Hugging Face's logo

charlescao2023
/
vicuna-13b-v1.5-medusa1-3head

Model card Files Community
charlescao2023 commited on
Commit
58cb504
·
verified ·
1 Parent(s): 892faa8

Upload folder using huggingface_hub

Browse files
Files changed (13) hide show
  1. README.md +13 -0
  2. base/README.md +110 -0
  3. base/config.json +31 -0
  4. base/pytorch_model-00001-of-00003.bin +3 -0
  5. base/pytorch_model-00002-of-00003.bin +3 -0
  6. base/pytorch_model-00003-of-00003.bin +3 -0
  7. base/pytorch_model.bin.index.json +379 -0
  8. base/special_tokens_map.json +30 -0
  9. base/tokenizer.model +3 -0
  10. base/tokenizer_config.json +44 -0
  11. head/config.json +7 -0
  12. head/medusa_lm_head.pt +3 -0
  13. medusa_model_demo.py +56 -0
README.md ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ base: include Llama base model and medusa head
2
+ head: only medusa head
3
+
4
+ ## convert
5
+ 1. prepare medusa
6
+ ```bash
7
+ git clone https://github.com/FasterDecoding/Medusa.git
8
+ cd Medusa
9
+ pip install -e .
10
+ ```
11
+
12
+ 2. python3 medusa_model_demo.py -m base -o head --use-full-key --verbose
13
+
base/README.md ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ tags:
3
+ - generated_from_trainer
4
+ model-index:
5
+ - name: vicuna_13b_stage1
6
+ results: []
7
+ ---
8
+
9
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
10
+ should probably proofread and complete it, then remove this comment. -->
11
+
12
+ [<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)
13
+ # vicuna_13b_stage1
14
+
15
+ This model was trained from scratch on the None dataset.
16
+ It achieves the following results on the evaluation set:
17
+ - Loss: 1.2017
18
+
19
+ ## Model description
20
+
21
+ More information needed
22
+
23
+ ## Intended uses & limitations
24
+
25
+ More information needed
26
+
27
+ ## Training and evaluation data
28
+
29
+ More information needed
30
+
31
+ ## Training procedure
32
+
33
+ ### Training hyperparameters
34
+
35
+ The following hyperparameters were used during training:
36
+ - learning_rate: 0.0005
37
+ - train_batch_size: 2
38
+ - eval_batch_size: 2
39
+ - seed: 42
40
+ - distributed_type: multi-GPU
41
+ - num_devices: 2
42
+ - gradient_accumulation_steps: 4
43
+ - total_train_batch_size: 16
44
+ - total_eval_batch_size: 4
45
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
+ - lr_scheduler_type: cosine
47
+ - lr_scheduler_warmup_steps: 40
48
+ - num_epochs: 1
49
+
50
+ ### Training results
51
+
52
+ | Training Loss | Epoch | Step | Validation Loss |
53
+ |:-------------:|:-----:|:----:|:---------------:|
54
+ | 1.9535 | 0.02 | 40 | 1.9456 |
55
+ | 1.8556 | 0.04 | 80 | 1.7714 |
56
+ | 1.791 | 0.06 | 120 | 1.7425 |
57
+ | 1.6622 | 0.08 | 160 | 1.7164 |
58
+ | 1.8169 | 0.1 | 200 | 1.7154 |
59
+ | 1.7356 | 0.12 | 240 | 1.7026 |
60
+ | 1.6051 | 0.14 | 280 | 1.7104 |
61
+ | 1.7925 | 0.16 | 320 | 1.7127 |
62
+ | 1.8257 | 0.18 | 360 | 1.7055 |
63
+ | 1.7057 | 0.2 | 400 | 1.6906 |
64
+ | 1.9282 | 0.22 | 440 | 1.6746 |
65
+ | 1.668 | 0.24 | 480 | 1.7052 |
66
+ | 1.6273 | 0.26 | 520 | 1.6620 |
67
+ | 1.6136 | 0.28 | 560 | 1.6616 |
68
+ | 1.4754 | 0.3 | 600 | 1.6389 |
69
+ | 1.4024 | 0.32 | 640 | 1.6038 |
70
+ | 1.6773 | 0.34 | 680 | 1.5743 |
71
+ | 1.6008 | 0.36 | 720 | 1.5607 |
72
+ | 1.568 | 0.39 | 760 | 1.5236 |
73
+ | 1.4922 | 0.41 | 800 | 1.5158 |
74
+ | 1.4667 | 0.43 | 840 | 1.4938 |
75
+ | 1.5653 | 0.45 | 880 | 1.4692 |
76
+ | 1.331 | 0.47 | 920 | 1.4581 |
77
+ | 1.4019 | 0.49 | 960 | 1.4290 |
78
+ | 1.4925 | 0.51 | 1000 | 1.4087 |
79
+ | 1.4772 | 0.53 | 1040 | 1.3961 |
80
+ | 1.4728 | 0.55 | 1080 | 1.3817 |
81
+ | 1.4555 | 0.57 | 1120 | 1.3559 |
82
+ | 1.5487 | 0.59 | 1160 | 1.3399 |
83
+ | 1.3888 | 0.61 | 1200 | 1.3212 |
84
+ | 1.2544 | 0.63 | 1240 | 1.3099 |
85
+ | 1.2657 | 0.65 | 1280 | 1.2972 |
86
+ | 1.3641 | 0.67 | 1320 | 1.2815 |
87
+ | 1.2915 | 0.69 | 1360 | 1.2687 |
88
+ | 1.4182 | 0.71 | 1400 | 1.2541 |
89
+ | 1.2515 | 0.73 | 1440 | 1.2427 |
90
+ | 1.2287 | 0.75 | 1480 | 1.2352 |
91
+ | 1.1886 | 0.77 | 1520 | 1.2285 |
92
+ | 1.2651 | 0.79 | 1560 | 1.2219 |
93
+ | 1.3341 | 0.81 | 1600 | 1.2145 |
94
+ | 1.2357 | 0.83 | 1640 | 1.2107 |
95
+ | 1.0767 | 0.85 | 1680 | 1.2080 |
96
+ | 1.2158 | 0.87 | 1720 | 1.2051 |
97
+ | 1.2042 | 0.89 | 1760 | 1.2034 |
98
+ | 1.1887 | 0.91 | 1800 | 1.2023 |
99
+ | 1.2662 | 0.93 | 1840 | 1.2018 |
100
+ | 1.1866 | 0.95 | 1880 | 1.2017 |
101
+ | 1.1798 | 0.97 | 1920 | 1.2017 |
102
+ | 1.336 | 0.99 | 1960 | 1.2017 |
103
+
104
+
105
+ ### Framework versions
106
+
107
+ - Transformers 4.34.1
108
+ - Pytorch 2.3.1+cu121
109
+ - Datasets 2.14.7
110
+ - Tokenizers 0.14.1
base/config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "vicuna-13b-v1.5",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "bos_token_id": 1,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 5120,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 13824,
13
+ "max_length": 4096,
14
+ "max_position_embeddings": 4096,
15
+ "medusa_num_heads": 3,
16
+ "medusa_num_layers": 1,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 40,
19
+ "num_hidden_layers": 40,
20
+ "num_key_value_heads": 40,
21
+ "pad_token_id": 0,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-05,
24
+ "rope_scaling": null,
25
+ "rope_theta": 10000.0,
26
+ "tie_word_embeddings": false,
27
+ "torch_dtype": "bfloat16",
28
+ "transformers_version": "4.34.1",
29
+ "use_cache": false,
30
+ "vocab_size": 32000
31
+ }
base/pytorch_model-00001-of-00003.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41a1924f2e72d81283cceb852ab0bdb2dba5166ed6fbf962c62cb8eb114b266f
3
+ size 9948721840
base/pytorch_model-00002-of-00003.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:eb643d70e034911e9aee1700d132250da26d4a35ca47123d523b9793347b673d
3
+ size 9904157815
base/pytorch_model-00003-of-00003.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd8dd403ad067a81ed22f54e8652df78cd325911721b0c3e738ab375c95172e4
3
+ size 7319340180
base/pytorch_model.bin.index.json ADDED
@@ -0,0 +1,379 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 27172085760
4
+ },
5
+ "weight_map": {
6
+ "lm_head.weight": "pytorch_model-00003-of-00003.bin",
7
+ "medusa_head.0.0.linear.bias": "pytorch_model-00003-of-00003.bin",
8
+ "medusa_head.0.0.linear.weight": "pytorch_model-00003-of-00003.bin",
9
+ "medusa_head.0.1.weight": "pytorch_model-00003-of-00003.bin",
10
+ "medusa_head.1.0.linear.bias": "pytorch_model-00003-of-00003.bin",
11
+ "medusa_head.1.0.linear.weight": "pytorch_model-00003-of-00003.bin",
12
+ "medusa_head.1.1.weight": "pytorch_model-00003-of-00003.bin",
13
+ "medusa_head.2.0.linear.bias": "pytorch_model-00003-of-00003.bin",
14
+ "medusa_head.2.0.linear.weight": "pytorch_model-00003-of-00003.bin",
15
+ "medusa_head.2.1.weight": "pytorch_model-00003-of-00003.bin",
16
+ "model.embed_tokens.weight": "pytorch_model-00001-of-00003.bin",
17
+ "model.layers.0.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
18
+ "model.layers.0.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
19
+ "model.layers.0.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
20
+ "model.layers.0.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
21
+ "model.layers.0.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
22
+ "model.layers.0.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
23
+ "model.layers.0.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
24
+ "model.layers.0.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
25
+ "model.layers.0.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
26
+ "model.layers.1.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
27
+ "model.layers.1.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
28
+ "model.layers.1.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
29
+ "model.layers.1.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
30
+ "model.layers.1.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
31
+ "model.layers.1.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
32
+ "model.layers.1.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
33
+ "model.layers.1.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
34
+ "model.layers.1.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
35
+ "model.layers.10.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
36
+ "model.layers.10.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
37
+ "model.layers.10.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
38
+ "model.layers.10.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
39
+ "model.layers.10.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
40
+ "model.layers.10.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
41
+ "model.layers.10.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
42
+ "model.layers.10.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
43
+ "model.layers.10.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
44
+ "model.layers.11.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
45
+ "model.layers.11.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
46
+ "model.layers.11.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
47
+ "model.layers.11.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
48
+ "model.layers.11.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
49
+ "model.layers.11.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
50
+ "model.layers.11.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
51
+ "model.layers.11.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
52
+ "model.layers.11.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
53
+ "model.layers.12.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
54
+ "model.layers.12.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
55
+ "model.layers.12.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
56
+ "model.layers.12.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
57
+ "model.layers.12.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
58
+ "model.layers.12.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
59
+ "model.layers.12.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
60
+ "model.layers.12.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
61
+ "model.layers.12.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
62
+ "model.layers.13.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
63
+ "model.layers.13.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
64
+ "model.layers.13.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
65
+ "model.layers.13.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
66
+ "model.layers.13.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
67
+ "model.layers.13.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
68
+ "model.layers.13.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
69
+ "model.layers.13.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
70
+ "model.layers.13.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
71
+ "model.layers.14.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
72
+ "model.layers.14.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
73
+ "model.layers.14.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
74
+ "model.layers.14.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
75
+ "model.layers.14.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
76
+ "model.layers.14.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
77
+ "model.layers.14.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
78
+ "model.layers.14.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
79
+ "model.layers.14.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
80
+ "model.layers.15.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
81
+ "model.layers.15.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
82
+ "model.layers.15.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
83
+ "model.layers.15.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
84
+ "model.layers.15.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
85
+ "model.layers.15.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
86
+ "model.layers.15.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
87
+ "model.layers.15.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
88
+ "model.layers.15.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
89
+ "model.layers.16.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
90
+ "model.layers.16.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
91
+ "model.layers.16.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
92
+ "model.layers.16.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
93
+ "model.layers.16.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
94
+ "model.layers.16.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
95
+ "model.layers.16.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
96
+ "model.layers.16.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
97
+ "model.layers.16.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
98
+ "model.layers.17.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
99
+ "model.layers.17.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
100
+ "model.layers.17.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
101
+ "model.layers.17.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
102
+ "model.layers.17.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
103
+ "model.layers.17.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
104
+ "model.layers.17.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
105
+ "model.layers.17.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
106
+ "model.layers.17.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
107
+ "model.layers.18.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
108
+ "model.layers.18.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
109
+ "model.layers.18.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
110
+ "model.layers.18.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
111
+ "model.layers.18.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
112
+ "model.layers.18.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
113
+ "model.layers.18.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
114
+ "model.layers.18.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
115
+ "model.layers.18.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
116
+ "model.layers.19.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
117
+ "model.layers.19.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
118
+ "model.layers.19.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
119
+ "model.layers.19.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
120
+ "model.layers.19.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
121
+ "model.layers.19.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
122
+ "model.layers.19.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
123
+ "model.layers.19.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
124
+ "model.layers.19.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
125
+ "model.layers.2.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
126
+ "model.layers.2.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
127
+ "model.layers.2.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
128
+ "model.layers.2.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
129
+ "model.layers.2.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
130
+ "model.layers.2.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
131
+ "model.layers.2.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
132
+ "model.layers.2.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
133
+ "model.layers.2.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
134
+ "model.layers.20.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
135
+ "model.layers.20.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
136
+ "model.layers.20.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
137
+ "model.layers.20.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
138
+ "model.layers.20.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
139
+ "model.layers.20.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
140
+ "model.layers.20.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
141
+ "model.layers.20.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
142
+ "model.layers.20.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
143
+ "model.layers.21.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
144
+ "model.layers.21.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
145
+ "model.layers.21.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
146
+ "model.layers.21.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
147
+ "model.layers.21.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
148
+ "model.layers.21.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
149
+ "model.layers.21.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
150
+ "model.layers.21.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
151
+ "model.layers.21.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
152
+ "model.layers.22.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
153
+ "model.layers.22.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
154
+ "model.layers.22.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
155
+ "model.layers.22.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
156
+ "model.layers.22.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
157
+ "model.layers.22.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
158
+ "model.layers.22.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
159
+ "model.layers.22.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
160
+ "model.layers.22.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
161
+ "model.layers.23.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
162
+ "model.layers.23.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
163
+ "model.layers.23.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
164
+ "model.layers.23.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
165
+ "model.layers.23.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
166
+ "model.layers.23.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
167
+ "model.layers.23.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
168
+ "model.layers.23.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
169
+ "model.layers.23.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
170
+ "model.layers.24.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
171
+ "model.layers.24.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
172
+ "model.layers.24.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
173
+ "model.layers.24.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
174
+ "model.layers.24.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
175
+ "model.layers.24.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
176
+ "model.layers.24.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
177
+ "model.layers.24.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
178
+ "model.layers.24.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
179
+ "model.layers.25.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
180
+ "model.layers.25.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
181
+ "model.layers.25.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
182
+ "model.layers.25.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
183
+ "model.layers.25.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
184
+ "model.layers.25.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
185
+ "model.layers.25.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
186
+ "model.layers.25.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
187
+ "model.layers.25.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
188
+ "model.layers.26.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
189
+ "model.layers.26.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
190
+ "model.layers.26.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
191
+ "model.layers.26.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
192
+ "model.layers.26.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
193
+ "model.layers.26.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
194
+ "model.layers.26.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
195
+ "model.layers.26.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
196
+ "model.layers.26.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
197
+ "model.layers.27.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
198
+ "model.layers.27.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
199
+ "model.layers.27.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
200
+ "model.layers.27.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
201
+ "model.layers.27.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
202
+ "model.layers.27.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
203
+ "model.layers.27.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
204
+ "model.layers.27.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
205
+ "model.layers.27.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
206
+ "model.layers.28.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
207
+ "model.layers.28.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
208
+ "model.layers.28.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
209
+ "model.layers.28.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
210
+ "model.layers.28.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
211
+ "model.layers.28.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
212
+ "model.layers.28.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
213
+ "model.layers.28.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
214
+ "model.layers.28.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
215
+ "model.layers.29.input_layernorm.weight": "pytorch_model-00002-of-00003.bin",
216
+ "model.layers.29.mlp.down_proj.weight": "pytorch_model-00002-of-00003.bin",
217
+ "model.layers.29.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
218
+ "model.layers.29.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
219
+ "model.layers.29.post_attention_layernorm.weight": "pytorch_model-00002-of-00003.bin",
220
+ "model.layers.29.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
221
+ "model.layers.29.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
222
+ "model.layers.29.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
223
+ "model.layers.29.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
224
+ "model.layers.3.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
225
+ "model.layers.3.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
226
+ "model.layers.3.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
227
+ "model.layers.3.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
228
+ "model.layers.3.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
229
+ "model.layers.3.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
230
+ "model.layers.3.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
231
+ "model.layers.3.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
232
+ "model.layers.3.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
233
+ "model.layers.30.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
234
+ "model.layers.30.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
235
+ "model.layers.30.mlp.gate_proj.weight": "pytorch_model-00002-of-00003.bin",
236
+ "model.layers.30.mlp.up_proj.weight": "pytorch_model-00002-of-00003.bin",
237
+ "model.layers.30.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
238
+ "model.layers.30.self_attn.k_proj.weight": "pytorch_model-00002-of-00003.bin",
239
+ "model.layers.30.self_attn.o_proj.weight": "pytorch_model-00002-of-00003.bin",
240
+ "model.layers.30.self_attn.q_proj.weight": "pytorch_model-00002-of-00003.bin",
241
+ "model.layers.30.self_attn.v_proj.weight": "pytorch_model-00002-of-00003.bin",
242
+ "model.layers.31.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
243
+ "model.layers.31.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
244
+ "model.layers.31.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
245
+ "model.layers.31.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
246
+ "model.layers.31.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
247
+ "model.layers.31.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
248
+ "model.layers.31.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
249
+ "model.layers.31.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
250
+ "model.layers.31.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
251
+ "model.layers.32.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
252
+ "model.layers.32.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
253
+ "model.layers.32.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
254
+ "model.layers.32.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
255
+ "model.layers.32.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
256
+ "model.layers.32.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
257
+ "model.layers.32.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
258
+ "model.layers.32.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
259
+ "model.layers.32.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
260
+ "model.layers.33.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
261
+ "model.layers.33.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
262
+ "model.layers.33.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
263
+ "model.layers.33.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
264
+ "model.layers.33.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
265
+ "model.layers.33.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
266
+ "model.layers.33.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
267
+ "model.layers.33.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
268
+ "model.layers.33.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
269
+ "model.layers.34.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
270
+ "model.layers.34.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
271
+ "model.layers.34.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
272
+ "model.layers.34.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
273
+ "model.layers.34.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
274
+ "model.layers.34.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
275
+ "model.layers.34.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
276
+ "model.layers.34.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
277
+ "model.layers.34.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
278
+ "model.layers.35.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
279
+ "model.layers.35.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
280
+ "model.layers.35.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
281
+ "model.layers.35.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
282
+ "model.layers.35.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
283
+ "model.layers.35.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
284
+ "model.layers.35.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
285
+ "model.layers.35.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
286
+ "model.layers.35.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
287
+ "model.layers.36.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
288
+ "model.layers.36.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
289
+ "model.layers.36.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
290
+ "model.layers.36.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
291
+ "model.layers.36.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
292
+ "model.layers.36.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
293
+ "model.layers.36.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
294
+ "model.layers.36.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
295
+ "model.layers.36.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
296
+ "model.layers.37.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
297
+ "model.layers.37.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
298
+ "model.layers.37.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
299
+ "model.layers.37.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
300
+ "model.layers.37.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
301
+ "model.layers.37.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
302
+ "model.layers.37.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
303
+ "model.layers.37.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
304
+ "model.layers.37.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
305
+ "model.layers.38.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
306
+ "model.layers.38.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
307
+ "model.layers.38.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
308
+ "model.layers.38.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
309
+ "model.layers.38.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
310
+ "model.layers.38.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
311
+ "model.layers.38.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
312
+ "model.layers.38.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
313
+ "model.layers.38.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
314
+ "model.layers.39.input_layernorm.weight": "pytorch_model-00003-of-00003.bin",
315
+ "model.layers.39.mlp.down_proj.weight": "pytorch_model-00003-of-00003.bin",
316
+ "model.layers.39.mlp.gate_proj.weight": "pytorch_model-00003-of-00003.bin",
317
+ "model.layers.39.mlp.up_proj.weight": "pytorch_model-00003-of-00003.bin",
318
+ "model.layers.39.post_attention_layernorm.weight": "pytorch_model-00003-of-00003.bin",
319
+ "model.layers.39.self_attn.k_proj.weight": "pytorch_model-00003-of-00003.bin",
320
+ "model.layers.39.self_attn.o_proj.weight": "pytorch_model-00003-of-00003.bin",
321
+ "model.layers.39.self_attn.q_proj.weight": "pytorch_model-00003-of-00003.bin",
322
+ "model.layers.39.self_attn.v_proj.weight": "pytorch_model-00003-of-00003.bin",
323
+ "model.layers.4.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
324
+ "model.layers.4.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
325
+ "model.layers.4.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
326
+ "model.layers.4.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
327
+ "model.layers.4.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
328
+ "model.layers.4.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
329
+ "model.layers.4.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
330
+ "model.layers.4.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
331
+ "model.layers.4.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
332
+ "model.layers.5.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
333
+ "model.layers.5.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
334
+ "model.layers.5.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
335
+ "model.layers.5.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
336
+ "model.layers.5.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
337
+ "model.layers.5.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
338
+ "model.layers.5.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
339
+ "model.layers.5.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
340
+ "model.layers.5.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
341
+ "model.layers.6.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
342
+ "model.layers.6.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
343
+ "model.layers.6.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
344
+ "model.layers.6.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
345
+ "model.layers.6.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
346
+ "model.layers.6.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
347
+ "model.layers.6.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
348
+ "model.layers.6.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
349
+ "model.layers.6.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
350
+ "model.layers.7.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
351
+ "model.layers.7.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
352
+ "model.layers.7.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
353
+ "model.layers.7.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
354
+ "model.layers.7.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
355
+ "model.layers.7.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
356
+ "model.layers.7.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
357
+ "model.layers.7.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
358
+ "model.layers.7.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
359
+ "model.layers.8.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
360
+ "model.layers.8.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
361
+ "model.layers.8.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
362
+ "model.layers.8.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
363
+ "model.layers.8.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
364
+ "model.layers.8.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
365
+ "model.layers.8.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
366
+ "model.layers.8.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
367
+ "model.layers.8.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
368
+ "model.layers.9.input_layernorm.weight": "pytorch_model-00001-of-00003.bin",
369
+ "model.layers.9.mlp.down_proj.weight": "pytorch_model-00001-of-00003.bin",
370
+ "model.layers.9.mlp.gate_proj.weight": "pytorch_model-00001-of-00003.bin",
371
+ "model.layers.9.mlp.up_proj.weight": "pytorch_model-00001-of-00003.bin",
372
+ "model.layers.9.post_attention_layernorm.weight": "pytorch_model-00001-of-00003.bin",
373
+ "model.layers.9.self_attn.k_proj.weight": "pytorch_model-00001-of-00003.bin",
374
+ "model.layers.9.self_attn.o_proj.weight": "pytorch_model-00001-of-00003.bin",
375
+ "model.layers.9.self_attn.q_proj.weight": "pytorch_model-00001-of-00003.bin",
376
+ "model.layers.9.self_attn.v_proj.weight": "pytorch_model-00001-of-00003.bin",
377
+ "model.norm.weight": "pytorch_model-00003-of-00003.bin"
378
+ }
379
+ }
base/special_tokens_map.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": {
17
+ "content": "<unk>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "unk_token": {
24
+ "content": "<unk>",
25
+ "lstrip": false,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ }
30
+ }
base/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
base/tokenizer_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "added_tokens_decoder": {
5
+ "0": {
6
+ "content": "<unk>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "1": {
14
+ "content": "<s>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "2": {
22
+ "content": "</s>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ }
29
+ },
30
+ "bos_token": "<s>",
31
+ "clean_up_tokenization_spaces": false,
32
+ "eos_token": "</s>",
33
+ "legacy": false,
34
+ "model_max_length": 4096,
35
+ "pad_token": "<unk>",
36
+ "padding_side": "right",
37
+ "sp_model_kwargs": {},
38
+ "spaces_between_special_tokens": false,
39
+ "tokenizer_class": "LlamaTokenizer",
40
+ "trust_remote_code": false,
41
+ "unk_token": "<unk>",
42
+ "use_default_system_prompt": true,
43
+ "use_fast": true
44
+ }
head/config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "medusa_num_heads": 3,
3
+ "medusa_num_layers": 1,
4
+ "transformers_version": "4.34.1",
5
+ "base_model_name_or_path": "vicuna-13b-v1.5",
6
+ "vocab_size": 32000
7
+ }
head/medusa_lm_head.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32dedf425ba6a4e71a35bbfacb536a03fe70c66f5cfb18fcbde83dd4262f454c
3
+ size 2280717655
medusa_model_demo.py ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ from medusa.model.medusa_model_legacy import MedusaConfig
2
+ from medusa.model.medusa_model_new import MedusaModel
3
+ import torch
4
+ import os
5
+ import json
6
+ import argparse
7
+
8
+ def read_json(file_path):
9
+ with open(file_path, 'r') as file:
10
+ data = json.load(file)
11
+ return data
12
+
13
+ def write_json(data, file_path):
14
+ with open(file_path, 'w') as file:
15
+ json.dump(data, file, indent=4)
16
+
17
+
18
+ os.environ["CUDA_VISIBLE_DEVICES"] = "7"
19
+ def main():
20
+ parser = argparse.ArgumentParser(description='Example extract medusa head')
21
+ parser.add_argument('-m', '--model-path', type=str, required=True, help='Path to new medusa models')
22
+ parser.add_argument('-o', '--output-dir', type=str, required=True, help='Path to save medusa head')
23
+ parser.add_argument('--use-full-key', action='store_true', help='medusa head keys with medusa_head. or not')
24
+ parser.add_argument('--verbose', action='store_true', help='Enable verbose output')
25
+ args = parser.parse_args()
26
+ config = read_json(os.path.join(args.model_path, "config.json"))
27
+ new_config = {}
28
+ for k, v in config.items():
29
+ if k.startswith("medusa"):
30
+ new_config[k] = v
31
+ new_config["transformers_version"] = config["transformers_version"]
32
+ new_config["base_model_name_or_path"] = config["_name_or_path"]
33
+
34
+ model = MedusaModel.from_pretrained(pretrained_model_name_or_path=args.model_path)
35
+
36
+ if args.verbose:
37
+ print(model)
38
+ medusa_lm_head = dict()
39
+ for k, v in model.state_dict().items():
40
+ if k.startswith("medusa"):
41
+ if args.verbose: print(k)
42
+ if args.use_full_key:
43
+ medusa_lm_head[k] = v
44
+ else:
45
+ medusa_lm_head[k[len("medusa_head."):]] = v
46
+ if not os.path.exists(args.output_dir):
47
+ os.makedirs(args.output_dir)
48
+ write_json(new_config, f"{args.output_dir}/config.json")
49
+ model_save_path = f"{args.output_dir}/medusa_lm_head.pt"
50
+ torch.save(medusa_lm_head, model_save_path)
51
+
52
+ print(f"Model saved to {model_save_path}")
53
+
54
+
55
+ if __name__ == "__main__":
56
+ main()