Update config.json
Browse files- config.json +51 -0
config.json
ADDED
|
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"ReformerModelWithLMHead"
|
| 4 |
+
],
|
| 5 |
+
"attention_head_size": 64,
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"attn_layers": [
|
| 8 |
+
"local",
|
| 9 |
+
"lsh",
|
| 10 |
+
"local",
|
| 11 |
+
"lsh",
|
| 12 |
+
"local",
|
| 13 |
+
"lsh"
|
| 14 |
+
],
|
| 15 |
+
"axial_norm_std": 1.0,
|
| 16 |
+
"axial_pos_embds": true,
|
| 17 |
+
"axial_pos_embds_dim": [
|
| 18 |
+
64,
|
| 19 |
+
192
|
| 20 |
+
],
|
| 21 |
+
"axial_pos_shape": [
|
| 22 |
+
512,
|
| 23 |
+
1024
|
| 24 |
+
],
|
| 25 |
+
"chunk_size_feed_forward": 0,
|
| 26 |
+
"chunk_size_lm_head": 0,
|
| 27 |
+
"feed_forward_size": 512,
|
| 28 |
+
"hidden_act": "relu",
|
| 29 |
+
"hidden_dropout_prob": 0.05,
|
| 30 |
+
"hidden_size": 256,
|
| 31 |
+
"initializer_range": 0.02,
|
| 32 |
+
"is_decoder": true,
|
| 33 |
+
"layer_norm_eps": 1e-12,
|
| 34 |
+
"local_attn_chunk_length": 64,
|
| 35 |
+
"lsh_attn_chunk_length": 64,
|
| 36 |
+
"max_position_embeddings": 524288,
|
| 37 |
+
"model_type": "reformer",
|
| 38 |
+
"num_attention_heads": 2,
|
| 39 |
+
"num_buckets": [
|
| 40 |
+
64,
|
| 41 |
+
128
|
| 42 |
+
],
|
| 43 |
+
"num_chunks_after": 0,
|
| 44 |
+
"num_chunks_before": 1,
|
| 45 |
+
"num_hashes": 1,
|
| 46 |
+
"num_hidden_layers": 6,
|
| 47 |
+
"pad_token_id": 0,
|
| 48 |
+
"seed": null,
|
| 49 |
+
"sinusoidal_pos_embds": false,
|
| 50 |
+
"vocab_size": 320
|
| 51 |
+
}
|