Upload folder using huggingface_hub
Browse files- README.md +1 -1
- history.csv +3 -3
- metrics.json +109 -109
- model.pt +1 -1
README.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
# Week 2 MoE Seq2Seq (hash routing)
|
| 2 |
|
| 3 |
-
* Best validation loss: 5.
|
| 4 |
* Top-k: 1
|
| 5 |
* Aux loss coef: 0.0
|
| 6 |
|
|
|
|
| 1 |
# Week 2 MoE Seq2Seq (hash routing)
|
| 2 |
|
| 3 |
+
* Best validation loss: 5.6068
|
| 4 |
* Top-k: 1
|
| 5 |
* Aux loss coef: 0.0
|
| 6 |
|
history.csv
CHANGED
|
@@ -1,4 +1,4 @@
|
|
| 1 |
epoch,train_loss,train_aux_loss,train_perplexity,val_loss,val_aux_loss,val_perplexity
|
| 2 |
-
1,6.
|
| 3 |
-
2,5.
|
| 4 |
-
3,5.
|
|
|
|
| 1 |
epoch,train_loss,train_aux_loss,train_perplexity,val_loss,val_aux_loss,val_perplexity
|
| 2 |
+
1,6.783949644676193,0.0,883.5515563058923,6.120485806811677,0.0,455.08572455110965
|
| 3 |
+
2,5.72200544220322,0.0,305.51700570051946,5.783777441596244,0.0,324.984484631975
|
| 4 |
+
3,5.255330442014935,0.0,191.58478305751783,5.606828254184237,0.0,272.279266370978
|
metrics.json
CHANGED
|
@@ -2,120 +2,120 @@
|
|
| 2 |
"history": [
|
| 3 |
{
|
| 4 |
"epoch": 1,
|
| 5 |
-
"train_loss": 6.
|
| 6 |
"train_aux_loss": 0.0,
|
| 7 |
-
"train_perplexity":
|
| 8 |
-
"val_loss": 6.
|
| 9 |
"val_aux_loss": 0.0,
|
| 10 |
-
"val_perplexity":
|
| 11 |
},
|
| 12 |
{
|
| 13 |
"epoch": 2,
|
| 14 |
-
"train_loss": 5.
|
| 15 |
"train_aux_loss": 0.0,
|
| 16 |
-
"train_perplexity":
|
| 17 |
-
"val_loss": 5.
|
| 18 |
"val_aux_loss": 0.0,
|
| 19 |
-
"val_perplexity":
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 3,
|
| 23 |
-
"train_loss": 5.
|
| 24 |
"train_aux_loss": 0.0,
|
| 25 |
-
"train_perplexity":
|
| 26 |
-
"val_loss": 5.
|
| 27 |
"val_aux_loss": 0.0,
|
| 28 |
-
"val_perplexity": 272.
|
| 29 |
}
|
| 30 |
],
|
| 31 |
"train_expert_usage": [
|
| 32 |
{
|
| 33 |
"encoder": [
|
| 34 |
[
|
| 35 |
-
1.
|
| 36 |
-
1.
|
| 37 |
-
1.
|
| 38 |
-
1.
|
| 39 |
],
|
| 40 |
[
|
| 41 |
-
1.
|
| 42 |
-
|
| 43 |
-
1.
|
| 44 |
-
|
| 45 |
]
|
| 46 |
],
|
| 47 |
"decoder": [
|
| 48 |
[
|
| 49 |
-
0.
|
| 50 |
-
0.
|
| 51 |
-
0.
|
| 52 |
-
0.
|
| 53 |
],
|
| 54 |
[
|
| 55 |
-
0.
|
| 56 |
-
0.
|
| 57 |
-
0.
|
| 58 |
-
0.
|
| 59 |
]
|
| 60 |
]
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"encoder": [
|
| 64 |
[
|
| 65 |
-
1.
|
| 66 |
-
1.
|
| 67 |
-
|
| 68 |
-
1.
|
| 69 |
],
|
| 70 |
[
|
| 71 |
-
1.
|
| 72 |
-
1.
|
| 73 |
-
1.
|
| 74 |
-
1.
|
| 75 |
]
|
| 76 |
],
|
| 77 |
"decoder": [
|
| 78 |
[
|
| 79 |
-
0.
|
| 80 |
-
0.
|
| 81 |
-
0.
|
| 82 |
-
0.
|
| 83 |
],
|
| 84 |
[
|
| 85 |
-
0.
|
| 86 |
-
0.
|
| 87 |
-
0.
|
| 88 |
-
0.
|
| 89 |
]
|
| 90 |
]
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"encoder": [
|
| 94 |
[
|
| 95 |
-
1.
|
| 96 |
-
1.
|
| 97 |
-
|
| 98 |
-
1.
|
| 99 |
],
|
| 100 |
[
|
| 101 |
-
1.
|
| 102 |
-
1.
|
| 103 |
-
1.
|
| 104 |
-
1.
|
| 105 |
]
|
| 106 |
],
|
| 107 |
"decoder": [
|
| 108 |
[
|
| 109 |
-
0.
|
| 110 |
-
0.
|
| 111 |
-
0.
|
| 112 |
-
0.
|
| 113 |
],
|
| 114 |
[
|
| 115 |
-
0.
|
| 116 |
-
0.
|
| 117 |
-
0.
|
| 118 |
-
0.
|
| 119 |
]
|
| 120 |
]
|
| 121 |
}
|
|
@@ -124,95 +124,95 @@
|
|
| 124 |
{
|
| 125 |
"encoder": [
|
| 126 |
[
|
| 127 |
-
1.
|
| 128 |
-
1.
|
| 129 |
-
|
| 130 |
-
1.
|
| 131 |
],
|
| 132 |
[
|
| 133 |
-
|
| 134 |
-
|
| 135 |
-
1.
|
| 136 |
-
1.
|
| 137 |
]
|
| 138 |
],
|
| 139 |
"decoder": [
|
| 140 |
[
|
| 141 |
-
0.
|
| 142 |
-
0.
|
| 143 |
-
0.
|
| 144 |
-
0.
|
| 145 |
],
|
| 146 |
[
|
| 147 |
-
0.
|
| 148 |
-
0.
|
| 149 |
-
0.
|
| 150 |
-
0.
|
| 151 |
]
|
| 152 |
]
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"encoder": [
|
| 156 |
[
|
| 157 |
-
1.
|
| 158 |
-
1.
|
| 159 |
-
|
| 160 |
-
1.
|
| 161 |
],
|
| 162 |
[
|
| 163 |
-
1.
|
| 164 |
-
1.
|
| 165 |
-
1.
|
| 166 |
-
1.
|
| 167 |
]
|
| 168 |
],
|
| 169 |
"decoder": [
|
| 170 |
[
|
| 171 |
-
0.
|
| 172 |
-
0.
|
| 173 |
-
0.
|
| 174 |
-
0.
|
| 175 |
],
|
| 176 |
[
|
| 177 |
-
0.
|
| 178 |
-
0.
|
| 179 |
-
0.
|
| 180 |
-
0.
|
| 181 |
]
|
| 182 |
]
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"encoder": [
|
| 186 |
[
|
| 187 |
-
1.
|
| 188 |
-
1.
|
| 189 |
-
|
| 190 |
-
1.
|
| 191 |
],
|
| 192 |
[
|
| 193 |
-
1.
|
| 194 |
-
1.
|
| 195 |
-
1.
|
| 196 |
-
|
| 197 |
]
|
| 198 |
],
|
| 199 |
"decoder": [
|
| 200 |
[
|
| 201 |
-
0.
|
| 202 |
-
0.
|
| 203 |
-
0.
|
| 204 |
-
0.
|
| 205 |
],
|
| 206 |
[
|
| 207 |
-
0.
|
| 208 |
-
0.
|
| 209 |
-
0.
|
| 210 |
-
0.
|
| 211 |
]
|
| 212 |
]
|
| 213 |
}
|
| 214 |
],
|
| 215 |
-
"best_val_loss": 5.
|
| 216 |
"config": {
|
| 217 |
"tokenizer": "bert-base-uncased",
|
| 218 |
"max_seq_len": 128,
|
|
|
|
| 2 |
"history": [
|
| 3 |
{
|
| 4 |
"epoch": 1,
|
| 5 |
+
"train_loss": 6.783949644676193,
|
| 6 |
"train_aux_loss": 0.0,
|
| 7 |
+
"train_perplexity": 883.5515563058923,
|
| 8 |
+
"val_loss": 6.120485806811677,
|
| 9 |
"val_aux_loss": 0.0,
|
| 10 |
+
"val_perplexity": 455.08572455110965
|
| 11 |
},
|
| 12 |
{
|
| 13 |
"epoch": 2,
|
| 14 |
+
"train_loss": 5.72200544220322,
|
| 15 |
"train_aux_loss": 0.0,
|
| 16 |
+
"train_perplexity": 305.51700570051946,
|
| 17 |
+
"val_loss": 5.783777441596244,
|
| 18 |
"val_aux_loss": 0.0,
|
| 19 |
+
"val_perplexity": 324.984484631975
|
| 20 |
},
|
| 21 |
{
|
| 22 |
"epoch": 3,
|
| 23 |
+
"train_loss": 5.255330442014935,
|
| 24 |
"train_aux_loss": 0.0,
|
| 25 |
+
"train_perplexity": 191.58478305751783,
|
| 26 |
+
"val_loss": 5.606828254184237,
|
| 27 |
"val_aux_loss": 0.0,
|
| 28 |
+
"val_perplexity": 272.279266370978
|
| 29 |
}
|
| 30 |
],
|
| 31 |
"train_expert_usage": [
|
| 32 |
{
|
| 33 |
"encoder": [
|
| 34 |
[
|
| 35 |
+
1.2391363382339478,
|
| 36 |
+
1.0983222723007202,
|
| 37 |
+
1.0011056661605835,
|
| 38 |
+
1.2266035079956055
|
| 39 |
],
|
| 40 |
[
|
| 41 |
+
1.2305195331573486,
|
| 42 |
+
0.7604071497917175,
|
| 43 |
+
1.1566280126571655,
|
| 44 |
+
1.4176130294799805
|
| 45 |
]
|
| 46 |
],
|
| 47 |
"decoder": [
|
| 48 |
[
|
| 49 |
+
0.7768203616142273,
|
| 50 |
+
0.23009872436523438,
|
| 51 |
+
0.22161749005317688,
|
| 52 |
+
0.26393088698387146
|
| 53 |
],
|
| 54 |
[
|
| 55 |
+
0.07864928245544434,
|
| 56 |
+
0.7649437785148621,
|
| 57 |
+
0.16985277831554413,
|
| 58 |
+
0.4790216386318207
|
| 59 |
]
|
| 60 |
]
|
| 61 |
},
|
| 62 |
{
|
| 63 |
"encoder": [
|
| 64 |
[
|
| 65 |
+
1.2392860651016235,
|
| 66 |
+
1.1015535593032837,
|
| 67 |
+
0.9848208427429199,
|
| 68 |
+
1.2395071983337402
|
| 69 |
],
|
| 70 |
[
|
| 71 |
+
1.2409480810165405,
|
| 72 |
+
1.0997846126556396,
|
| 73 |
+
1.1543740034103394,
|
| 74 |
+
1.0700610876083374
|
| 75 |
]
|
| 76 |
],
|
| 77 |
"decoder": [
|
| 78 |
[
|
| 79 |
+
0.7814996838569641,
|
| 80 |
+
0.21621061861515045,
|
| 81 |
+
0.2551786005496979,
|
| 82 |
+
0.2495078146457672
|
| 83 |
],
|
| 84 |
[
|
| 85 |
+
0.15144944190979004,
|
| 86 |
+
0.5814668536186218,
|
| 87 |
+
0.28734877705574036,
|
| 88 |
+
0.4821316599845886
|
| 89 |
]
|
| 90 |
]
|
| 91 |
},
|
| 92 |
{
|
| 93 |
"encoder": [
|
| 94 |
[
|
| 95 |
+
1.2064954042434692,
|
| 96 |
+
1.114086389541626,
|
| 97 |
+
0.9930238723754883,
|
| 98 |
+
1.2515621185302734
|
| 99 |
],
|
| 100 |
[
|
| 101 |
+
1.157797932624817,
|
| 102 |
+
1.213179111480713,
|
| 103 |
+
1.1462993621826172,
|
| 104 |
+
1.0478914976119995
|
| 105 |
]
|
| 106 |
],
|
| 107 |
"decoder": [
|
| 108 |
[
|
| 109 |
+
0.7839962244033813,
|
| 110 |
+
0.2211395800113678,
|
| 111 |
+
0.2677685022354126,
|
| 112 |
+
0.2284652441740036
|
| 113 |
],
|
| 114 |
[
|
| 115 |
+
0.19157297909259796,
|
| 116 |
+
0.4467729926109314,
|
| 117 |
+
0.3304468095302582,
|
| 118 |
+
0.5325767397880554
|
| 119 |
]
|
| 120 |
]
|
| 121 |
}
|
|
|
|
| 124 |
{
|
| 125 |
"encoder": [
|
| 126 |
[
|
| 127 |
+
1.2822349071502686,
|
| 128 |
+
1.1881206035614014,
|
| 129 |
+
0.9720463156700134,
|
| 130 |
+
1.1625773906707764
|
| 131 |
],
|
| 132 |
[
|
| 133 |
+
1.4352425336837769,
|
| 134 |
+
0.8882213234901428,
|
| 135 |
+
1.248740792274475,
|
| 136 |
+
1.0327744483947754
|
| 137 |
]
|
| 138 |
],
|
| 139 |
"decoder": [
|
| 140 |
[
|
| 141 |
+
0.8861706852912903,
|
| 142 |
+
0.21258454024791718,
|
| 143 |
+
0.2510792911052704,
|
| 144 |
+
0.2656497359275818
|
| 145 |
],
|
| 146 |
[
|
| 147 |
+
0.07652179896831512,
|
| 148 |
+
0.5462296605110168,
|
| 149 |
+
0.2719815671443939,
|
| 150 |
+
0.720751166343689
|
| 151 |
]
|
| 152 |
]
|
| 153 |
},
|
| 154 |
{
|
| 155 |
"encoder": [
|
| 156 |
[
|
| 157 |
+
1.2491365671157837,
|
| 158 |
+
1.2312203645706177,
|
| 159 |
+
0.9285868406295776,
|
| 160 |
+
1.196035385131836
|
| 161 |
],
|
| 162 |
[
|
| 163 |
+
1.1820405721664429,
|
| 164 |
+
1.1934090852737427,
|
| 165 |
+
1.1655993461608887,
|
| 166 |
+
1.0639300346374512
|
| 167 |
]
|
| 168 |
],
|
| 169 |
"decoder": [
|
| 170 |
[
|
| 171 |
+
0.9253849387168884,
|
| 172 |
+
0.21632608771324158,
|
| 173 |
+
0.24233703315258026,
|
| 174 |
+
0.23143617808818817
|
| 175 |
],
|
| 176 |
[
|
| 177 |
+
0.14991365373134613,
|
| 178 |
+
0.43304792046546936,
|
| 179 |
+
0.35336020588874817,
|
| 180 |
+
0.6791624426841736
|
| 181 |
]
|
| 182 |
]
|
| 183 |
},
|
| 184 |
{
|
| 185 |
"encoder": [
|
| 186 |
[
|
| 187 |
+
1.2165778875350952,
|
| 188 |
+
1.2050294876098633,
|
| 189 |
+
0.9812203049659729,
|
| 190 |
+
1.2021514177322388
|
| 191 |
],
|
| 192 |
[
|
| 193 |
+
1.1700963973999023,
|
| 194 |
+
1.3727154731750488,
|
| 195 |
+
1.0938984155654907,
|
| 196 |
+
0.9682688117027283
|
| 197 |
]
|
| 198 |
],
|
| 199 |
"decoder": [
|
| 200 |
[
|
| 201 |
+
0.8936178088188171,
|
| 202 |
+
0.21582241356372833,
|
| 203 |
+
0.2633832097053528,
|
| 204 |
+
0.24266082048416138
|
| 205 |
],
|
| 206 |
[
|
| 207 |
+
0.15016549825668335,
|
| 208 |
+
0.48665276169776917,
|
| 209 |
+
0.4501367211341858,
|
| 210 |
+
0.5285292863845825
|
| 211 |
]
|
| 212 |
]
|
| 213 |
}
|
| 214 |
],
|
| 215 |
+
"best_val_loss": 5.606828254184237,
|
| 216 |
"config": {
|
| 217 |
"tokenizer": "bert-base-uncased",
|
| 218 |
"max_seq_len": 128,
|
model.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 85979282
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1cfe54bf46ce4b9250bb8e6372ceb5a2c5b18eb665e7d65cbf3648127f096699
|
| 3 |
size 85979282
|