augustocsc commited on
Commit
ef9a8b1
·
verified ·
1 Parent(s): b38065a

Model save

Browse files
Files changed (3) hide show
  1. README.md +191 -53
  2. all_results.json +6 -6
  3. train_results.json +6 -6
README.md CHANGED
@@ -1,22 +1,22 @@
1
  ---
2
  library_name: peft
3
  license: mit
4
- base_model: gpt2-medium
5
  tags:
6
  - generated_from_trainer
7
  model-index:
8
- - name: Se124M100KInfPrompt_WT_EOS_medium
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
- # Se124M100KInfPrompt_WT_EOS_medium
16
 
17
- This model is a fine-tuned version of [gpt2-medium](https://huggingface.co/gpt2-medium) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
- - Loss: 0.7127
20
 
21
  ## Model description
22
 
@@ -36,11 +36,11 @@ More information needed
36
 
37
  The following hyperparameters were used during training:
38
  - learning_rate: 5e-05
39
- - train_batch_size: 64
40
- - eval_batch_size: 64
41
  - seed: 42
42
  - gradient_accumulation_steps: 4
43
- - total_train_batch_size: 256
44
  - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
45
  - lr_scheduler_type: cosine
46
  - lr_scheduler_warmup_ratio: 0.03
@@ -50,51 +50,189 @@ The following hyperparameters were used during training:
50
 
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:------:|:----:|:---------------:|
53
- | 2.8652 | 0.0655 | 20 | 2.6742 |
54
- | 2.6735 | 0.1309 | 40 | 2.4205 |
55
- | 2.3498 | 0.1964 | 60 | 2.0554 |
56
- | 1.9542 | 0.2619 | 80 | 1.6239 |
57
- | 1.5661 | 0.3273 | 100 | 1.2791 |
58
- | 1.3052 | 0.3928 | 120 | 1.0776 |
59
- | 1.1291 | 0.4583 | 140 | 0.9537 |
60
- | 1.0151 | 0.5237 | 160 | 0.8837 |
61
- | 0.9431 | 0.5892 | 180 | 0.8324 |
62
- | 0.8821 | 0.6547 | 200 | 0.8044 |
63
- | 0.8536 | 0.7201 | 220 | 0.7846 |
64
- | 0.8371 | 0.7856 | 240 | 0.7712 |
65
- | 0.8281 | 0.8511 | 260 | 0.7628 |
66
- | 0.8077 | 0.9165 | 280 | 0.7553 |
67
- | 0.8013 | 0.9820 | 300 | 0.7501 |
68
- | 0.7948 | 1.0458 | 320 | 0.7447 |
69
- | 0.783 | 1.1113 | 340 | 0.7394 |
70
- | 0.7727 | 1.1768 | 360 | 0.7372 |
71
- | 0.777 | 1.2422 | 380 | 0.7331 |
72
- | 0.7711 | 1.3077 | 400 | 0.7309 |
73
- | 0.7642 | 1.3732 | 420 | 0.7289 |
74
- | 0.7631 | 1.4386 | 440 | 0.7267 |
75
- | 0.7581 | 1.5041 | 460 | 0.7250 |
76
- | 0.7606 | 1.5696 | 480 | 0.7233 |
77
- | 0.7578 | 1.6350 | 500 | 0.7223 |
78
- | 0.7562 | 1.7005 | 520 | 0.7208 |
79
- | 0.7497 | 1.7660 | 540 | 0.7195 |
80
- | 0.7508 | 1.8314 | 560 | 0.7179 |
81
- | 0.7476 | 1.8969 | 580 | 0.7168 |
82
- | 0.7503 | 1.9624 | 600 | 0.7165 |
83
- | 0.7414 | 2.0262 | 620 | 0.7164 |
84
- | 0.7425 | 2.0917 | 640 | 0.7159 |
85
- | 0.7451 | 2.1571 | 660 | 0.7146 |
86
- | 0.7452 | 2.2226 | 680 | 0.7147 |
87
- | 0.7446 | 2.2881 | 700 | 0.7138 |
88
- | 0.7437 | 2.3535 | 720 | 0.7140 |
89
- | 0.7397 | 2.4190 | 740 | 0.7131 |
90
- | 0.7426 | 2.4845 | 760 | 0.7130 |
91
- | 0.7421 | 2.5499 | 780 | 0.7127 |
92
- | 0.7408 | 2.6154 | 800 | 0.7135 |
93
- | 0.7413 | 2.6809 | 820 | 0.7135 |
94
- | 0.7404 | 2.7463 | 840 | 0.7131 |
95
- | 0.7373 | 2.8118 | 860 | 0.7128 |
96
- | 0.7451 | 2.8773 | 880 | 0.7134 |
97
- | 0.7407 | 2.9427 | 900 | 0.7127 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
98
 
99
 
100
  ### Framework versions
 
1
  ---
2
  library_name: peft
3
  license: mit
4
+ base_model: gpt2
5
  tags:
6
  - generated_from_trainer
7
  model-index:
8
+ - name: Se124M100KInfPrompt
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
+ # Se124M100KInfPrompt
16
 
17
+ This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on an unknown dataset.
18
  It achieves the following results on the evaluation set:
19
+ - Loss: 0.3662
20
 
21
  ## Model description
22
 
 
36
 
37
  The following hyperparameters were used during training:
38
  - learning_rate: 5e-05
39
+ - train_batch_size: 16
40
+ - eval_batch_size: 16
41
  - seed: 42
42
  - gradient_accumulation_steps: 4
43
+ - total_train_batch_size: 64
44
  - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
45
  - lr_scheduler_type: cosine
46
  - lr_scheduler_warmup_ratio: 0.03
 
50
 
51
  | Training Loss | Epoch | Step | Validation Loss |
52
  |:-------------:|:------:|:----:|:---------------:|
53
+ | 2.8209 | 0.0164 | 20 | 2.4755 |
54
+ | 2.7724 | 0.0327 | 40 | 2.4510 |
55
+ | 2.7076 | 0.0491 | 60 | 2.3560 |
56
+ | 2.5366 | 0.0655 | 80 | 2.1926 |
57
+ | 2.2739 | 0.0818 | 100 | 1.9387 |
58
+ | 1.9339 | 0.0982 | 120 | 1.6141 |
59
+ | 1.5753 | 0.1146 | 140 | 1.2455 |
60
+ | 1.254 | 0.1309 | 160 | 0.9377 |
61
+ | 1.0138 | 0.1473 | 180 | 0.7895 |
62
+ | 0.866 | 0.1637 | 200 | 0.6805 |
63
+ | 0.7571 | 0.1800 | 220 | 0.5998 |
64
+ | 0.6901 | 0.1964 | 240 | 0.5554 |
65
+ | 0.629 | 0.2128 | 260 | 0.5310 |
66
+ | 0.5879 | 0.2291 | 280 | 0.5076 |
67
+ | 0.5703 | 0.2455 | 300 | 0.4930 |
68
+ | 0.5546 | 0.2619 | 320 | 0.4845 |
69
+ | 0.5419 | 0.2782 | 340 | 0.4762 |
70
+ | 0.5179 | 0.2946 | 360 | 0.4656 |
71
+ | 0.5211 | 0.3110 | 380 | 0.4595 |
72
+ | 0.5039 | 0.3273 | 400 | 0.4543 |
73
+ | 0.4992 | 0.3437 | 420 | 0.4524 |
74
+ | 0.4937 | 0.3601 | 440 | 0.4477 |
75
+ | 0.4801 | 0.3764 | 460 | 0.4409 |
76
+ | 0.4805 | 0.3928 | 480 | 0.4380 |
77
+ | 0.4805 | 0.4092 | 500 | 0.4354 |
78
+ | 0.468 | 0.4255 | 520 | 0.4343 |
79
+ | 0.4759 | 0.4419 | 540 | 0.4319 |
80
+ | 0.4614 | 0.4583 | 560 | 0.4284 |
81
+ | 0.4622 | 0.4746 | 580 | 0.4272 |
82
+ | 0.4608 | 0.4910 | 600 | 0.4267 |
83
+ | 0.4621 | 0.5074 | 620 | 0.4236 |
84
+ | 0.4569 | 0.5237 | 640 | 0.4238 |
85
+ | 0.4519 | 0.5401 | 660 | 0.4219 |
86
+ | 0.4478 | 0.5565 | 680 | 0.4189 |
87
+ | 0.4524 | 0.5728 | 700 | 0.4167 |
88
+ | 0.4489 | 0.5892 | 720 | 0.4147 |
89
+ | 0.4452 | 0.6056 | 740 | 0.4150 |
90
+ | 0.4424 | 0.6219 | 760 | 0.4118 |
91
+ | 0.4355 | 0.6383 | 780 | 0.4117 |
92
+ | 0.4396 | 0.6547 | 800 | 0.4112 |
93
+ | 0.4432 | 0.6710 | 820 | 0.4078 |
94
+ | 0.44 | 0.6874 | 840 | 0.4051 |
95
+ | 0.4341 | 0.7038 | 860 | 0.4050 |
96
+ | 0.4425 | 0.7201 | 880 | 0.4018 |
97
+ | 0.4387 | 0.7365 | 900 | 0.4016 |
98
+ | 0.4369 | 0.7529 | 920 | 0.4031 |
99
+ | 0.437 | 0.7692 | 940 | 0.3967 |
100
+ | 0.4314 | 0.7856 | 960 | 0.4007 |
101
+ | 0.4371 | 0.8020 | 980 | 0.3943 |
102
+ | 0.4364 | 0.8183 | 1000 | 0.3986 |
103
+ | 0.4292 | 0.8347 | 1020 | 0.3970 |
104
+ | 0.427 | 0.8511 | 1040 | 0.3951 |
105
+ | 0.431 | 0.8674 | 1060 | 0.3941 |
106
+ | 0.4327 | 0.8838 | 1080 | 0.3958 |
107
+ | 0.4263 | 0.9002 | 1100 | 0.3930 |
108
+ | 0.429 | 0.9165 | 1120 | 0.3901 |
109
+ | 0.4277 | 0.9329 | 1140 | 0.3907 |
110
+ | 0.4251 | 0.9493 | 1160 | 0.3906 |
111
+ | 0.4279 | 0.9656 | 1180 | 0.3891 |
112
+ | 0.4249 | 0.9820 | 1200 | 0.3884 |
113
+ | 0.4213 | 0.9984 | 1220 | 0.3891 |
114
+ | 0.4192 | 1.0147 | 1240 | 0.3870 |
115
+ | 0.4263 | 1.0311 | 1260 | 0.3852 |
116
+ | 0.4219 | 1.0475 | 1280 | 0.3897 |
117
+ | 0.4256 | 1.0638 | 1300 | 0.3846 |
118
+ | 0.4129 | 1.0802 | 1320 | 0.3855 |
119
+ | 0.4184 | 1.0966 | 1340 | 0.3841 |
120
+ | 0.4207 | 1.1129 | 1360 | 0.3835 |
121
+ | 0.418 | 1.1293 | 1380 | 0.3808 |
122
+ | 0.4153 | 1.1457 | 1400 | 0.3827 |
123
+ | 0.4247 | 1.1620 | 1420 | 0.3812 |
124
+ | 0.421 | 1.1784 | 1440 | 0.3807 |
125
+ | 0.4127 | 1.1948 | 1460 | 0.3802 |
126
+ | 0.4233 | 1.2111 | 1480 | 0.3794 |
127
+ | 0.4235 | 1.2275 | 1500 | 0.3782 |
128
+ | 0.4184 | 1.2439 | 1520 | 0.3785 |
129
+ | 0.4171 | 1.2602 | 1540 | 0.3796 |
130
+ | 0.4181 | 1.2766 | 1560 | 0.3811 |
131
+ | 0.4126 | 1.2930 | 1580 | 0.3780 |
132
+ | 0.4188 | 1.3093 | 1600 | 0.3760 |
133
+ | 0.4162 | 1.3257 | 1620 | 0.3769 |
134
+ | 0.4192 | 1.3421 | 1640 | 0.3770 |
135
+ | 0.4153 | 1.3584 | 1660 | 0.3763 |
136
+ | 0.4187 | 1.3748 | 1680 | 0.3737 |
137
+ | 0.4138 | 1.3912 | 1700 | 0.3755 |
138
+ | 0.4115 | 1.4075 | 1720 | 0.3755 |
139
+ | 0.4118 | 1.4239 | 1740 | 0.3756 |
140
+ | 0.4036 | 1.4403 | 1760 | 0.3742 |
141
+ | 0.4161 | 1.4566 | 1780 | 0.3731 |
142
+ | 0.4102 | 1.4730 | 1800 | 0.3740 |
143
+ | 0.4118 | 1.4894 | 1820 | 0.3731 |
144
+ | 0.4102 | 1.5057 | 1840 | 0.3732 |
145
+ | 0.4143 | 1.5221 | 1860 | 0.3744 |
146
+ | 0.4118 | 1.5385 | 1880 | 0.3729 |
147
+ | 0.4179 | 1.5548 | 1900 | 0.3721 |
148
+ | 0.4092 | 1.5712 | 1920 | 0.3716 |
149
+ | 0.4109 | 1.5876 | 1940 | 0.3726 |
150
+ | 0.4137 | 1.6039 | 1960 | 0.3713 |
151
+ | 0.4067 | 1.6203 | 1980 | 0.3714 |
152
+ | 0.4131 | 1.6367 | 2000 | 0.3725 |
153
+ | 0.4103 | 1.6530 | 2020 | 0.3702 |
154
+ | 0.4044 | 1.6694 | 2040 | 0.3711 |
155
+ | 0.4105 | 1.6858 | 2060 | 0.3727 |
156
+ | 0.4063 | 1.7021 | 2080 | 0.3712 |
157
+ | 0.4109 | 1.7185 | 2100 | 0.3709 |
158
+ | 0.4114 | 1.7349 | 2120 | 0.3706 |
159
+ | 0.4148 | 1.7512 | 2140 | 0.3711 |
160
+ | 0.4081 | 1.7676 | 2160 | 0.3693 |
161
+ | 0.4062 | 1.7840 | 2180 | 0.3694 |
162
+ | 0.4152 | 1.8003 | 2200 | 0.3699 |
163
+ | 0.4043 | 1.8167 | 2220 | 0.3686 |
164
+ | 0.4046 | 1.8331 | 2240 | 0.3705 |
165
+ | 0.4136 | 1.8494 | 2260 | 0.3684 |
166
+ | 0.4073 | 1.8658 | 2280 | 0.3701 |
167
+ | 0.4089 | 1.8822 | 2300 | 0.3689 |
168
+ | 0.4075 | 1.8985 | 2320 | 0.3679 |
169
+ | 0.409 | 1.9149 | 2340 | 0.3694 |
170
+ | 0.4096 | 1.9313 | 2360 | 0.3677 |
171
+ | 0.4114 | 1.9476 | 2380 | 0.3686 |
172
+ | 0.4083 | 1.9640 | 2400 | 0.3676 |
173
+ | 0.4066 | 1.9804 | 2420 | 0.3696 |
174
+ | 0.4053 | 1.9967 | 2440 | 0.3677 |
175
+ | 0.4087 | 2.0131 | 2460 | 0.3688 |
176
+ | 0.4055 | 2.0295 | 2480 | 0.3680 |
177
+ | 0.4103 | 2.0458 | 2500 | 0.3678 |
178
+ | 0.4031 | 2.0622 | 2520 | 0.3685 |
179
+ | 0.4111 | 2.0786 | 2540 | 0.3674 |
180
+ | 0.413 | 2.0949 | 2560 | 0.3675 |
181
+ | 0.4135 | 2.1113 | 2580 | 0.3674 |
182
+ | 0.4085 | 2.1277 | 2600 | 0.3664 |
183
+ | 0.4029 | 2.1440 | 2620 | 0.3683 |
184
+ | 0.4023 | 2.1604 | 2640 | 0.3677 |
185
+ | 0.4087 | 2.1768 | 2660 | 0.3673 |
186
+ | 0.4088 | 2.1931 | 2680 | 0.3678 |
187
+ | 0.4064 | 2.2095 | 2700 | 0.3664 |
188
+ | 0.4067 | 2.2259 | 2720 | 0.3669 |
189
+ | 0.4047 | 2.2422 | 2740 | 0.3662 |
190
+ | 0.4069 | 2.2586 | 2760 | 0.3666 |
191
+ | 0.4028 | 2.2750 | 2780 | 0.3663 |
192
+ | 0.4101 | 2.2913 | 2800 | 0.3664 |
193
+ | 0.4061 | 2.3077 | 2820 | 0.3663 |
194
+ | 0.4056 | 2.3241 | 2840 | 0.3657 |
195
+ | 0.4073 | 2.3404 | 2860 | 0.3660 |
196
+ | 0.4096 | 2.3568 | 2880 | 0.3665 |
197
+ | 0.4034 | 2.3732 | 2900 | 0.3667 |
198
+ | 0.4067 | 2.3895 | 2920 | 0.3668 |
199
+ | 0.4032 | 2.4059 | 2940 | 0.3673 |
200
+ | 0.4082 | 2.4223 | 2960 | 0.3666 |
201
+ | 0.4048 | 2.4386 | 2980 | 0.3660 |
202
+ | 0.4058 | 2.4550 | 3000 | 0.3661 |
203
+ | 0.4066 | 2.4714 | 3020 | 0.3663 |
204
+ | 0.4128 | 2.4877 | 3040 | 0.3662 |
205
+ | 0.4104 | 2.5041 | 3060 | 0.3658 |
206
+ | 0.4057 | 2.5205 | 3080 | 0.3658 |
207
+ | 0.408 | 2.5368 | 3100 | 0.3660 |
208
+ | 0.4053 | 2.5532 | 3120 | 0.3660 |
209
+ | 0.3998 | 2.5696 | 3140 | 0.3664 |
210
+ | 0.4007 | 2.5859 | 3160 | 0.3659 |
211
+ | 0.402 | 2.6023 | 3180 | 0.3660 |
212
+ | 0.4017 | 2.6187 | 3200 | 0.3660 |
213
+ | 0.4069 | 2.6350 | 3220 | 0.3659 |
214
+ | 0.4028 | 2.6514 | 3240 | 0.3662 |
215
+ | 0.4014 | 2.6678 | 3260 | 0.3663 |
216
+ | 0.4023 | 2.6841 | 3280 | 0.3666 |
217
+ | 0.4025 | 2.7005 | 3300 | 0.3668 |
218
+ | 0.4027 | 2.7169 | 3320 | 0.3661 |
219
+ | 0.404 | 2.7332 | 3340 | 0.3659 |
220
+ | 0.4064 | 2.7496 | 3360 | 0.3663 |
221
+ | 0.4059 | 2.7660 | 3380 | 0.3659 |
222
+ | 0.4007 | 2.7823 | 3400 | 0.3659 |
223
+ | 0.4044 | 2.7987 | 3420 | 0.3663 |
224
+ | 0.4075 | 2.8151 | 3440 | 0.3658 |
225
+ | 0.4053 | 2.8314 | 3460 | 0.3660 |
226
+ | 0.4003 | 2.8478 | 3480 | 0.3664 |
227
+ | 0.4078 | 2.8642 | 3500 | 0.3662 |
228
+ | 0.4067 | 2.8805 | 3520 | 0.3661 |
229
+ | 0.4025 | 2.8969 | 3540 | 0.3660 |
230
+ | 0.4018 | 2.9133 | 3560 | 0.3658 |
231
+ | 0.4014 | 2.9296 | 3580 | 0.3661 |
232
+ | 0.4036 | 2.9460 | 3600 | 0.3661 |
233
+ | 0.4031 | 2.9624 | 3620 | 0.3657 |
234
+ | 0.4022 | 2.9787 | 3640 | 0.3660 |
235
+ | 0.4036 | 2.9951 | 3660 | 0.3662 |
236
 
237
 
238
  ### Framework versions
all_results.json CHANGED
@@ -1,13 +1,13 @@
1
  {
2
- "epoch": 2.9918166939443536,
3
  "eval_loss": 0.7127311825752258,
4
  "eval_runtime": 27.5772,
5
  "eval_samples_per_second": 607.423,
6
  "eval_steps_per_second": 9.501,
7
  "perplexity": 2.0395540531763667,
8
- "total_flos": 5.437860691102925e+16,
9
- "train_loss": 0.9607737994584881,
10
- "train_runtime": 2198.2213,
11
- "train_samples_per_second": 106.683,
12
- "train_steps_per_second": 0.416
13
  }
 
1
  {
2
+ "epoch": 3.0,
3
  "eval_loss": 0.7127311825752258,
4
  "eval_runtime": 27.5772,
5
  "eval_samples_per_second": 607.423,
6
  "eval_steps_per_second": 9.501,
7
  "perplexity": 2.0395540531763667,
8
+ "total_flos": 1.5045077015543808e+16,
9
+ "train_loss": 0.5129519925375031,
10
+ "train_runtime": 8559.3775,
11
+ "train_samples_per_second": 27.411,
12
+ "train_steps_per_second": 0.428
13
  }
train_results.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "epoch": 2.9918166939443536,
3
- "total_flos": 5.437860691102925e+16,
4
- "train_loss": 0.9607737994584881,
5
- "train_runtime": 2198.2213,
6
- "train_samples_per_second": 106.683,
7
- "train_steps_per_second": 0.416
8
  }
 
1
  {
2
+ "epoch": 3.0,
3
+ "total_flos": 1.5045077015543808e+16,
4
+ "train_loss": 0.5129519925375031,
5
+ "train_runtime": 8559.3775,
6
+ "train_samples_per_second": 27.411,
7
+ "train_steps_per_second": 0.428
8
  }