Hugging Face's logo

frankmorales2020
/
POC-Meta-Llama-3-8B-MEDAL-flash-attention-2-cosine-evaldata

PEFT
TensorBoard
Safetensors
trl
sft
Generated from Trainer
Model card Files Metrics Community
frankmorales2020 commited on
Commit
f27305a
·
verified ·
1 Parent(s): 4950bfa

Model save

Browse files
Files changed (3) hide show
  1. README.md +108 -0
  2. adapter_model.safetensors +1 -1
  3. logs/events.out.tfevents.1719114650.6ada53b7557c.59036.0 +2 -2
README.md ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model: meta-llama/Meta-Llama-3-8B
3
+ datasets:
4
+ - generator
5
+ library_name: peft
6
+ license: llama3
7
+ tags:
8
+ - trl
9
+ - sft
10
+ - generated_from_trainer
11
+ model-index:
12
+ - name: POC-Meta-Llama-3-8B-MEDAL-flash-attention-2-cosine-evaldata
13
+ results: []
14
+ ---
15
+
16
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
+ should probably proofread and complete it, then remove this comment. -->
18
+
19
+ # POC-Meta-Llama-3-8B-MEDAL-flash-attention-2-cosine-evaldata
20
+
21
+ This model is a fine-tuned version of [meta-llama/Meta-Llama-3-8B](https://huggingface.co/meta-llama/Meta-Llama-3-8B) on the generator dataset.
22
+ It achieves the following results on the evaluation set:
23
+ - Loss: 2.2897
24
+
25
+ ## Model description
26
+
27
+ More information needed
28
+
29
+ ## Intended uses & limitations
30
+
31
+ More information needed
32
+
33
+ ## Training and evaluation data
34
+
35
+ More information needed
36
+
37
+ ## Training procedure
38
+
39
+ ### Training hyperparameters
40
+
41
+ The following hyperparameters were used during training:
42
+ - learning_rate: 5e-05
43
+ - train_batch_size: 1
44
+ - eval_batch_size: 8
45
+ - seed: 42
46
+ - gradient_accumulation_steps: 8
47
+ - total_train_batch_size: 8
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_ratio: 0.15
51
+ - num_epochs: 0.3
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:------:|:----:|:---------------:|
57
+ | 2.7645 | 0.0069 | 100 | 2.6720 |
58
+ | 2.5917 | 0.0138 | 200 | 2.5243 |
59
+ | 2.5054 | 0.0207 | 300 | 2.4705 |
60
+ | 2.4406 | 0.0277 | 400 | 2.4379 |
61
+ | 2.4272 | 0.0346 | 500 | 2.4136 |
62
+ | 2.4171 | 0.0415 | 600 | 2.3942 |
63
+ | 2.3908 | 0.0484 | 700 | 2.3793 |
64
+ | 2.3808 | 0.0553 | 800 | 2.3664 |
65
+ | 2.3588 | 0.0622 | 900 | 2.3571 |
66
+ | 2.3595 | 0.0692 | 1000 | 2.3494 |
67
+ | 2.3411 | 0.0761 | 1100 | 2.3421 |
68
+ | 2.3308 | 0.0830 | 1200 | 2.3369 |
69
+ | 2.3358 | 0.0899 | 1300 | 2.3320 |
70
+ | 2.3295 | 0.0968 | 1400 | 2.3270 |
71
+ | 2.337 | 0.1037 | 1500 | 2.3228 |
72
+ | 2.3182 | 0.1106 | 1600 | 2.3195 |
73
+ | 2.3334 | 0.1176 | 1700 | 2.3161 |
74
+ | 2.3278 | 0.1245 | 1800 | 2.3128 |
75
+ | 2.3151 | 0.1314 | 1900 | 2.3101 |
76
+ | 2.3245 | 0.1383 | 2000 | 2.3075 |
77
+ | 2.3073 | 0.1452 | 2100 | 2.3053 |
78
+ | 2.3094 | 0.1521 | 2200 | 2.3036 |
79
+ | 2.3101 | 0.1590 | 2300 | 2.3013 |
80
+ | 2.3102 | 0.1660 | 2400 | 2.2995 |
81
+ | 2.3042 | 0.1729 | 2500 | 2.2980 |
82
+ | 2.2942 | 0.1798 | 2600 | 2.2965 |
83
+ | 2.2876 | 0.1867 | 2700 | 2.2951 |
84
+ | 2.3077 | 0.1936 | 2800 | 2.2941 |
85
+ | 2.2851 | 0.2005 | 2900 | 2.2931 |
86
+ | 2.2766 | 0.2075 | 3000 | 2.2923 |
87
+ | 2.2873 | 0.2144 | 3100 | 2.2916 |
88
+ | 2.2971 | 0.2213 | 3200 | 2.2910 |
89
+ | 2.2942 | 0.2282 | 3300 | 2.2906 |
90
+ | 2.2872 | 0.2351 | 3400 | 2.2903 |
91
+ | 2.2996 | 0.2420 | 3500 | 2.2901 |
92
+ | 2.2855 | 0.2489 | 3600 | 2.2899 |
93
+ | 2.2969 | 0.2559 | 3700 | 2.2898 |
94
+ | 2.2871 | 0.2628 | 3800 | 2.2898 |
95
+ | 2.2905 | 0.2697 | 3900 | 2.2897 |
96
+ | 2.2915 | 0.2766 | 4000 | 2.2897 |
97
+ | 2.2921 | 0.2835 | 4100 | 2.2897 |
98
+ | 2.3087 | 0.2904 | 4200 | 2.2897 |
99
+ | 2.3017 | 0.2974 | 4300 | 2.2897 |
100
+
101
+
102
+ ### Framework versions
103
+
104
+ - PEFT 0.11.1
105
+ - Transformers 4.41.2
106
+ - Pytorch 2.3.0+cu121
107
+ - Datasets 2.20.0
108
+ - Tokenizers 0.19.1
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:bfab1070e46c8865b35133c422d1f8e29f94855b738e9f8f77389e6e7d6f491c
3
  size 671150064
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41feff0a94b1ebddc10ca22a68e4ad4dff9299b5f7a48e4e94ca5c4dcf464aed
3
  size 671150064
logs/events.out.tfevents.1719114650.6ada53b7557c.59036.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:346c69e345b82ac46c3ddf383242b73199c190a5dc3916673bc5439be1100365
3
- size 26308
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca654f75c311728fe388403456fa8a5ab7477e5d01f98978e9cc454b13227e6a
3
+ size 26662