End of training
Browse files
README.md
CHANGED
@@ -16,14 +16,14 @@ This student model is distilled from the teacher model [gpt2](https://huggingfac
|
|
16 |
The [Distily](https://github.com/lapp0/distily) library was used for this distillation.
|
17 |
|
18 |
It achieves the following results on the evaluation set:
|
19 |
-
- eval_enwikippl:
|
20 |
-
- eval_frwikippl:
|
21 |
-
- eval_zhwikippl:
|
22 |
-
- eval_tinystoriesppl:
|
23 |
-
- eval_loss:
|
24 |
-
- eval_runtime: 25.
|
25 |
-
- eval_samples_per_second:
|
26 |
-
- eval_steps_per_second: 12.
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
should probably proofread and complete it, then remove this comment.
|
@@ -53,8 +53,8 @@ The following hyperparameters were used during training:
|
|
53 |
- eval_batch_size: 8
|
54 |
- seed: 42
|
55 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
56 |
-
- lr_scheduler_type:
|
57 |
-
- lr_scheduler_warmup_ratio: 0.
|
58 |
- num_epochs: 1.0
|
59 |
|
60 |
### Resource Usage
|
@@ -64,69 +64,69 @@ Peak GPU Memory: 7.2012 GB
|
|
64 |
| step | epoch | enwikippl | frwikippl | loss | runtime | samples_per_second | steps_per_second | tinystoriesppl | zhwikippl |
|
65 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
|
66 |
| **teacher eval** | | 43.25 | 61.25 | | | | | 11.6875 | 19.125 |
|
67 |
-
| 0 | 0 |
|
68 |
-
| 1000 | 0.0162 |
|
69 |
-
| 2000 | 0.0323 |
|
70 |
-
| 3000 | 0.0485 |
|
71 |
-
| 4000 | 0.0646 |
|
72 |
-
| 5000 | 0.0808 |
|
73 |
-
| 6000 | 0.0970 |
|
74 |
-
| 7000 | 0.1131 |
|
75 |
-
| 8000 | 0.1293 |
|
76 |
-
| 9000 | 0.1455 |
|
77 |
-
| 10000 | 0.1616 |
|
78 |
-
| 11000 | 0.1778 |
|
79 |
-
| 12000 | 0.1939 |
|
80 |
-
| 13000 | 0.2101 |
|
81 |
-
| 14000 | 0.2263 |
|
82 |
-
| 15000 | 0.2424 |
|
83 |
-
| 16000 | 0.2586 |
|
84 |
-
| 17000 | 0.2747 |
|
85 |
-
| 18000 | 0.2909 |
|
86 |
-
| 19000 | 0.3071 |
|
87 |
-
| 20000 | 0.3232 |
|
88 |
-
| 21000 | 0.3394 |
|
89 |
-
| 22000 | 0.3556 |
|
90 |
-
| 23000 | 0.3717 |
|
91 |
-
| 24000 | 0.3879 |
|
92 |
-
| 25000 | 0.4040 |
|
93 |
-
| 26000 | 0.4202 |
|
94 |
-
| 27000 | 0.4364 |
|
95 |
-
| 28000 | 0.4525 |
|
96 |
-
| 29000 | 0.4687 |
|
97 |
-
| 30000 | 0.4848 |
|
98 |
-
| 31000 | 0.5010 |
|
99 |
-
| 32000 | 0.5172 |
|
100 |
-
| 33000 | 0.5333 |
|
101 |
-
| 34000 | 0.5495 |
|
102 |
-
| 35000 | 0.5657 |
|
103 |
-
| 36000 | 0.5818 |
|
104 |
-
| 37000 | 0.5980 |
|
105 |
-
| 38000 | 0.6141 |
|
106 |
-
| 39000 | 0.6303 |
|
107 |
-
| 40000 | 0.6465 |
|
108 |
-
| 41000 | 0.6626 |
|
109 |
-
| 42000 | 0.6788 |
|
110 |
-
| 43000 | 0.6949 |
|
111 |
-
| 44000 | 0.7111 |
|
112 |
-
| 45000 | 0.7273 |
|
113 |
-
| 46000 | 0.7434 |
|
114 |
-
| 47000 | 0.7596 |
|
115 |
-
| 48000 | 0.7758 |
|
116 |
-
| 49000 | 0.7919 |
|
117 |
-
| 50000 | 0.8081 |
|
118 |
-
| 51000 | 0.8242 |
|
119 |
-
| 52000 | 0.8404 |
|
120 |
-
| 53000 | 0.8566 |
|
121 |
-
| 54000 | 0.8727 |
|
122 |
-
| 55000 | 0.8889 |
|
123 |
-
| 56000 | 0.9051 |
|
124 |
-
| 57000 | 0.9212 |
|
125 |
-
| 58000 | 0.9374 |
|
126 |
-
| 59000 | 0.9535 |
|
127 |
-
| 60000 | 0.9697 |
|
128 |
-
| 61000 | 0.9859 |
|
129 |
-
| 61875 | 1.0 |
|
130 |
|
131 |
### Framework versions
|
132 |
- Distily 0.2.0
|
|
|
16 |
The [Distily](https://github.com/lapp0/distily) library was used for this distillation.
|
17 |
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- eval_enwikippl: 216.0
|
20 |
+
- eval_frwikippl: 876.0
|
21 |
+
- eval_zhwikippl: 173.0
|
22 |
+
- eval_tinystoriesppl: 178.0
|
23 |
+
- eval_loss: 1.2693
|
24 |
+
- eval_runtime: 25.4344
|
25 |
+
- eval_samples_per_second: 98.292
|
26 |
+
- eval_steps_per_second: 12.306
|
27 |
|
28 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
29 |
should probably proofread and complete it, then remove this comment.
|
|
|
53 |
- eval_batch_size: 8
|
54 |
- seed: 42
|
55 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
56 |
+
- lr_scheduler_type: linear
|
57 |
+
- lr_scheduler_warmup_ratio: 0.5
|
58 |
- num_epochs: 1.0
|
59 |
|
60 |
### Resource Usage
|
|
|
64 |
| step | epoch | enwikippl | frwikippl | loss | runtime | samples_per_second | steps_per_second | tinystoriesppl | zhwikippl |
|
65 |
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
|
66 |
| **teacher eval** | | 43.25 | 61.25 | | | | | 11.6875 | 19.125 |
|
67 |
+
| 0 | 0 | 1606317768704.0 | 119297011613696.0 | 20.7344 | 25.4007 | 98.423 | 12.323 | 5066719232.0 | 19104014532608.0 |
|
68 |
+
| 1000 | 0.0162 | 81408.0 | 958464.0 | 5.1263 | 25.4848 | 98.098 | 12.282 | 16384.0 | 1253376.0 |
|
69 |
+
| 2000 | 0.0323 | 4032.0 | 47360.0 | 3.2154 | 25.5066 | 98.014 | 12.271 | 2368.0 | 256000.0 |
|
70 |
+
| 3000 | 0.0485 | 1104.0 | 5792.0 | 2.3506 | 25.4684 | 98.161 | 12.29 | 724.0 | 8960.0 |
|
71 |
+
| 4000 | 0.0646 | 632.0 | 3984.0 | 1.9871 | 25.4984 | 98.046 | 12.275 | 410.0 | 2368.0 |
|
72 |
+
| 5000 | 0.0808 | 418.0 | 2096.0 | 1.6792 | 25.4486 | 98.237 | 12.299 | 310.0 | 572.0 |
|
73 |
+
| 6000 | 0.0970 | 310.0 | 1384.0 | 1.5115 | 25.5353 | 97.904 | 12.258 | 241.0 | 284.0 |
|
74 |
+
| 7000 | 0.1131 | 256.0 | 1040.0 | 1.3917 | 25.5364 | 97.899 | 12.257 | 203.0 | 205.0 |
|
75 |
+
| 8000 | 0.1293 | 229.0 | 892.0 | 1.3296 | 25.4687 | 98.16 | 12.29 | 183.0 | 224.0 |
|
76 |
+
| 9000 | 0.1455 | 216.0 | 876.0 | 1.2693 | 25.4344 | 98.292 | 12.306 | 178.0 | 173.0 |
|
77 |
+
| 10000 | 0.1616 | 192.0 | 716.0 | 1.1774 | 25.4691 | 98.158 | 12.289 | 165.0 | 198.0 |
|
78 |
+
| 11000 | 0.1778 | 255.0 | 824.0 | 1.0750 | 25.5641 | 97.793 | 12.244 | 296.0 | 169.0 |
|
79 |
+
| 12000 | 0.1939 | 264.0 | 744.0 | 0.9937 | 25.5508 | 97.844 | 12.25 | 292.0 | 147.0 |
|
80 |
+
| 13000 | 0.2101 | 231.0 | 608.0 | 0.9388 | 25.6185 | 97.586 | 12.218 | 192.0 | 158.0 |
|
81 |
+
| 14000 | 0.2263 | 221.0 | 552.0 | 0.8904 | 25.5619 | 97.802 | 12.245 | 178.0 | 127.5 |
|
82 |
+
| 15000 | 0.2424 | 221.0 | 608.0 | 0.8620 | 25.5139 | 97.986 | 12.268 | 145.0 | 128.0 |
|
83 |
+
| 16000 | 0.2586 | 194.0 | 564.0 | 0.8357 | 25.4869 | 98.09 | 12.281 | 123.0 | 152.0 |
|
84 |
+
| 17000 | 0.2747 | 143.0 | 468.0 | 0.8056 | 25.4845 | 98.099 | 12.282 | 88.0 | 131.0 |
|
85 |
+
| 18000 | 0.2909 | 135.0 | 512.0 | 0.7808 | 25.6085 | 97.624 | 12.222 | 81.5 | 121.5 |
|
86 |
+
| 19000 | 0.3071 | 130.0 | 484.0 | 0.8070 | 25.6119 | 97.611 | 12.221 | 80.5 | 123.5 |
|
87 |
+
| 20000 | 0.3232 | 102.5 | 414.0 | 0.7258 | 25.5902 | 97.694 | 12.231 | 73.0 | 136.0 |
|
88 |
+
| 21000 | 0.3394 | 92.5 | 370.0 | 0.6690 | 25.484 | 98.101 | 12.282 | 72.5 | 117.0 |
|
89 |
+
| 22000 | 0.3556 | 89.5 | 320.0 | 0.6363 | 25.4959 | 98.055 | 12.277 | 72.0 | 109.5 |
|
90 |
+
| 23000 | 0.3717 | 93.5 | 302.0 | 0.6073 | 25.5285 | 97.93 | 12.261 | 76.0 | 105.5 |
|
91 |
+
| 24000 | 0.3879 | 93.5 | 306.0 | 0.5868 | 25.52 | 97.962 | 12.265 | 76.5 | 106.0 |
|
92 |
+
| 25000 | 0.4040 | 126.5 | 336.0 | 0.5717 | 25.5413 | 97.881 | 12.255 | 108.0 | 132.0 |
|
93 |
+
| 26000 | 0.4202 | 72.0 | 250.0 | 0.5578 | 25.5841 | 97.717 | 12.234 | 61.25 | 114.0 |
|
94 |
+
| 27000 | 0.4364 | 87.5 | 262.0 | 0.5462 | 25.5236 | 97.948 | 12.263 | 63.75 | 116.0 |
|
95 |
+
| 28000 | 0.4525 | 72.0 | 228.0 | 0.5431 | 25.5362 | 97.9 | 12.257 | 57.75 | 107.0 |
|
96 |
+
| 29000 | 0.4687 | 75.5 | 247.0 | 0.5524 | 25.5375 | 97.895 | 12.256 | 56.75 | 105.5 |
|
97 |
+
| 30000 | 0.4848 | 72.0 | 222.0 | 0.5459 | 25.5244 | 97.945 | 12.263 | 53.25 | 114.5 |
|
98 |
+
| 31000 | 0.5010 | 71.5 | 230.0 | 0.5460 | 25.4562 | 98.208 | 12.296 | 54.25 | 151.0 |
|
99 |
+
| 32000 | 0.5172 | 69.5 | 246.0 | 0.5391 | 25.5258 | 97.94 | 12.262 | 54.5 | 194.0 |
|
100 |
+
| 33000 | 0.5333 | 70.5 | 242.0 | 0.5283 | 25.4905 | 98.076 | 12.279 | 53.75 | 117.5 |
|
101 |
+
| 34000 | 0.5495 | 76.5 | 246.0 | 0.5147 | 25.4567 | 98.206 | 12.295 | 62.75 | 114.0 |
|
102 |
+
| 35000 | 0.5657 | 68.0 | 216.0 | 0.5014 | 25.4939 | 98.063 | 12.277 | 50.25 | 87.0 |
|
103 |
+
| 36000 | 0.5818 | 71.0 | 225.0 | 0.4986 | 25.4871 | 98.089 | 12.281 | 52.75 | 90.0 |
|
104 |
+
| 37000 | 0.5980 | 79.5 | 231.0 | 0.4939 | 25.509 | 98.005 | 12.27 | 54.0 | 112.5 |
|
105 |
+
| 38000 | 0.6141 | 75.5 | 232.0 | 0.4880 | 25.4997 | 98.04 | 12.275 | 58.75 | 116.5 |
|
106 |
+
| 39000 | 0.6303 | 82.0 | 241.0 | 0.4879 | 25.519 | 97.966 | 12.265 | 61.75 | 123.0 |
|
107 |
+
| 40000 | 0.6465 | 73.5 | 220.0 | 0.4759 | 25.4982 | 98.046 | 12.275 | 55.25 | 126.0 |
|
108 |
+
| 41000 | 0.6626 | 70.5 | 212.0 | 0.4738 | 25.5094 | 98.003 | 12.27 | 55.5 | 102.0 |
|
109 |
+
| 42000 | 0.6788 | 79.0 | 227.0 | 0.4699 | 25.527 | 97.935 | 12.262 | 59.25 | 82.5 |
|
110 |
+
| 43000 | 0.6949 | 82.0 | 227.0 | 0.4613 | 25.5453 | 97.865 | 12.253 | 59.25 | 92.5 |
|
111 |
+
| 44000 | 0.7111 | 78.0 | 210.0 | 0.4227 | 25.5427 | 97.875 | 12.254 | 61.5 | 79.0 |
|
112 |
+
| 45000 | 0.7273 | 72.0 | 192.0 | 0.4068 | 25.4354 | 98.288 | 12.306 | 55.5 | 72.0 |
|
113 |
+
| 46000 | 0.7434 | 72.0 | 185.0 | 0.3952 | 25.5194 | 97.965 | 12.265 | 56.0 | 53.0 |
|
114 |
+
| 47000 | 0.7596 | 68.0 | 178.0 | 0.3928 | 25.5018 | 98.032 | 12.274 | 51.25 | 54.25 |
|
115 |
+
| 48000 | 0.7758 | 71.5 | 183.0 | 0.3873 | 25.5018 | 98.032 | 12.274 | 54.5 | 63.0 |
|
116 |
+
| 49000 | 0.7919 | 71.0 | 177.0 | 0.3853 | 25.5149 | 97.982 | 12.267 | 52.5 | 66.5 |
|
117 |
+
| 50000 | 0.8081 | 71.0 | 174.0 | 0.3802 | 25.4786 | 98.122 | 12.285 | 54.75 | 61.75 |
|
118 |
+
| 51000 | 0.8242 | 73.5 | 184.0 | 0.3797 | 25.4575 | 98.203 | 12.295 | 56.25 | 59.0 |
|
119 |
+
| 52000 | 0.8404 | 75.5 | 187.0 | 0.3784 | 25.4298 | 98.31 | 12.308 | 57.5 | 68.5 |
|
120 |
+
| 53000 | 0.8566 | 74.5 | 185.0 | 0.3741 | 25.5041 | 98.023 | 12.273 | 55.75 | 59.75 |
|
121 |
+
| 54000 | 0.8727 | 73.0 | 178.0 | 0.3686 | 25.5431 | 97.874 | 12.254 | 54.75 | 56.25 |
|
122 |
+
| 55000 | 0.8889 | 72.0 | 180.0 | 0.3649 | 25.447 | 98.244 | 12.3 | 55.0 | 55.0 |
|
123 |
+
| 56000 | 0.9051 | 71.5 | 179.0 | 0.3636 | 25.5222 | 97.954 | 12.264 | 53.75 | 53.0 |
|
124 |
+
| 57000 | 0.9212 | 71.5 | 176.0 | 0.3619 | 25.4887 | 98.083 | 12.28 | 54.25 | 50.5 |
|
125 |
+
| 58000 | 0.9374 | 72.0 | 177.0 | 0.3605 | 25.4451 | 98.251 | 12.301 | 55.0 | 50.25 |
|
126 |
+
| 59000 | 0.9535 | 72.0 | 177.0 | 0.3598 | 25.5041 | 98.023 | 12.273 | 54.5 | 50.5 |
|
127 |
+
| 60000 | 0.9697 | 72.0 | 177.0 | 0.3589 | 25.5086 | 98.006 | 12.27 | 54.0 | 50.5 |
|
128 |
+
| 61000 | 0.9859 | 72.0 | 177.0 | 0.3588 | 25.4042 | 98.409 | 12.321 | 54.5 | 50.5 |
|
129 |
+
| 61875 | 1.0 | 72.0 | 177.0 | 0.3587 | 25.56 | 97.809 | 12.246 | 54.25 | 50.25 |
|
130 |
|
131 |
### Framework versions
|
132 |
- Distily 0.2.0
|
logs/lr_scheduler_type=linear, warmup_ratio=0.5/events.out.tfevents.1724151237.f383272e719b
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9d25613f58ba0775aca44b60d5e261053940f4379382de62faacfb2ea18a5672
|
3 |
+
size 312
|