out_features": 3072,
1603
+ "param_dtype": "bfloat16",
1604
+ "rank": 64
1605
+ },
1606
+ "layers.39.proj.ffn_down": {
1607
+ "alpha": 64.0,
1608
+ "dropout": 0.0,
1609
+ "in_features": 8192,
1610
+ "out_features": 3072,
1611
+ "param_dtype": "bfloat16",
1612
+ "rank": 64
1613
+ },
1614
+ "layers.39.proj.ffn_up_gate": {
1615
+ "alpha": 64.0,
1616
+ "dropout": 0.0,
1617
+ "in_features": 3072,
1618
+ "out_features": 16384,
1619
+ "param_dtype": "bfloat16",
1620
+ "rank": 64
1621
+ },
1622
+ "layers.39.proj.k": {
1623
+ "alpha": 64.0,
1624
+ "dropout": 0.0,
1625
+ "in_features": 3072,
1626
+ "out_features": 3072,
1627
+ "param_dtype": "bfloat16",
1628
+ "rank": 64
1629
+ },
1630
+ "layers.39.proj.o": {
1631
+ "alpha": 64.0,
1632
+ "dropout": 0.0,
1633
+ "in_features": 3072,
1634
+ "out_features": 3072,
1635
+ "param_dtype": "bfloat16",
1636
+ "rank": 64
1637
+ },
1638
+ "layers.39.proj.q": {
1639
+ "alpha": 64.0,
1640
+ "dropout": 0.0,
1641
+ "in_features": 3072,
1642
+ "out_features": 3072,
1643
+ "param_dtype": "bfloat16",
1644
+ "rank": 64
1645
+ },
1646
+ "layers.39.proj.v": {
1647
+ "alpha": 64.0,
1648
+ "dropout": 0.0,
1649
+ "in_features": 3072,
1650
+ "out_features": 3072,
1651
+ "param_dtype": "bfloat16",
1652
+ "rank": 64
1653
+ },
1654
+ "layers.4.proj.ffn_down": {
1655
+ "alpha": 64.0,
1656
+ "dropout": 0.0,
1657
+ "in_features": 8192,
1658
+ "out_features": 3072,
1659
+ "param_dtype": "bfloat16",
1660
+ "rank": 64
1661
+ },
1662
+ "layers.4.proj.ffn_up_gate": {
1663
+ "alpha": 64.0,
1664
+ "dropout": 0.0,
1665
+ "in_features": 3072,
1666
+ "out_features": 16384,
1667
+ "param_dtype": "bfloat16",
1668
+ "rank": 64
1669
+ },
1670
+ "layers.4.proj.k": {
1671
+ "alpha": 64.0,
1672
+ "dropout": 0.0,
1673
+ "in_features": 3072,
1674
+ "out_features": 3072,
1675
+ "param_dtype": "bfloat16",
1676
+ "rank": 64
1677
+ },
1678
+ "layers.4.proj.o": {
1679
+ "alpha": 64.0,
1680
+ "dropout": 0.0,
1681
+ "in_features": 3072,
1682
+ "out_features": 3072,
1683
+ "param_dtype": "bfloat16",
1684
+ "rank": 64
1685
+ },
1686
+ "layers.4.proj.q": {
1687
+ "alpha": 64.0,
1688
+ "dropout": 0.0,
1689
+ "in_features": 3072,
1690
+ "out_features": 3072,
1691
+ "param_dtype": "bfloat16",
1692
+ "rank": 64
1693
+ },
1694
+ "layers.4.proj.v": {
1695
+ "alpha": 64.0,
1696
+ "dropout": 0.0,
1697
+ "in_features": 3072,
1698
+ "out_features": 3072,
1699
+ "param_dtype": "bfloat16",
1700
+ "rank": 64
1701
+ },
1702
+ "layers.40.proj.ffn_down": {
1703
+ "alpha": 64.0,
1704
+ "dropout": 0.0,
1705
+ "in_features": 8192,
1706
+ "out_features": 3072,
1707
+ "param_dtype": "bfloat16",
1708
+ "rank": 64
1709
+ },
1710
+ "layers.40.proj.ffn_up_gate": {
1711
+ "alpha": 64.0,
1712
+ "dropout": 0.0,
1713
+ "in_features": 3072,
1714
+ "out_features": 16384,
1715
+ "param_dtype": "bfloat16",
1716
+ "rank": 64
1717
+ },
1718
+ "layers.40.proj.k": {
1719
+ "alpha": 64.0,
1720
+ "dropout": 0.0,
1721
+ "in_features": 3072,
1722
+ "out_features": 3072,
1723
+ "param_dtype": "bfloat16",
1724
+ "rank": 64
1725
+ },
1726
+ "layers.40.proj.o": {
1727
+ "alpha": 64.0,
1728
+ "dropout": 0.0,
1729
+ "in_features": 3072,
1730
+ "out_features": 3072,
1731
+ "param_dtype": "bfloat16",
1732
+ "rank": 64
1733
+ },
1734
+ "layers.40.proj.q": {
1735
+ "alpha": 64.0,
1736
+ "dropout": 0.0,
1737
+ "in_features": 3072,
1738
+ "out_features": 3072,
1739
+ "param_dtype": "bfloat16",
1740
+ "rank": 64
1741
+ },
1742
+ "layers.40.proj.v": {
1743
+ "alpha": 64.0,
1744
+ "dropout": 0.0,
1745
+ "in_features": 3072,
1746
+ "out_features": 3072,
1747
+ "param_dtype": "bfloat16",
1748
+ "rank": 64
1749
+ },
1750
+ "layers.41.proj.ffn_down": {
1751
+ "alpha": 64.0,
1752
+ "dropout": 0.0,
1753
+ "in_features": 8192,
1754
+ "out_features": 3072,
1755
+ "param_dtype": "bfloat16",
1756
+ "rank": 64
1757
+ },
1758
+ "layers.41.proj.ffn_up_gate": {
1759
+ "alpha": 64.0,
1760
+ "dropout": 0.0,
1761
+ "in_features": 3072,
1762
+ "out_features": 16384,
1763
+ "param_dtype": "bfloat16",
1764
+ "rank": 64
1765
+ },
1766
+ "layers.41.proj.k": {
1767
+ "alpha": 64.0,
1768
+ "dropout": 0.0,
1769
+ "in_features": 3072,
1770
+ "out_features": 3072,
1771
+ "param_dtype": "bfloat16",
1772
+ "rank": 64
1773
+ },
1774
+ "layers.41.proj.o": {
1775
+ "alpha": 64.0,
1776
+ "dropout": 0.0,
1777
+ "in_features": 3072,
1778
+ "out_features": 3072,
1779
+ "param_dtype": "bfloat16",
1780
+ "rank": 64
1781
+ },
1782
+ "layers.41.proj.q": {
1783
+ "alpha": 64.0,
1784
+ "dropout": 0.0,
1785
+ "in_features": 3072,
1786
+ "out_features": 3072,
1787
+ "param_dtype": "bfloat16",
1788
+ "rank": 64
1789
+ },
1790
+ "layers.41.proj.v": {
1791
+ "alpha": 64.0,
1792
+ "dropout": 0.0,
1793
+ "in_features": 3072,
1794
+ "out_features": 3072,
1795
+ "param_dtype": "bfloat16",
1796
+ "rank": 64
1797
+ },
1798
+ "layers.5.proj.ffn_down": {
1799
+ "alpha": 64.0,
1800
+ "dropout": 0.0,
1801
+ "in_features": 8192,
1802
+ "out_features": 3072,
1803
+ "param_dtype": "bfloat16",
1804
+ "rank": 64
1805
+ },
1806
+ "layers.5.proj.ffn_up_gate": {
1807
+ "alpha": 64.0,
1808
+ "dropout": 0.0,
1809
+ "in_features": 3072,
1810
+ "out_features": 16384,
1811
+ "param_dtype": "bfloat16",
1812
+ "rank": 64
1813
+ },
1814
+ "layers.5.proj.k": {
1815
+ "alpha": 64.0,
1816
+ "dropout": 0.0,
1817
+ "in_features": 3072,
1818
+ "out_features": 3072,
1819
+ "param_dtype": "bfloat16",
1820
+ "rank": 64
1821
+ },
1822
+ "layers.5.proj.o": {
1823
+ "alpha": 64.0,
1824
+ "dropout": 0.0,
1825
+ "in_features": 3072,
1826
+ "out_features": 3072,
1827
+ "param_dtype": "bfloat16",
1828
+ "rank": 64
1829
+ },
1830
+ "layers.5.proj.q": {
1831
+ "alpha": 64.0,
1832
+ "dropout": 0.0,
1833
+ "in_features": 3072,
1834
+ "out_features": 3072,
1835
+ "param_dtype": "bfloat16",
1836
+ "rank": 64
1837
+ },
1838
+ "layers.5.proj.v": {
1839
+ "alpha": 64.0,
1840
+ "dropout": 0.0,
1841
+ "in_features": 3072,
1842
+ "out_features": 3072,
1843
+ "param_dtype": "bfloat16",
1844
+ "rank": 64
1845
+ },
1846
+ "layers.6.proj.ffn_down": {
1847
+ "alpha": 64.0,
1848
+ "dropout": 0.0,
1849
+ "in_features": 8192,
1850
+ "out_features": 3072,
1851
+ "param_dtype": "bfloat16",
1852
+ "rank": 64
1853
+ },
1854
+ "layers.6.proj.ffn_up_gate": {
1855
+ "alpha": 64.0,
1856
+ "dropout": 0.0,
1857
+ "in_features": 3072,
1858
+ "out_features": 16384,
1859
+ "param_dtype": "bfloat16",
1860
+ "rank": 64
1861
+ },
1862
+ "layers.6.proj.k": {
1863
+ "alpha": 64.0,
1864
+ "dropout": 0.0,
1865
+ "in_features": 3072,
1866
+ "out_features": 3072,
1867
+ "param_dtype": "bfloat16",
1868
+ "rank": 64
1869
+ },
1870
+ "layers.6.proj.o": {
1871
+ "alpha": 64.0,
1872
+ "dropout": 0.0,
1873
+ "in_features": 3072,
1874
+ "out_features": 3072,
1875
+ "param_dtype": "bfloat16",
1876
+ "rank": 64
1877
+ },
1878
+ "layers.6.proj.q": {
1879
+ "alpha": 64.0,
1880
+ "dropout": 0.0,
1881
+ "in_features": 3072,
1882
+ "out_features": 3072,
1883
+ "param_dtype": "bfloat16",
1884
+ "rank": 64
1885
+ },
1886
+ "layers.6.proj.v": {
1887
+ "alpha": 64.0,
1888
+ "dropout": 0.0,
1889
+ "in_features": 3072,
1890
+ "out_features": 3072,
1891
+ "param_dtype": "bfloat16",
1892
+ "rank": 64
1893
+ },
1894
+ "layers.7.proj.ffn_down": {
1895
+ "alpha": 64.0,
1896
+ "dropout": 0.0,
1897
+ "in_features": 8192,
1898
+ "out_features": 3072,
1899
+ "param_dtype": "bfloat16",
1900
+ "rank": 64
1901
+ },
1902
+ "layers.7.proj.ffn_up_gate": {
1903
+ "alpha": 64.0,
1904
+ "dropout": 0.0,
1905
+ "in_features": 3072,
1906
+ "out_features": 16384,
1907
+ "param_dtype": "bfloat16",
1908
+ "rank": 64
1909
+ },
1910
+ "layers.7.proj.k": {
1911
+ "alpha": 64.0,
1912
+ "dropout": 0.0,
1913
+ "in_features": 3072,
1914
+ "out_features": 3072,
1915
+ "param_dtype": "bfloat16",
1916
+ "rank": 64
1917
+ },
1918
+ "layers.7.proj.o": {
1919
+ "alpha": 64.0,
1920
+ "dropout": 0.0,
1921
+ "in_features": 3072,
1922
+ "out_features": 3072,
1923
+ "param_dtype": "bfloat16",
1924
+ "rank": 64
1925
+ },
1926
+ "layers.7.proj.q": {
1927
+ "alpha": 64.0,
1928
+ "dropout": 0.0,
1929
+ "in_features": 3072,
1930
+ "out_features": 3072,
1931
+ "param_dtype": "bfloat16",
1932
+ "rank": 64
1933
+ },
1934
+ "layers.7.proj.v": {
1935
+ "alpha": 64.0,
1936
+ "dropout": 0.0,
1937
+ "in_features": 3072,
1938
+ "out_features": 3072,
1939
+ "param_dtype": "bfloat16",
1940
+ "rank": 64
1941
+ },
1942
+ "layers.8.proj.ffn_down": {
1943
+ "alpha": 64.0,
1944
+ "dropout": 0.0,
1945
+ "in_features": 8192,
1946
+ "out_features": 3072,
1947
+ "param_dtype": "bfloat16",
1948
+ "rank": 64
1949
+ },
1950
+ "layers.8.proj.ffn_up_gate": {
1951
+ "alpha": 64.0,
1952
+ "dropout": 0.0,
1953
+ "in_features": 3072,
1954
+ "out_features": 16384,
1955
+ "param_dtype": "bfloat16",
1956
+ "rank": 64
1957
+ },
1958
+ "layers.8.proj.k": {
1959
+ "alpha": 64.0,
1960
+ "dropout": 0.0,
1961
+ "in_features": 3072,
1962
+ "out_features": 3072,
1963
+ "param_dtype": "bfloat16",
1964
+ "rank": 64
1965
+ },
1966
+ "layers.8.proj.o": {
1967
+ "alpha": 64.0,
1968
+ "dropout": 0.0,
1969
+ "in_features": 3072,
1970
+ "out_features": 3072,
1971
+ "param_dtype": "bfloat16",
1972
+ "rank": 64
1973
+ },
1974
+ "layers.8.proj.q": {
1975
+ "alpha": 64.0,
1976
+ "dropout": 0.0,
1977
+ "in_features": 3072,
1978
+ "out_features": 3072,
1979
+ "param_dtype": "bfloat16",
1980
+ "rank": 64
1981
+ },
1982
+ "layers.8.proj.v": {
1983
+ "alpha": 64.0,
1984
+ "dropout": 0.0,
1985
+ "in_features": 3072,
1986
+ "out_features": 3072,
1987
+ "param_dtype": "bfloat16",
1988
+ "rank": 64
1989
+ },
1990
+ "layers.9.proj.ffn_down": {
1991
+ "alpha": 64.0,
1992
+ "dropout": 0.0,
1993
+ "in_features": 8192,
1994
+ "out_features": 3072,
1995
+ "param_dtype": "bfloat16",
1996
+ "rank": 64
1997
+ },
1998
+ "layers.9.proj.ffn_up_gate": {
1999
+ "alpha": 64.0,
2000
+ "dropout": 0.0,
2001
+ "in_features": 3072,
2002
+ "out_features": 16384,
2003
+ "param_dtype": "bfloat16",
2004
+ "rank": 64
2005
+ },
2006
+ "layers.9.proj.k": {
2007
+ "alpha": 64.0,
2008
+ "dropout": 0.0,
2009
+ "in_features": 3072,
2010
+ "out_features": 3072,
2011
+ "param_dtype": "bfloat16",
2012
+ "rank": 64
2013
+ },
2014
+ "layers.9.proj.o": {
2015
+ "alpha": 64.0,
2016
+ "dropout": 0.0,
2017
+ "in_features": 3072,
2018
+ "out_features": 3072,
2019
+ "param_dtype": "bfloat16",
2020
+ "rank": 64
2021
+ },
2022
+ "layers.9.proj.q": {
2023
+ "alpha": 64.0,
2024
+ "dropout": 0.0,
2025
+ "in_features": 3072,
2026
+ "out_features": 3072,
2027
+ "param_dtype": "bfloat16",
2028
+ "rank": 64
2029
+ },
2030
+ "layers.9.proj.v": {
2031
+ "alpha": 64.0,
2032
+ "dropout": 0.0,
2033
+ "in_features": 3072,
2034
+ "out_features": 3072,
2035
+ "param_dtype": "bfloat16",
2036
+ "rank": 64
2037
+ }
2038
+ },
2039
+ "targets": [
2040
+ "layers.0.proj.ffn_down",
2041
+ "layers.0.proj.ffn_up_gate",
2042
+ "layers.0.proj.k",
2043
+ "layers.0.proj.o",
2044
+ "layers.0.proj.q",
2045
+ "layers.0.proj.v",
2046
+ "layers.1.proj.ffn_down",
2047
+ "layers.1.proj.ffn_up_gate",
2048
+ "layers.1.proj.k",
2049
+ "layers.1.proj.o",
2050
+ "layers.1.proj.q",
2051
+ "layers.1.proj.v",
2052
+ "layers.10.proj.ffn_down",
2053
+ "layers.10.proj.ffn_up_gate",
2054
+ "layers.10.proj.k",
2055
+ "layers.10.proj.o",
2056
+ "layers.10.proj.q",
2057
+ "layers.10.proj.v",
2058
+ "layers.11.proj.ffn_down",
2059
+ "layers.11.proj.ffn_up_gate",
2060
+ "layers.11.proj.k",
2061
+ "layers.11.proj.o",
2062
+ "layers.11.proj.q",
2063
+ "layers.11.proj.v",
2064
+ "layers.12.proj.ffn_down",
2065
+ "layers.12.proj.ffn_up_gate",
2066
+ "layers.12.proj.k",
2067
+ "layers.12.proj.o",
2068
+ "layers.12.proj.q",
2069
+ "layers.12.proj.v",
2070
+ "layers.13.proj.ffn_down",
2071
+ "layers.13.proj.ffn_up_gate",
2072
+ "layers.13.proj.k",
2073
+ "layers.13.proj.o",
2074
+ "layers.13.proj.q",
2075
+ "layers.13.proj.v",
2076
+ "layers.14.proj.ffn_down",
2077
+ "layers.14.proj.ffn_up_gate",
2078
+ "layers.14.proj.k",
2079
+ "layers.14.proj.o",
2080
+ "layers.14.proj.q",
2081
+ "layers.14.proj.v",
2082
+ "layers.15.proj.ffn_down",
2083
+ "layers.15.proj.ffn_up_gate",
2084
+ "layers.15.proj.k",
2085
+ "layers.15.proj.o",
2086
+ "layers.15.proj.q",
2087
+ "layers.15.proj.v",
2088
+ "layers.16.proj.ffn_down",
2089
+ "layers.16.proj.ffn_up_gate",
2090
+ "layers.16.proj.k",
2091
+ "layers.16.proj.o",
2092
+ "layers.16.proj.q",
2093
+ "layers.16.proj.v",
2094
+ "layers.17.proj.ffn_down",
2095
+ "layers.17.proj.ffn_up_gate",
2096
+ "layers.17.proj.k",
2097
+ "layers.17.proj.o",
2098
+ "layers.17.proj.q",
2099
+ "layers.17.proj.v",
2100
+ "layers.18.proj.ffn_down",
2101
+ "layers.18.proj.ffn_up_gate",
2102
+ "layers.18.proj.k",
2103
+ "layers.18.proj.o",
2104
+ "layers.18.proj.q",
2105
+ "layers.18.proj.v",
2106
+ "layers.19.proj.ffn_down",
2107
+ "layers.19.proj.ffn_up_gate",
2108
+ "layers.19.proj.k",
2109
+ "layers.19.proj.o",
2110
+ "layers.19.proj.q",
2111
+ "layers.19.proj.v",
2112
+ "layers.2.proj.ffn_down",
2113
+ "layers.2.proj.ffn_up_gate",
2114
+ "layers.2.proj.k",
2115
+ "layers.2.proj.o",
2116
+ "layers.2.proj.q",
2117
+ "layers.2.proj.v",
2118
+ "layers.20.proj.ffn_down",
2119
+ "layers.20.proj.ffn_up_gate",
2120
+ "layers.20.proj.k",
2121
+ "layers.20.proj.o",
2122
+ "layers.20.proj.q",
2123
+ "layers.20.proj.v",
2124
+ "layers.21.proj.ffn_down",
2125
+ "layers.21.proj.ffn_up_gate",
2126
+ "layers.21.proj.k",
2127
+ "layers.21.proj.o",
2128
+ "layers.21.proj.q",
2129
+ "layers.21.proj.v",
2130
+ "layers.22.proj.ffn_down",
2131
+ "layers.22.proj.ffn_up_gate",
2132
+ "layers.22.proj.k",
2133
+ "layers.22.proj.o",
2134
+ "layers.22.proj.q",
2135
+ "layers.22.proj.v",
2136
+ "layers.23.proj.ffn_down",
2137
+ "layers.23.proj.ffn_up_gate",
2138
+ "layers.23.proj.k",
2139
+ "layers.23.proj.o",
2140
+ "layers.23.proj.q",
2141
+ "layers.23.proj.v",
2142
+ "layers.24.proj.ffn_down",
2143
+ "layers.24.proj.ffn_up_gate",
2144
+ "layers.24.proj.k",
2145
+ "layers.24.proj.o",
2146
+ "layers.24.proj.q",
2147
+ "layers.24.proj.v",
2148
+ "layers.25.proj.ffn_down",
2149
+ "layers.25.proj.ffn_up_gate",
2150
+ "layers.25.proj.k",
2151
+ "layers.25.proj.o",
2152
+ "layers.25.proj.q",
2153
+ "layers.25.proj.v",
2154
+ "layers.26.proj.ffn_down",
2155
+ "layers.26.proj.ffn_up_gate",
2156
+ "layers.26.proj.k",
2157
+ "layers.26.proj.o",
2158
+ "layers.26.proj.q",
2159
+ "layers.26.proj.v",
2160
+ "layers.27.proj.ffn_down",
2161
+ "layers.27.proj.ffn_up_gate",
2162
+ "layers.27.proj.k",
2163
+ "layers.27.proj.o",
2164
+ "layers.27.proj.q",
2165
+ "layers.27.proj.v",
2166
+ "layers.28.proj.ffn_down",
2167
+ "layers.28.proj.ffn_up_gate",
2168
+ "layers.28.proj.k",
2169
+ "layers.28.proj.o",
2170
+ "layers.28.proj.q",
2171
+ "layers.28.proj.v",
2172
+ "layers.29.proj.ffn_down",
2173
+ "layers.29.proj.ffn_up_gate",
2174
+ "layers.29.proj.k",
2175
+ "layers.29.proj.o",
2176
+ "layers.29.proj.q",
2177
+ "layers.29.proj.v",
2178
+ "layers.3.proj.ffn_down",
2179
+ "layers.3.proj.ffn_up_gate",
2180
+ "layers.3.proj.k",
2181
+ "layers.3.proj.o",
2182
+ "layers.3.proj.q",
2183
+ "layers.3.proj.v",
2184
+ "layers.30.proj.ffn_down",
2185
+ "layers.30.proj.ffn_up_gate",
2186
+ "layers.30.proj.k",
2187
+ "layers.30.proj.o",
2188
+ "layers.30.proj.q",
2189
+ "layers.30.proj.v",
2190
+ "layers.31.proj.ffn_down",
2191
+ "layers.31.proj.ffn_up_gate",
2192
+ "layers.31.proj.k",
2193
+ "layers.31.proj.o",
2194
+ "layers.31.proj.q",
2195
+ "layers.31.proj.v",
2196
+ "layers.32.proj.ffn_down",
2197
+ "layers.32.proj.ffn_up_gate",
2198
+ "layers.32.proj.k",
2199
+ "layers.32.proj.o",
2200
+ "layers.32.proj.q",
2201
+ "layers.32.proj.v",
2202
+ "layers.33.proj.ffn_down",
2203
+ "layers.33.proj.ffn_up_gate",
2204
+ "layers.33.proj.k",
2205
+ "layers.33.proj.o",
2206
+ "layers.33.proj.q",
2207
+ "layers.33.proj.v",
2208
+ "layers.34.proj.ffn_down",
2209
+ "layers.34.proj.ffn_up_gate",
2210
+ "layers.34.proj.k",
2211
+ "layers.34.proj.o",
2212
+ "layers.34.proj.q",
2213
+ "layers.34.proj.v",
2214
+ "layers.35.proj.ffn_down",
2215
+ "layers.35.proj.ffn_up_gate",
2216
+ "layers.35.proj.k",
2217
+ "layers.35.proj.o",
2218
+ "layers.35.proj.q",
2219
+ "layers.35.proj.v",
2220
+ "layers.36.proj.ffn_down",
2221
+ "layers.36.proj.ffn_up_gate",
2222
+ "layers.36.proj.k",
2223
+ "layers.36.proj.o",
2224
+ "layers.36.proj.q",
2225
+ "layers.36.proj.v",
2226
+ "layers.37.proj.ffn_down",
2227
+ "layers.37.proj.ffn_up_gate",
2228
+ "layers.37.proj.k",
2229
+ "layers.37.proj.o",
2230
+ "layers.37.proj.q",
2231
+ "layers.37.proj.v",
2232
+ "layers.38.proj.ffn_down",
2233
+ "layers.38.proj.ffn_up_gate",
2234
+ "layers.38.proj.k",
2235
+ "layers.38.proj.o",
2236
+ "layers.38.proj.q",
2237
+ "layers.38.proj.v",
2238
+ "layers.39.proj.ffn_down",
2239
+ "layers.39.proj.ffn_up_gate",
2240
+ "layers.39.proj.k",
2241
+ "layers.39.proj.o",
2242
+ "layers.39.proj.q",
2243
+ "layers.39.proj.v",
2244
+ "layers.4.proj.ffn_down",
2245
+ "layers.4.proj.ffn_up_gate",
2246
+ "layers.4.proj.k",
2247
+ "layers.4.proj.o",
2248
+ "layers.4.proj.q",
2249
+ "layers.4.proj.v",
2250
+ "layers.40.proj.ffn_down",
2251
+ "layers.40.proj.ffn_up_gate",
2252
+ "layers.40.proj.k",
2253
+ "layers.40.proj.o",
2254
+ "layers.40.proj.q",
2255
+ "layers.40.proj.v",
2256
+ "layers.41.proj.ffn_down",
2257
+ "layers.41.proj.ffn_up_gate",
2258
+ "layers.41.proj.k",
2259
+ "layers.41.proj.o",
2260
+ "layers.41.proj.q",
2261
+ "layers.41.proj.v",
2262
+ "layers.5.proj.ffn_down",
2263
+ "layers.5.proj.ffn_up_gate",
2264
+ "layers.5.proj.k",
2265
+ "layers.5.proj.o",
2266
+ "layers.5.proj.q",
2267
+ "layers.5.proj.v",
2268
+ "layers.6.proj.ffn_down",
2269
+ "layers.6.proj.ffn_up_gate",
2270
+ "layers.6.proj.k",
2271
+ "layers.6.proj.o",
2272
+ "layers.6.proj.q",
2273
+ "layers.6.proj.v",
2274
+ "layers.7.proj.ffn_down",
2275
+ "layers.7.proj.ffn_up_gate",
2276
+ "layers.7.proj.k",
2277
+ "layers.7.proj.o",
2278
+ "layers.7.proj.q",
2279
+ "layers.7.proj.v",
2280
+ "layers.8.proj.ffn_down",
2281
+ "layers.8.proj.ffn_up_gate",
2282
+ "layers.8.proj.k",
2283
+ "layers.8.proj.o",
2284
+ "layers.8.proj.q",
2285
+ "layers.8.proj.v",
2286
+ "layers.9.proj.ffn_down",
2287
+ "layers.9.proj.ffn_up_gate",
2288
+ "layers.9.proj.k",
2289
+ "layers.9.proj.o",
2290
+ "layers.9.proj.q",
2291
+ "layers.9.proj.v"
2292
+ ],
2293
+ "weights_sha256": "17a48b8c605dcffdedb92a2b887d18cf316505be90802717669fb6556098dcdf"
2294
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:17a48b8c605dcffdedb92a2b887d18cf316505be90802717669fb6556098dcdf
3
+ size 297382216
assets/moss-image-edit-banner-cases.png ADDED

Git LFS Details

  • SHA256: 65debff7419176c40c33f87d5fa2184e60ffedfdfe8253e159f53669d3ce4663
  • Pointer size: 132 Bytes
  • Size of remote file: 1.34 MB
assets/moss-image-edit-showcase.jpg ADDED

Git LFS Details

  • SHA256: 41baea80dc3931dba401d0692ef8322ea9c5bdb4e4cefe245db4063cb6fccaa7
  • Pointer size: 132 Bytes
  • Size of remote file: 4.19 MB