Upload 5 files
Browse files- xsmall/README.md +1 -0
- xsmall/discriminator.bin +3 -0
- xsmall/generator.bin +3 -0
- xsmall/model_config.json +61 -0
- xsmall/training_RTD.log +0 -0
xsmall/README.md
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
Bản xmall đc train trên 30GB text
|
xsmall/discriminator.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:977b70cb677b737ab9d35e1e1c9e0c9730a617ad7bfc4e693ac7f0776014ef42
|
| 3 |
+
size 81059807
|
xsmall/generator.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9be0bbd57a269c079f56fc98bd7b62b4791067e700170181e4e450cea1742d85
|
| 3 |
+
size 40924839
|
xsmall/model_config.json
ADDED
|
@@ -0,0 +1,61 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"attention_probs_dropout_prob": 0.1,
|
| 3 |
+
"discriminator": {
|
| 4 |
+
"attention_head_size": 64,
|
| 5 |
+
"attention_probs_dropout_prob": 0.1,
|
| 6 |
+
"hidden_act": "gelu",
|
| 7 |
+
"hidden_dropout_prob": 0.1,
|
| 8 |
+
"hidden_size": 384,
|
| 9 |
+
"initializer_range": 0.02,
|
| 10 |
+
"intermediate_size": 1536,
|
| 11 |
+
"layer_norm_eps": 1e-07,
|
| 12 |
+
"max_position_embeddings": 512,
|
| 13 |
+
"max_relative_positions": -1,
|
| 14 |
+
"norm_rel_ebd": "layer_norm",
|
| 15 |
+
"num_attention_heads": 6,
|
| 16 |
+
"num_hidden_layers": 12,
|
| 17 |
+
"pos_att_type": "p2c|c2p",
|
| 18 |
+
"position_biased_input": false,
|
| 19 |
+
"position_buckets": 256,
|
| 20 |
+
"relative_attention": true,
|
| 21 |
+
"share_att_key": true,
|
| 22 |
+
"type_vocab_size": 0,
|
| 23 |
+
"vocab_size": 24064
|
| 24 |
+
},
|
| 25 |
+
"embedding_sharing": "gdes",
|
| 26 |
+
"generator": {
|
| 27 |
+
"attention_head_size": 64,
|
| 28 |
+
"attention_probs_dropout_prob": 0.1,
|
| 29 |
+
"hidden_act": "gelu",
|
| 30 |
+
"hidden_dropout_prob": 0.1,
|
| 31 |
+
"hidden_size": 384,
|
| 32 |
+
"initializer_range": 0.02,
|
| 33 |
+
"intermediate_size": 1536,
|
| 34 |
+
"layer_norm_eps": 1e-07,
|
| 35 |
+
"max_position_embeddings": 512,
|
| 36 |
+
"max_relative_positions": -1,
|
| 37 |
+
"norm_rel_ebd": "layer_norm",
|
| 38 |
+
"num_attention_heads": 6,
|
| 39 |
+
"num_hidden_layers": 6,
|
| 40 |
+
"pos_att_type": "p2c|c2p",
|
| 41 |
+
"position_biased_input": false,
|
| 42 |
+
"position_buckets": 256,
|
| 43 |
+
"relative_attention": true,
|
| 44 |
+
"share_att_key": true,
|
| 45 |
+
"type_vocab_size": 0,
|
| 46 |
+
"vocab_size": 24064
|
| 47 |
+
},
|
| 48 |
+
"hidden_act": "gelu",
|
| 49 |
+
"hidden_dropout_prob": 0.1,
|
| 50 |
+
"hidden_size": 768,
|
| 51 |
+
"initializer_range": 0.02,
|
| 52 |
+
"intermediate_size": 3072,
|
| 53 |
+
"layer_norm_eps": 1e-07,
|
| 54 |
+
"max_position_embeddings": 512,
|
| 55 |
+
"num_attention_heads": 12,
|
| 56 |
+
"num_hidden_layers": 12,
|
| 57 |
+
"padding_idx": 0,
|
| 58 |
+
"type_vocab_size": 0,
|
| 59 |
+
"vocab_size": -1
|
| 60 |
+
}
|
| 61 |
+
|
xsmall/training_RTD.log
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|