lsmille commited on
Commit
32c517d
1 Parent(s): f391cc0

lsmille/lora_evo_ta_all_layers_16

Browse files
README.md ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ library_name: peft
4
+ tags:
5
+ - generated_from_trainer
6
+ base_model: togethercomputer/evo-1-8k-base
7
+ model-index:
8
+ - name: lora_evo_ta_all_layers_16
9
+ results: []
10
+ ---
11
+
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # lora_evo_ta_all_layers_16
16
+
17
+ This model is a fine-tuned version of [togethercomputer/evo-1-8k-base](https://huggingface.co/togethercomputer/evo-1-8k-base) on the None dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 2.5463
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ ### Training hyperparameters
36
+
37
+ The following hyperparameters were used during training:
38
+ - learning_rate: 0.0003
39
+ - train_batch_size: 2
40
+ - eval_batch_size: 2
41
+ - seed: 42
42
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
+ - lr_scheduler_type: constant
44
+ - lr_scheduler_warmup_steps: 500
45
+ - num_epochs: 3
46
+
47
+ ### Training results
48
+
49
+ | Training Loss | Epoch | Step | Validation Loss |
50
+ |:-------------:|:------:|:-----:|:---------------:|
51
+ | 2.8598 | 0.4998 | 1999 | 2.6289 |
52
+ | 2.5927 | 0.9995 | 3998 | 2.5852 |
53
+ | 2.5467 | 1.4992 | 5997 | 2.5717 |
54
+ | 2.5487 | 1.999 | 7996 | 2.5554 |
55
+ | 2.4987 | 2.4988 | 9995 | 2.5546 |
56
+ | 2.4934 | 2.9985 | 11994 | 2.5463 |
57
+
58
+
59
+ ### Framework versions
60
+
61
+ - PEFT 0.11.1
62
+ - Transformers 4.41.2
63
+ - Pytorch 2.3.0+cu121
64
+ - Datasets 2.19.1
65
+ - Tokenizers 0.19.1
adapter_config.json ADDED
@@ -0,0 +1,129 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": {
4
+ "base_model_class": "StripedHyenaModelForCausalLM",
5
+ "parent_library": "transformers_modules.togethercomputer.evo-1-131k-base.567369e9825aa08b3de4b122fca34fac6a890602.modeling_hyena"
6
+ },
7
+ "base_model_name_or_path": "togethercomputer/evo-1-8k-base",
8
+ "bias": "none",
9
+ "fan_in_fan_out": false,
10
+ "inference_mode": true,
11
+ "init_lora_weights": "gaussian",
12
+ "layer_replication": null,
13
+ "layers_pattern": null,
14
+ "layers_to_transform": null,
15
+ "loftq_config": {},
16
+ "lora_alpha": 128,
17
+ "lora_dropout": 0.1,
18
+ "megatron_config": null,
19
+ "megatron_core": "megatron.core",
20
+ "modules_to_save": null,
21
+ "peft_type": "LORA",
22
+ "r": 128,
23
+ "rank_pattern": {},
24
+ "revision": null,
25
+ "target_modules": [
26
+ "backbone.blocks.22.mlp.l1",
27
+ "backbone.blocks.12.mlp.l3",
28
+ "backbone.blocks.16.mlp.l3",
29
+ "backbone.blocks.16.inner_mha_cls.Wqkv",
30
+ "backbone.blocks.1.mlp.l2",
31
+ "backbone.blocks.4.mlp.l3",
32
+ "backbone.blocks.24.inner_mha_cls.out_proj",
33
+ "backbone.blocks.6.mlp.l1",
34
+ "backbone.blocks.7.mlp.l2",
35
+ "backbone.blocks.8.inner_mha_cls.out_proj",
36
+ "backbone.blocks.3.mlp.l3",
37
+ "backbone.blocks.31.mlp.l2",
38
+ "backbone.blocks.27.mlp.l2",
39
+ "backbone.blocks.15.mlp.l3",
40
+ "backbone.blocks.15.mlp.l1",
41
+ "backbone.blocks.24.mlp.l2",
42
+ "backbone.blocks.14.mlp.l1",
43
+ "backbone.blocks.11.mlp.l3",
44
+ "backbone.blocks.25.mlp.l2",
45
+ "backbone.blocks.31.mlp.l3",
46
+ "backbone.blocks.23.mlp.l1",
47
+ "backbone.blocks.19.mlp.l2",
48
+ "backbone.blocks.19.mlp.l3",
49
+ "backbone.blocks.18.mlp.l1",
50
+ "backbone.blocks.17.mlp.l3",
51
+ "backbone.blocks.2.mlp.l2",
52
+ "backbone.blocks.25.mlp.l1",
53
+ "backbone.blocks.22.mlp.l3",
54
+ "backbone.blocks.29.mlp.l2",
55
+ "backbone.blocks.8.inner_mha_cls.Wqkv",
56
+ "backbone.blocks.20.mlp.l1",
57
+ "backbone.blocks.27.mlp.l1",
58
+ "backbone.blocks.14.mlp.l3",
59
+ "backbone.blocks.30.mlp.l2",
60
+ "backbone.blocks.27.mlp.l3",
61
+ "backbone.blocks.24.inner_mha_cls.Wqkv",
62
+ "backbone.blocks.29.mlp.l1",
63
+ "backbone.blocks.22.mlp.l2",
64
+ "backbone.blocks.20.mlp.l3",
65
+ "backbone.blocks.6.mlp.l3",
66
+ "backbone.blocks.10.mlp.l3",
67
+ "backbone.blocks.5.mlp.l2",
68
+ "backbone.blocks.26.mlp.l1",
69
+ "backbone.blocks.17.mlp.l1",
70
+ "backbone.blocks.5.mlp.l3",
71
+ "backbone.blocks.21.mlp.l2",
72
+ "backbone.blocks.4.mlp.l1",
73
+ "backbone.blocks.28.mlp.l3",
74
+ "backbone.blocks.23.mlp.l3",
75
+ "backbone.blocks.9.mlp.l3",
76
+ "backbone.blocks.16.mlp.l1",
77
+ "backbone.blocks.13.mlp.l3",
78
+ "backbone.blocks.15.mlp.l2",
79
+ "backbone.blocks.20.mlp.l2",
80
+ "backbone.blocks.17.mlp.l2",
81
+ "backbone.blocks.1.mlp.l3",
82
+ "backbone.blocks.11.mlp.l1",
83
+ "backbone.blocks.18.mlp.l3",
84
+ "backbone.blocks.9.mlp.l2",
85
+ "backbone.blocks.21.mlp.l1",
86
+ "backbone.blocks.1.mlp.l1",
87
+ "backbone.blocks.6.mlp.l2",
88
+ "backbone.blocks.28.mlp.l1",
89
+ "backbone.blocks.13.mlp.l2",
90
+ "backbone.blocks.5.mlp.l1",
91
+ "backbone.blocks.14.mlp.l2",
92
+ "backbone.blocks.24.mlp.l3",
93
+ "backbone.blocks.12.mlp.l2",
94
+ "backbone.blocks.31.mlp.l1",
95
+ "backbone.blocks.19.mlp.l1",
96
+ "backbone.blocks.3.mlp.l1",
97
+ "backbone.blocks.16.inner_mha_cls.out_proj",
98
+ "backbone.blocks.4.mlp.l2",
99
+ "backbone.blocks.26.mlp.l2",
100
+ "backbone.blocks.26.mlp.l3",
101
+ "backbone.blocks.18.mlp.l2",
102
+ "backbone.blocks.30.mlp.l3",
103
+ "backbone.blocks.3.mlp.l2",
104
+ "backbone.blocks.21.mlp.l3",
105
+ "backbone.blocks.25.mlp.l3",
106
+ "backbone.blocks.7.mlp.l3",
107
+ "backbone.blocks.16.mlp.l2",
108
+ "backbone.blocks.10.mlp.l1",
109
+ "backbone.blocks.10.mlp.l2",
110
+ "backbone.blocks.28.mlp.l2",
111
+ "backbone.blocks.8.mlp.l2",
112
+ "backbone.blocks.8.mlp.l3",
113
+ "backbone.blocks.2.mlp.l3",
114
+ "backbone.blocks.11.mlp.l2",
115
+ "backbone.blocks.23.mlp.l2",
116
+ "backbone.blocks.30.mlp.l1",
117
+ "backbone.blocks.7.mlp.l1",
118
+ "backbone.blocks.29.mlp.l3",
119
+ "backbone.blocks.13.mlp.l1",
120
+ "backbone.blocks.2.mlp.l1",
121
+ "backbone.blocks.12.mlp.l1",
122
+ "backbone.blocks.9.mlp.l1",
123
+ "backbone.blocks.8.mlp.l1",
124
+ "backbone.blocks.24.mlp.l1"
125
+ ],
126
+ "task_type": null,
127
+ "use_dora": false,
128
+ "use_rslora": false
129
+ }
adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85cf93c6e8702c7f76a52ff26833d4705ce8a88271c22cf7a34665226e70ac4a
3
+ size 376592024
special_tokens_map.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "eos_token": {
3
+ "content": " ",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "pad_token": " "
10
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_commit_hash": "cb2b3eb17365b1c22deab9114aec7afb2746c9f1",
3
+ "_from_auto": true,
4
+ "added_tokens_decoder": {
5
+ "32": {
6
+ "content": " ",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ }
13
+ },
14
+ "byte_level": true,
15
+ "clean_up_tokenization_spaces": true,
16
+ "eos_token": " ",
17
+ "model_max_length": 1000000000000000019884624838656,
18
+ "pad_token": " ",
19
+ "tokenizer_class": "ByteTokenizer",
20
+ "trust_remote_code": true
21
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed691de93d31ca5af6881c236a89f884554dfd73c5b889de2856e28a7175d7f0
3
+ size 5048