Hugging Face's logo

Litzy619
/
V0413TUNE

Safetensors
Generated from Trainer
Model card Files
xet
 Community
Litzy619 commited on
Commit
4be7d67
·
verified ·
1 Parent(s): ca86742

End of training

Browse files
Files changed (5) hide show
  1. README.md +27 -27
  2. adapter_config.json +1 -1
  3. adapter_model.bin +2 -2
  4. model.safetensors +2 -2
  5. training_args.bin +1 -1
README.md CHANGED
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
15
 
16
  This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
- - Loss: 0.1402
19
 
20
  ## Model description
21
 
@@ -38,40 +38,40 @@ The following hyperparameters were used during training:
38
  - train_batch_size: 8
39
  - eval_batch_size: 8
40
  - seed: 42
41
- - gradient_accumulation_steps: 16
42
- - total_train_batch_size: 128
43
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
  - lr_scheduler_type: cosine_with_restarts
45
- - lr_scheduler_warmup_steps: 60
46
- - num_epochs: 4
47
  - mixed_precision_training: Native AMP
48
 
49
  ### Training results
50
 
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:-----:|:----:|:---------------:|
53
- | 0.3053 | 0.18 | 20 | 1.2898 |
54
- | 2.4718 | 0.36 | 40 | 4.1484 |
55
- | 2.4944 | 0.54 | 60 | 0.4804 |
56
- | 0.3301 | 0.73 | 80 | 0.3197 |
57
- | 0.2944 | 0.91 | 100 | 0.2742 |
58
- | 0.2574 | 1.09 | 120 | 0.2777 |
59
- | 0.272 | 1.27 | 140 | 0.2909 |
60
- | 0.2601 | 1.45 | 160 | 0.2258 |
61
- | 0.1581 | 1.63 | 180 | 0.1577 |
62
- | 0.1471 | 1.81 | 200 | 0.1484 |
63
- | 0.1567 | 1.99 | 220 | 0.1509 |
64
- | 0.1529 | 2.18 | 240 | 0.1481 |
65
- | 0.1425 | 2.36 | 260 | 0.1444 |
66
- | 0.147 | 2.54 | 280 | 0.1421 |
67
- | 0.142 | 2.72 | 300 | 0.1435 |
68
- | 0.1387 | 2.9 | 320 | 0.1439 |
69
- | 0.1387 | 3.08 | 340 | 0.1426 |
70
- | 0.1357 | 3.26 | 360 | 0.1407 |
71
- | 0.1448 | 3.44 | 380 | 0.1405 |
72
- | 0.1435 | 3.63 | 400 | 0.1415 |
73
- | 0.1401 | 3.81 | 420 | 0.1408 |
74
- | 0.1459 | 3.99 | 440 | 0.1402 |
75
 
76
 
77
  ### Framework versions
 
15
 
16
  This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 0.0656
19
 
20
  ## Model description
21
 
 
38
  - train_batch_size: 8
39
  - eval_batch_size: 8
40
  - seed: 42
41
+ - gradient_accumulation_steps: 8
42
+ - total_train_batch_size: 64
43
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
44
  - lr_scheduler_type: cosine_with_restarts
45
+ - lr_scheduler_warmup_steps: 100
46
+ - num_epochs: 2
47
  - mixed_precision_training: Native AMP
48
 
49
  ### Training results
50
 
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:-----:|:----:|:---------------:|
53
+ | 0.1876 | 0.09 | 20 | 0.2016 |
54
+ | 4.5112 | 0.18 | 40 | 1.8999 |
55
+ | 0.7446 | 0.27 | 60 | 0.4058 |
56
+ | 0.3379 | 0.36 | 80 | 0.3014 |
57
+ | 0.2506 | 0.45 | 100 | 0.2398 |
58
+ | 0.1919 | 0.54 | 120 | 0.2093 |
59
+ | 0.194 | 0.63 | 140 | 0.1571 |
60
+ | 0.1427 | 0.73 | 160 | 0.1332 |
61
+ | 0.1141 | 0.82 | 180 | 0.1500 |
62
+ | 0.1198 | 0.91 | 200 | 0.1212 |
63
+ | 0.1003 | 1.0 | 220 | 0.0844 |
64
+ | 0.1177 | 1.09 | 240 | 0.0906 |
65
+ | 0.1133 | 1.18 | 260 | 0.1179 |
66
+ | 0.0873 | 1.27 | 280 | 0.0808 |
67
+ | 0.0764 | 1.36 | 300 | 0.0756 |
68
+ | 0.0855 | 1.45 | 320 | 0.0729 |
69
+ | 0.0809 | 1.54 | 340 | 0.0699 |
70
+ | 0.0757 | 1.63 | 360 | 0.0709 |
71
+ | 0.0744 | 1.72 | 380 | 0.0684 |
72
+ | 0.0624 | 1.81 | 400 | 0.0676 |
73
+ | 0.0678 | 1.9 | 420 | 0.0654 |
74
+ | 0.0682 | 1.99 | 440 | 0.0656 |
75
 
76
 
77
  ### Framework versions
adapter_config.json CHANGED
@@ -15,7 +15,7 @@
15
  "ot_diversified_prefix": false,
16
  "peft_type": "M_PREFIX",
17
  "prefix_projection": true,
18
- "scale": 64,
19
  "task_type": "CAUSAL_LM",
20
  "token_dim": 2560
21
  }
 
15
  "ot_diversified_prefix": false,
16
  "peft_type": "M_PREFIX",
17
  "prefix_projection": true,
18
+ "scale": 128,
19
  "task_type": "CAUSAL_LM",
20
  "token_dim": 2560
21
  }
adapter_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:0f622066511d49496d55cd44114e2d4c4e621413d9848c5cdb7e1ad14956e7e7
3
- size 84237364
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e964b589fb9768f9395cd503482313c7c959208eed8daa6bbfb3e268f4dc7fa4
3
+ size 77888436
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:13561e8a88b5c20619d1f5cb50699c6b9d830590179796558824fffcc02f350a
3
- size 5643644816
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a67631b3875667ef6a46e449e5e8e930f250c78a903a079561c1215cf09cb5d4
3
+ size 5637295888
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:52e54f1b8b4f34adc09340f145f28dcbf5e1cdb1f4714d55c9e88c421bfe2f79
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6eafb9a2c9e081851d086e9ae1c68b7fc126a2edbc74268710b201c4930bef8
3
  size 5112