Hugging Face's logo

Gunslinger3D
/
fine-tuning-Phi2-with-webglm-qa-with-lora

PEFT
Safetensors
Generated from Trainer
Model card Files Community
Gunslinger3D commited on
Commit
0ba7fe3
·
verified ·
1 Parent(s): 911c10c

fine-tuning-Phi2-with-webglm-qa-with-lora

Browse files
Files changed (3) hide show
  1. README.md +51 -51
  2. adapter_config.json +3 -3
  3. adapter_model.safetensors +1 -1
README.md CHANGED
@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
16
 
17
  This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.1016
20
 
21
  ## Model description
22
 
@@ -51,56 +51,56 @@ The following hyperparameters were used during training:
51
 
52
  | Training Loss | Epoch | Step | Validation Loss |
53
  |:-------------:|:-----:|:----:|:---------------:|
54
- | No log | 0.2 | 10 | 8.0315 |
55
- | No log | 0.4 | 20 | 6.2055 |
56
- | No log | 0.6 | 30 | 2.7735 |
57
- | No log | 0.8 | 40 | 0.6055 |
58
- | 4.4745 | 1.0 | 50 | 0.5323 |
59
- | 4.4745 | 1.2 | 60 | 0.4631 |
60
- | 4.4745 | 1.39 | 70 | 0.4075 |
61
- | 4.4745 | 1.59 | 80 | 0.3566 |
62
- | 4.4745 | 1.79 | 90 | 0.3155 |
63
- | 0.3331 | 1.99 | 100 | 0.2869 |
64
- | 0.3331 | 2.19 | 110 | 0.2624 |
65
- | 0.3331 | 2.39 | 120 | 0.2453 |
66
- | 0.3331 | 2.59 | 130 | 0.2288 |
67
- | 0.3331 | 2.79 | 140 | 0.2095 |
68
- | 0.1947 | 2.99 | 150 | 0.1978 |
69
- | 0.1947 | 3.19 | 160 | 0.1886 |
70
- | 0.1947 | 3.39 | 170 | 0.1766 |
71
- | 0.1947 | 3.59 | 180 | 0.1691 |
72
- | 0.1947 | 3.78 | 190 | 0.1626 |
73
- | 0.1486 | 3.98 | 200 | 0.1562 |
74
- | 0.1486 | 4.18 | 210 | 0.1510 |
75
- | 0.1486 | 4.38 | 220 | 0.1489 |
76
- | 0.1486 | 4.58 | 230 | 0.1439 |
77
- | 0.1486 | 4.78 | 240 | 0.1364 |
78
- | 0.1232 | 4.98 | 250 | 0.1314 |
79
- | 0.1232 | 5.18 | 260 | 0.1306 |
80
- | 0.1232 | 5.38 | 270 | 0.1295 |
81
- | 0.1232 | 5.58 | 280 | 0.1256 |
82
- | 0.1232 | 5.78 | 290 | 0.1228 |
83
- | 0.1084 | 5.98 | 300 | 0.1195 |
84
- | 0.1084 | 6.18 | 310 | 0.1165 |
85
- | 0.1084 | 6.37 | 320 | 0.1156 |
86
- | 0.1084 | 6.57 | 330 | 0.1147 |
87
- | 0.1084 | 6.77 | 340 | 0.1120 |
88
- | 0.0964 | 6.97 | 350 | 0.1100 |
89
- | 0.0964 | 7.17 | 360 | 0.1100 |
90
- | 0.0964 | 7.37 | 370 | 0.1087 |
91
- | 0.0964 | 7.57 | 380 | 0.1080 |
92
- | 0.0964 | 7.77 | 390 | 0.1071 |
93
- | 0.0905 | 7.97 | 400 | 0.1065 |
94
- | 0.0905 | 8.17 | 410 | 0.1061 |
95
- | 0.0905 | 8.37 | 420 | 0.1053 |
96
- | 0.0905 | 8.57 | 430 | 0.1044 |
97
- | 0.0905 | 8.76 | 440 | 0.1036 |
98
- | 0.0843 | 8.96 | 450 | 0.1028 |
99
- | 0.0843 | 9.16 | 460 | 0.1021 |
100
- | 0.0843 | 9.36 | 470 | 0.1019 |
101
- | 0.0843 | 9.56 | 480 | 0.1018 |
102
- | 0.0843 | 9.76 | 490 | 0.1016 |
103
- | 0.0819 | 9.96 | 500 | 0.1016 |
104
 
105
 
106
  ### Framework versions
 
16
 
17
  This model is a fine-tuned version of [microsoft/phi-2](https://huggingface.co/microsoft/phi-2) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.1032
20
 
21
  ## Model description
22
 
 
51
 
52
  | Training Loss | Epoch | Step | Validation Loss |
53
  |:-------------:|:-----:|:----:|:---------------:|
54
+ | No log | 0.2 | 10 | 8.0361 |
55
+ | No log | 0.4 | 20 | 6.2064 |
56
+ | No log | 0.6 | 30 | 2.7739 |
57
+ | No log | 0.8 | 40 | 0.6071 |
58
+ | 4.4774 | 1.0 | 50 | 0.5329 |
59
+ | 4.4774 | 1.2 | 60 | 0.4635 |
60
+ | 4.4774 | 1.39 | 70 | 0.4081 |
61
+ | 4.4774 | 1.59 | 80 | 0.3576 |
62
+ | 4.4774 | 1.79 | 90 | 0.3173 |
63
+ | 0.3338 | 1.99 | 100 | 0.2889 |
64
+ | 0.3338 | 2.19 | 110 | 0.2645 |
65
+ | 0.3338 | 2.39 | 120 | 0.2471 |
66
+ | 0.3338 | 2.59 | 130 | 0.2301 |
67
+ | 0.3338 | 2.79 | 140 | 0.2121 |
68
+ | 0.1964 | 2.99 | 150 | 0.1992 |
69
+ | 0.1964 | 3.19 | 160 | 0.1913 |
70
+ | 0.1964 | 3.39 | 170 | 0.1793 |
71
+ | 0.1964 | 3.59 | 180 | 0.1713 |
72
+ | 0.1964 | 3.78 | 190 | 0.1642 |
73
+ | 0.1501 | 3.98 | 200 | 0.1579 |
74
+ | 0.1501 | 4.18 | 210 | 0.1531 |
75
+ | 0.1501 | 4.38 | 220 | 0.1511 |
76
+ | 0.1501 | 4.58 | 230 | 0.1455 |
77
+ | 0.1501 | 4.78 | 240 | 0.1379 |
78
+ | 0.1248 | 4.98 | 250 | 0.1333 |
79
+ | 0.1248 | 5.18 | 260 | 0.1313 |
80
+ | 0.1248 | 5.38 | 270 | 0.1308 |
81
+ | 0.1248 | 5.58 | 280 | 0.1271 |
82
+ | 0.1248 | 5.78 | 290 | 0.1244 |
83
+ | 0.1097 | 5.98 | 300 | 0.1208 |
84
+ | 0.1097 | 6.18 | 310 | 0.1178 |
85
+ | 0.1097 | 6.37 | 320 | 0.1164 |
86
+ | 0.1097 | 6.57 | 330 | 0.1155 |
87
+ | 0.1097 | 6.77 | 340 | 0.1125 |
88
+ | 0.0976 | 6.97 | 350 | 0.1108 |
89
+ | 0.0976 | 7.17 | 360 | 0.1109 |
90
+ | 0.0976 | 7.37 | 370 | 0.1093 |
91
+ | 0.0976 | 7.57 | 380 | 0.1085 |
92
+ | 0.0976 | 7.77 | 390 | 0.1079 |
93
+ | 0.0917 | 7.97 | 400 | 0.1072 |
94
+ | 0.0917 | 8.17 | 410 | 0.1064 |
95
+ | 0.0917 | 8.37 | 420 | 0.1058 |
96
+ | 0.0917 | 8.57 | 430 | 0.1054 |
97
+ | 0.0917 | 8.76 | 440 | 0.1047 |
98
+ | 0.0855 | 8.96 | 450 | 0.1040 |
99
+ | 0.0855 | 9.16 | 460 | 0.1034 |
100
+ | 0.0855 | 9.36 | 470 | 0.1032 |
101
+ | 0.0855 | 9.56 | 480 | 0.1032 |
102
+ | 0.0855 | 9.76 | 490 | 0.1032 |
103
+ | 0.0833 | 9.96 | 500 | 0.1032 |
104
 
105
 
106
  ### Framework versions
adapter_config.json CHANGED
@@ -19,12 +19,12 @@
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
22
- "q_proj",
23
  "k_proj",
24
  "v_proj",
25
- "fc2",
26
  "dense",
27
- "fc1"
 
 
28
  ],
29
  "task_type": "CAUSAL_LM"
30
  }
 
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
 
22
  "k_proj",
23
  "v_proj",
 
24
  "dense",
25
+ "fc1",
26
+ "q_proj",
27
+ "fc2"
28
  ],
29
  "task_type": "CAUSAL_LM"
30
  }
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eb85b33189faa8d8a482db2171584cc114b99d6eab73450f77c75ee1eff60b10
3
  size 94422368
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2940117eb1f424c3485633ca11ec57876aa73e4dcf297c5726a207484ed4272
3
  size 94422368