milmor commited on
Commit
a8443f8
·
1 Parent(s): d2ed733

Upload 5 files

Browse files
.gitattributes CHANGED
@@ -32,3 +32,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
32
  *.zip filter=lfs diff=lfs merge=lfs -text
33
  *.zst filter=lfs diff=lfs merge=lfs -text
34
  *tfevents* filter=lfs diff=lfs merge=lfs -text
35
+ openwt_512_d_512/best-ckpt/ckpt-616000.data-00000-of-00001 filter=lfs diff=lfs merge=lfs -text
openwt_512_d_512/best-ckpt/checkpoint ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ model_checkpoint_path: "ckpt-616000"
2
+ all_model_checkpoint_paths: "ckpt-594000"
3
+ all_model_checkpoint_paths: "ckpt-600000"
4
+ all_model_checkpoint_paths: "ckpt-616000"
5
+ all_model_checkpoint_timestamps: 1686872841.3177085
6
+ all_model_checkpoint_timestamps: 1686874219.3603818
7
+ all_model_checkpoint_timestamps: 1686877883.841565
8
+ last_preserved_timestamp: 1686731001.6954174
openwt_512_d_512/best-ckpt/ckpt-616000.data-00000-of-00001 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f571c03e942882fd9ebc837b05a8739d6879e07d6f3a1dd19169187a832b7889
3
+ size 810756746
openwt_512_d_512/best-ckpt/ckpt-616000.index ADDED
Binary file (30.8 kB). View file
 
openwt_512_d_512/log-dir/events.out.tfevents.1686731002.emam-beta.162770.0.v2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e3ab9a4f89a91d9c36e4386d67b3062e6cd054779b4ccc143fb7baa0525efb3
3
+ size 42704
openwt_512_d_512/openwt_512_d_512_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"batch_size": 16, "buffer_size": 40000, "shuffle_seed": 32, "vocab_file": "wiki_en_vocab", "min_seq_len": false, "ckpt_interval": 2000, "val_steps": 1000, "train_size": 95, "vocab_size": 50257, "seq_len": 512, "learning_rate": 0.001, "beta_1": 0.9, "beta_2": 0.95, "decay_lr": false, "decay_steps": 400000, "alpha": 0.1, "emb_dim": 512, "heads": 8, "mlp_dim": 512, "depth": 10, "dropout": 0.0, "initializer": "glorot_uniform", "embedding_initializer": "glorot_uniform", "eps": 1e-06, "mlp_activation": "gelu"}