Commit
•
40cc034
1
Parent(s):
6e4991c
Upload SwitchTransformersForConditionalGeneration
Browse files- config.json +43 -0
- generation_config.json +8 -0
- pytorch_model.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/marpha/research/switch-ul2-checkpoints/checkpoint-114500",
|
| 3 |
+
"add_router_probs": false,
|
| 4 |
+
"architectures": [
|
| 5 |
+
"SwitchTransformersForConditionalGeneration"
|
| 6 |
+
],
|
| 7 |
+
"d_ff": 2048,
|
| 8 |
+
"d_kv": 64,
|
| 9 |
+
"d_model": 768,
|
| 10 |
+
"decoder_sparse_step": 4,
|
| 11 |
+
"decoder_start_token_id": 0,
|
| 12 |
+
"dense_act_fn": "relu",
|
| 13 |
+
"dropout_rate": 0.1,
|
| 14 |
+
"encoder_sparse_step": 4,
|
| 15 |
+
"eos_token_id": 1,
|
| 16 |
+
"expert_capacity": 64,
|
| 17 |
+
"feed_forward_proj": "relu",
|
| 18 |
+
"initializer_factor": 1.0,
|
| 19 |
+
"is_encoder_decoder": true,
|
| 20 |
+
"is_gated_act": false,
|
| 21 |
+
"layer_norm_epsilon": 1e-06,
|
| 22 |
+
"max_length": 512,
|
| 23 |
+
"model_type": "switch_transformers",
|
| 24 |
+
"num_decoder_layers": 12,
|
| 25 |
+
"num_experts": 8,
|
| 26 |
+
"num_heads": 12,
|
| 27 |
+
"num_layers": 12,
|
| 28 |
+
"num_sparse_decoder_layers": 3,
|
| 29 |
+
"num_sparse_encoder_layers": 3,
|
| 30 |
+
"pad_token_id": 0,
|
| 31 |
+
"relative_attention_max_distance": 128,
|
| 32 |
+
"relative_attention_num_buckets": 32,
|
| 33 |
+
"router_aux_loss_coef": 0.001,
|
| 34 |
+
"router_bias": false,
|
| 35 |
+
"router_dtype": "float32",
|
| 36 |
+
"router_ignore_padding_tokens": false,
|
| 37 |
+
"router_jitter_noise": 0.01,
|
| 38 |
+
"router_z_loss_coef": 0.001,
|
| 39 |
+
"torch_dtype": "float32",
|
| 40 |
+
"transformers_version": "4.28.1",
|
| 41 |
+
"use_cache": true,
|
| 42 |
+
"vocab_size": 100000
|
| 43 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"decoder_start_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"max_length": 512,
|
| 6 |
+
"pad_token_id": 0,
|
| 7 |
+
"transformers_version": "4.28.1"
|
| 8 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8486e65754f057176bf19157a71e26d14b23f73bce254b23fe5aaf75370f6ae4
|
| 3 |
+
size 1477867321
|