Hugging Face's logo

AlpachinoNLP
/
Baichuan-13B-Instruction

Text Generation
Transformers
PyTorch
Chinese
English
baichuan
custom_code
text-generation-inference
Model card Files Community
2
Redflashing commited on
Commit
9582aaf
•
1 Parent(s): 88683ce

Upload folder using huggingface_hub

Browse files
Files changed (11) hide show
  1. Baichuan-13B-Instruction.md +111 -0
  2. alpachino.png +0 -0
  3. config.json +2 -2
  4. pytorch_model-00001-of-00007.bin +3 -0
  5. pytorch_model-00002-of-00007.bin +3 -0
  6. pytorch_model-00003-of-00007.bin +3 -0
  7. pytorch_model-00004-of-00007.bin +3 -0
  8. pytorch_model-00005-of-00007.bin +3 -0
  9. pytorch_model-00006-of-00007.bin +3 -0
  10. pytorch_model-00007-of-00007.bin +3 -0
  11. pytorch_model.bin.index.json +290 -0
Baichuan-13B-Instruction.md ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - zh
4
+ - en
5
+ pipeline_tag: text-generation
6
+ inference: false
7
+ ---
8
+ # Baichuan-13B-Instruction
9
+
10
+ ![](./alpachino.png)
11
+
12
+ <!-- Provide a quick summary of what the model is/does. -->
13
+
14
+ ## 介绍
15
+ Baichuan-13B-Instruction 为 Baichuan-13B 系列模型进行指令微调后的版本，预训练模型可见 [Baichuan-13B-Base](https://huggingface.co/baichuan-inc/Baichuan-13B-Base)。
16
+
17
+
18
+ ## 使用方式
19
+
20
+ 如下是一个使用Baichuan-13B-Chat进行对话的示例，正确输出为"乔戈里峰。世界第二高峰———乔戈里峰西方登山者称其为k2峰，海拔高度是8611米，位于喀喇昆仑山脉的中巴边境上"
21
+ ```python
22
+ import torch
23
+ from transformers import AutoModelForCausalLM, AutoTokenizer
24
+ from transformers.generation.utils import GenerationConfig
25
+ tokenizer = AutoTokenizer.from_pretrained("AlpachinoNLP/Baichuan-13B-Instruction", use_fast=False, trust_remote_code=True)
26
+ model = AutoModelForCausalLM.from_pretrained("AlpachinoNLP/Baichuan-13B-Instruction", device_map="auto", torch_dtype=torch.float16, trust_remote_code=True)
27
+ model.generation_config = GenerationConfig.from_pretrained("AlpachinoNLP/Baichuan-13B-Instruction")
28
+ messages = []
29
+ messages.append({"role": "Human", "content": "世界上第二高的山峰是哪座"})
30
+ response = model.chat(tokenizer, messages)
31
+ print(response)
32
+ ```
33
+
34
+ ## 量化部署
35
+
36
+ Baichuan-13B 支持 int8 和 int4 量化，用户只需在推理代码中简单修改两行即可实现。请注意，如果是为了节省显存而进行量化，应加载原始精度模型到 CPU 后再开始量化；避免在 `from_pretrained` 时添加 `device_map='auto'` 或者其它会导致把原始精度模型直接加载到 GPU 的行为的参数。
37
+
38
+ 使用 int8 量化 (To use int8 quantization):
39
+ ```python
40
+ model = AutoModelForCausalLM.from_pretrained("AlpachinoNLP/Baichuan-13B-Instruction", torch_dtype=torch.float16, trust_remote_code=True)
41
+ model = model.quantize(8).cuda()
42
+ ```
43
+
44
+ 同样的，如需使用 int4 量化 (Similarly, to use int4 quantization):
45
+ ```python
46
+ model = AutoModelForCausalLM.from_pretrained("AlpachinoNLP/Baichuan-13B-Instruction", torch_dtype=torch.float16, trust_remote_code=True)
47
+ model = model.quantize(4).cuda()
48
+ ```
49
+
50
+ ## 模型详情
51
+
52
+
53
+ ### 模型结构
54
+
55
+ <!-- Provide the basic links for the model. -->
56
+
57
+ 整体模型基于Baichuan-13B，为了获得更好的推理性能，Baichuan-13B 使用了 ALiBi 线性偏置技术，相对于 Rotary Embedding 计算量更小，对推理性能有显著提升；与标准的 LLaMA-13B 相比，生成 2000 个 tokens 的平均推理速度 (tokens/s)，实测提升 31.6%：
58
+
59
+ | Model | tokens/s |
60
+ | ------------ | -------- |
61
+ | LLaMA-13B | 19.4 |
62
+ | Baichuan-13B | 25.4 |
63
+
64
+ 具体参数和见下表
65
+ | 模型名称 | 隐含层维度 | 层数 | 头数 | 词表大小 | 总参数量 | 训练数据（tokens） | 位置编码 | 最大长度 |
66
+ | ------------ | ---------- | ---- | ---- | -------- | -------------- | ------------------ | ----------------------------------------- | -------- |
67
+ | Baichuan-7B | 4,096 | 32 | 32 | 64,000 | 7,000,559,616 | 1.2万亿 | [RoPE](https://arxiv.org/abs/2104.09864) | 4,096 |
68
+ | Baichuan-13B | 5,120 | 40 | 40 | 64,000 | 13,264,901,120 | 1.4万亿 | [ALiBi](https://arxiv.org/abs/2108.12409) | 4,096 |
69
+
70
+ ## 训练详情
71
+
72
+ 数据集主要由三部分组成：
73
+
74
+ * 在 [sharegpt_zh](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/ShareGPT) 数据集中筛选的出 13k 高质量数据。
75
+ * [lima](https://huggingface.co/datasets/GAIR/lima)
76
+ * 按照任务类型挑选的 2.3k 高质量中文数据集，每个任务类型的数据量在 100 条左右。
77
+
78
+ 硬件：8*A40
79
+
80
+ ## 测评结果
81
+
82
+ ## [CMMLU](https://github.com/haonan-li/CMMLU)
83
+
84
+ | Model 5-shot | STEM | Humanities | Social Sciences | Others | China Specific | Average |
85
+ | ---------------------------------------------------------- | :-------: | :--------: | :-------------: | :------: | :------------: | :------: |
86
+ | Baichuan-7B | 34.4 | 47.5 | 47.6 | 46.6 | 44.3 | 44.0 |
87
+ | Vicuna-13B | 31.8 | 36.2 | 37.6 | 39.5 | 34.3 | 36.3 |
88
+ | Chinese-Alpaca-Plus-13B | 29.8 | 33.4 | 33.2 | 37.9 | 32.1 | 33.4 |
89
+ | Chinese-LLaMA-Plus-13B | 28.1 | 33.1 | 35.4 | 35.1 | 33.5 | 33.0 |
90
+ | Ziya-LLaMA-13B-Pretrain | 29.0 | 30.7 | 33.8 | 34.4 | 31.9 | 32.1 |
91
+ | LLaMA-13B | 29.2 | 30.8 | 31.6 | 33.0 | 30.5 | 31.2 |
92
+ | moss-moon-003-base (16B) | 27.2 | 30.4 | 28.8 | 32.6 | 28.7 | 29.6 |
93
+ | Baichuan-13B-Base | 41.7 | 61.1 | 59.8 | 59.0 | 56.4 | 55.3 |
94
+ | Baichuan-13B-Chat | 42.8 | **62.6** | **59.7** | **59.0** | **56.1** | **55.8** |
95
+ | **Baichuan-13B-Instruction** | **44.50** | 61.16 | 59.07 | 58.34 | 55.55 | 55.61 |
96
+
97
+ | Model zero-shot | STEM | Humanities | Social Sciences | Others | China Specific | Average |
98
+ | ------------------------------------------------------------ | :-------: | :--------: | :-------------: | :-------: | :------------: | :-------: |
99
+ | [ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b) | 41.28 | 52.85 | 53.37 | 52.24 | 50.58 | 49.95 |
100
+ | [Baichuan-7B](https://github.com/baichuan-inc/baichuan-7B) | 32.79 | 44.43 | 46.78 | 44.79 | 43.11 | 42.33 |
101
+ | [ChatGLM-6B](https://github.com/THUDM/GLM-130B) | 32.22 | 42.91 | 44.81 | 42.60 | 41.93 | 40.79 |
102
+ | [BatGPT-15B](https://arxiv.org/abs/2307.00360) | 33.72 | 36.53 | 38.07 | 46.94 | 38.32 | 38.51 |
103
+ | [Chinese-LLaMA-13B](https://github.com/ymcui/Chinese-LLaMA-Alpaca) | 26.76 | 26.57 | 27.42 | 28.33 | 26.73 | 27.34 |
104
+ | [MOSS-SFT-16B](https://github.com/OpenLMLab/MOSS) | 25.68 | 26.35 | 27.21 | 27.92 | 26.70 | 26.88 |
105
+ | [Chinese-GLM-10B](https://github.com/THUDM/GLM) | 25.57 | 25.01 | 26.33 | 25.94 | 25.81 | 25.80 |
106
+ | [Baichuan-13B](https://github.com/baichuan-inc/Baichuan-13B) | 42.04 | 60.49 | 59.55 | 56.60 | 55.72 | 54.63 |
107
+ | [Baichuan-13B-Chat](https://github.com/baichuan-inc/Baichuan-13B) | 37.32 | 56.24 | 54.79 | 54.07 | 52.23 | 50.48 |
108
+ | **Baichuan-13B-Instruction** | **42.56** | **62.09** | **60.41** | **58.97** | **56.95** | **55.88** |
109
+
110
+ > 说明：CMMLU 是一个综合性的中文评估基准，专门用于评估语言模型在中文语境下的知识和推理能力。我们直接使用其官方的[评测脚本](https://github.com/haonan-li/CMMLU)对模型进行评测。Model zero-shot 表格中 [Baichuan-13B-Chat](https://github.com/baichuan-inc/Baichuan-13B) 的得分来自我们直接运行 CMMLU 官方的评测脚本得到，其他模型的的得分来自于 [CMMLU](https://github.com/haonan-li/CMMLU/tree/master) 官方的评测结果，Model 5-shot 中其他模型的得分来自于[Baichuan-13B](https://github.com/baichuan-inc/Baichuan-13B) 官方的评测结果。
111
+
alpachino.png ADDED
config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "_from_model_config": true,
3
- "_name_or_path": "/data/checkpoints/Baichuan-13B-Base",
4
  "architectures": [
5
  "BaichuanForCausalLM"
6
  ],
@@ -21,7 +21,7 @@
21
  "pad_token_id": 0,
22
  "rms_norm_eps": 1e-06,
23
  "tie_word_embeddings": false,
24
- "torch_dtype": "bfloat16",
25
  "transformers_version": "4.29.2",
26
  "use_cache": true,
27
  "vocab_size": 64000
 
1
  {
2
  "_from_model_config": true,
3
+ "_name_or_path": "/data/Mermaid/baichuan_train/checkpoints/baichuan-13b-zh-0717-sharegpt-plus-baichuan3k",
4
  "architectures": [
5
  "BaichuanForCausalLM"
6
  ],
 
21
  "pad_token_id": 0,
22
  "rms_norm_eps": 1e-06,
23
  "tie_word_embeddings": false,
24
+ "torch_dtype": "float16",
25
  "transformers_version": "4.29.2",
26
  "use_cache": true,
27
  "vocab_size": 64000
pytorch_model-00001-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:223f471338e4ac43b132170200516f78312682090386c4716855036195b0780e
3
+ size 3965042665
pytorch_model-00002-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf9e048d7ba6dab5937496340a82a0fc87b30d6aededdd47dcaf8285619f2524
3
+ size 3975551335
pytorch_model-00003-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:467f678c0c4c759e8c51f1ebc35a1def3f12b5277b30e9d4bbc281b593bf6aba
3
+ size 3923122197
pytorch_model-00004-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0836ec45735c9249608229f65e3d527a63688eef22a2e76adf9cbe2f88f52ba7
3
+ size 3923143353
pytorch_model-00005-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86eebec2b3ec035c2bec0cd1803e3bcbfb5639b3c4107afa1612767a31fbf9c5
3
+ size 3992590695
pytorch_model-00006-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:876b91915719acb81ffd6b935925ef4262aac7c047ab43333cbe9e51ab295262
3
+ size 3923122197
pytorch_model-00007-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:962ebb5b42e890770d800fd8e1e0dc46e96fe4eadacca52c9691d18da7de3bbd
3
+ size 2827324213
pytorch_model.bin.index.json ADDED
@@ -0,0 +1,290 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 26529802240
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "pytorch_model-00007-of-00007.bin",
7
+ "model.embed_tokens.weight": "pytorch_model-00001-of-00007.bin",
8
+ "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00007.bin",
9
+ "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00007.bin",
10
+ "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00007.bin",
11
+ "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00007.bin",
12
+ "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00007.bin",
13
+ "model.layers.0.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
14
+ "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00007.bin",
15
+ "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00007.bin",
16
+ "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00007.bin",
17
+ "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00007.bin",
18
+ "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00007.bin",
19
+ "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00007.bin",
20
+ "model.layers.1.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
21
+ "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00007.bin",
22
+ "model.layers.10.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
23
+ "model.layers.10.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
24
+ "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
25
+ "model.layers.10.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
26
+ "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
27
+ "model.layers.10.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
28
+ "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
29
+ "model.layers.11.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
30
+ "model.layers.11.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
31
+ "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
32
+ "model.layers.11.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
33
+ "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
34
+ "model.layers.11.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
35
+ "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
36
+ "model.layers.12.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
37
+ "model.layers.12.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
38
+ "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
39
+ "model.layers.12.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
40
+ "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
41
+ "model.layers.12.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
42
+ "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
43
+ "model.layers.13.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
44
+ "model.layers.13.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
45
+ "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
46
+ "model.layers.13.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
47
+ "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
48
+ "model.layers.13.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
49
+ "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
50
+ "model.layers.14.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
51
+ "model.layers.14.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
52
+ "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
53
+ "model.layers.14.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
54
+ "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
55
+ "model.layers.14.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
56
+ "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
57
+ "model.layers.15.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
58
+ "model.layers.15.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
59
+ "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
60
+ "model.layers.15.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
61
+ "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
62
+ "model.layers.15.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
63
+ "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
64
+ "model.layers.16.input_layernorm.weight": "pytorch_model-00003-of-00007.bin",
65
+ "model.layers.16.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
66
+ "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
67
+ "model.layers.16.mlp.up_proj.weight": "pytorch_model-00003-of-00007.bin",
68
+ "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00003-of-00007.bin",
69
+ "model.layers.16.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
70
+ "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
71
+ "model.layers.17.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
72
+ "model.layers.17.mlp.down_proj.weight": "pytorch_model-00003-of-00007.bin",
73
+ "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00003-of-00007.bin",
74
+ "model.layers.17.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
75
+ "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
76
+ "model.layers.17.self_attn.W_pack.weight": "pytorch_model-00003-of-00007.bin",
77
+ "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00003-of-00007.bin",
78
+ "model.layers.18.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
79
+ "model.layers.18.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
80
+ "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
81
+ "model.layers.18.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
82
+ "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
83
+ "model.layers.18.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
84
+ "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
85
+ "model.layers.19.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
86
+ "model.layers.19.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
87
+ "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
88
+ "model.layers.19.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
89
+ "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
90
+ "model.layers.19.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
91
+ "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
92
+ "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00007.bin",
93
+ "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00007.bin",
94
+ "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00007.bin",
95
+ "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00007.bin",
96
+ "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00007.bin",
97
+ "model.layers.2.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
98
+ "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00007.bin",
99
+ "model.layers.20.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
100
+ "model.layers.20.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
101
+ "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
102
+ "model.layers.20.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
103
+ "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
104
+ "model.layers.20.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
105
+ "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
106
+ "model.layers.21.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
107
+ "model.layers.21.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
108
+ "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
109
+ "model.layers.21.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
110
+ "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
111
+ "model.layers.21.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
112
+ "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
113
+ "model.layers.22.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
114
+ "model.layers.22.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
115
+ "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
116
+ "model.layers.22.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
117
+ "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
118
+ "model.layers.22.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
119
+ "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
120
+ "model.layers.23.input_layernorm.weight": "pytorch_model-00004-of-00007.bin",
121
+ "model.layers.23.mlp.down_proj.weight": "pytorch_model-00004-of-00007.bin",
122
+ "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00004-of-00007.bin",
123
+ "model.layers.23.mlp.up_proj.weight": "pytorch_model-00004-of-00007.bin",
124
+ "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00004-of-00007.bin",
125
+ "model.layers.23.self_attn.W_pack.weight": "pytorch_model-00004-of-00007.bin",
126
+ "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00004-of-00007.bin",
127
+ "model.layers.24.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
128
+ "model.layers.24.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
129
+ "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
130
+ "model.layers.24.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
131
+ "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
132
+ "model.layers.24.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
133
+ "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
134
+ "model.layers.25.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
135
+ "model.layers.25.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
136
+ "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
137
+ "model.layers.25.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
138
+ "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
139
+ "model.layers.25.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
140
+ "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
141
+ "model.layers.26.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
142
+ "model.layers.26.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
143
+ "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
144
+ "model.layers.26.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
145
+ "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
146
+ "model.layers.26.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
147
+ "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
148
+ "model.layers.27.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
149
+ "model.layers.27.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
150
+ "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
151
+ "model.layers.27.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
152
+ "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
153
+ "model.layers.27.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
154
+ "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
155
+ "model.layers.28.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
156
+ "model.layers.28.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
157
+ "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
158
+ "model.layers.28.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
159
+ "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
160
+ "model.layers.28.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
161
+ "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
162
+ "model.layers.29.input_layernorm.weight": "pytorch_model-00005-of-00007.bin",
163
+ "model.layers.29.mlp.down_proj.weight": "pytorch_model-00005-of-00007.bin",
164
+ "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00005-of-00007.bin",
165
+ "model.layers.29.mlp.up_proj.weight": "pytorch_model-00005-of-00007.bin",
166
+ "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00005-of-00007.bin",
167
+ "model.layers.29.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
168
+ "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
169
+ "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00007.bin",
170
+ "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00007.bin",
171
+ "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00007.bin",
172
+ "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00007.bin",
173
+ "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00007.bin",
174
+ "model.layers.3.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
175
+ "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00007.bin",
176
+ "model.layers.30.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
177
+ "model.layers.30.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
178
+ "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
179
+ "model.layers.30.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
180
+ "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
181
+ "model.layers.30.self_attn.W_pack.weight": "pytorch_model-00005-of-00007.bin",
182
+ "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00005-of-00007.bin",
183
+ "model.layers.31.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
184
+ "model.layers.31.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
185
+ "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
186
+ "model.layers.31.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
187
+ "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
188
+ "model.layers.31.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
189
+ "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
190
+ "model.layers.32.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
191
+ "model.layers.32.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
192
+ "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
193
+ "model.layers.32.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
194
+ "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
195
+ "model.layers.32.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
196
+ "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
197
+ "model.layers.33.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
198
+ "model.layers.33.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
199
+ "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
200
+ "model.layers.33.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
201
+ "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
202
+ "model.layers.33.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
203
+ "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
204
+ "model.layers.34.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
205
+ "model.layers.34.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
206
+ "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
207
+ "model.layers.34.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
208
+ "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
209
+ "model.layers.34.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
210
+ "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
211
+ "model.layers.35.input_layernorm.weight": "pytorch_model-00006-of-00007.bin",
212
+ "model.layers.35.mlp.down_proj.weight": "pytorch_model-00006-of-00007.bin",
213
+ "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
214
+ "model.layers.35.mlp.up_proj.weight": "pytorch_model-00006-of-00007.bin",
215
+ "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00006-of-00007.bin",
216
+ "model.layers.35.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
217
+ "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
218
+ "model.layers.36.input_layernorm.weight": "pytorch_model-00007-of-00007.bin",
219
+ "model.layers.36.mlp.down_proj.weight": "pytorch_model-00007-of-00007.bin",
220
+ "model.layers.36.mlp.gate_proj.weight": "pytorch_model-00006-of-00007.bin",
221
+ "model.layers.36.mlp.up_proj.weight": "pytorch_model-00007-of-00007.bin",
222
+ "model.layers.36.post_attention_layernorm.weight": "pytorch_model-00007-of-00007.bin",
223
+ "model.layers.36.self_attn.W_pack.weight": "pytorch_model-00006-of-00007.bin",
224
+ "model.layers.36.self_attn.o_proj.weight": "pytorch_model-00006-of-00007.bin",
225
+ "model.layers.37.input_layernorm.weight": "pytorch_model-00007-of-00007.bin",
226
+ "model.layers.37.mlp.down_proj.weight": "pytorch_model-00007-of-00007.bin",
227
+ "model.layers.37.mlp.gate_proj.weight": "pytorch_model-00007-of-00007.bin",
228
+ "model.layers.37.mlp.up_proj.weight": "pytorch_model-00007-of-00007.bin",
229
+ "model.layers.37.post_attention_layernorm.weight": "pytorch_model-00007-of-00007.bin",
230
+ "model.layers.37.self_attn.W_pack.weight": "pytorch_model-00007-of-00007.bin",
231
+ "model.layers.37.self_attn.o_proj.weight": "pytorch_model-00007-of-00007.bin",
232
+ "model.layers.38.input_layernorm.weight": "pytorch_model-00007-of-00007.bin",
233
+ "model.layers.38.mlp.down_proj.weight": "pytorch_model-00007-of-00007.bin",
234
+ "model.layers.38.mlp.gate_proj.weight": "pytorch_model-00007-of-00007.bin",
235
+ "model.layers.38.mlp.up_proj.weight": "pytorch_model-00007-of-00007.bin",
236
+ "model.layers.38.post_attention_layernorm.weight": "pytorch_model-00007-of-00007.bin",
237
+ "model.layers.38.self_attn.W_pack.weight": "pytorch_model-00007-of-00007.bin",
238
+ "model.layers.38.self_attn.o_proj.weight": "pytorch_model-00007-of-00007.bin",
239
+ "model.layers.39.input_layernorm.weight": "pytorch_model-00007-of-00007.bin",
240
+ "model.layers.39.mlp.down_proj.weight": "pytorch_model-00007-of-00007.bin",
241
+ "model.layers.39.mlp.gate_proj.weight": "pytorch_model-00007-of-00007.bin",
242
+ "model.layers.39.mlp.up_proj.weight": "pytorch_model-00007-of-00007.bin",
243
+ "model.layers.39.post_attention_layernorm.weight": "pytorch_model-00007-of-00007.bin",
244
+ "model.layers.39.self_attn.W_pack.weight": "pytorch_model-00007-of-00007.bin",
245
+ "model.layers.39.self_attn.o_proj.weight": "pytorch_model-00007-of-00007.bin",
246
+ "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00007.bin",
247
+ "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00007.bin",
248
+ "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00007.bin",
249
+ "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00007.bin",
250
+ "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00007.bin",
251
+ "model.layers.4.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
252
+ "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00007.bin",
253
+ "model.layers.5.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
254
+ "model.layers.5.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
255
+ "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
256
+ "model.layers.5.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
257
+ "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
258
+ "model.layers.5.self_attn.W_pack.weight": "pytorch_model-00001-of-00007.bin",
259
+ "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
260
+ "model.layers.6.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
261
+ "model.layers.6.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
262
+ "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
263
+ "model.layers.6.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
264
+ "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
265
+ "model.layers.6.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
266
+ "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
267
+ "model.layers.7.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
268
+ "model.layers.7.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
269
+ "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
270
+ "model.layers.7.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
271
+ "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
272
+ "model.layers.7.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
273
+ "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
274
+ "model.layers.8.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
275
+ "model.layers.8.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
276
+ "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
277
+ "model.layers.8.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
278
+ "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
279
+ "model.layers.8.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
280
+ "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
281
+ "model.layers.9.input_layernorm.weight": "pytorch_model-00002-of-00007.bin",
282
+ "model.layers.9.mlp.down_proj.weight": "pytorch_model-00002-of-00007.bin",
283
+ "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00002-of-00007.bin",
284
+ "model.layers.9.mlp.up_proj.weight": "pytorch_model-00002-of-00007.bin",
285
+ "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00002-of-00007.bin",
286
+ "model.layers.9.self_attn.W_pack.weight": "pytorch_model-00002-of-00007.bin",
287
+ "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00002-of-00007.bin",
288
+ "model.norm.weight": "pytorch_model-00007-of-00007.bin"
289
+ }
290
+ }