Hugging Face's logo

ISTA-DASLab
/
switch-large-128_qmoe

Text2Text Generation
Transformers
switch_transformers
Inference Endpoints
Model card Files Community
efrantar commited on
Commit
d1d5026
•
1 Parent(s): c4cd3d9

Added files

Browse files
Files changed (27) hide show
  1. config.json +41 -0
  2. exp00.pt +3 -0
  3. exp01.pt +3 -0
  4. exp02.pt +3 -0
  5. exp03.pt +3 -0
  6. exp04.pt +3 -0
  7. exp05.pt +3 -0
  8. exp06.pt +3 -0
  9. exp07.pt +3 -0
  10. exp08.pt +3 -0
  11. exp09.pt +3 -0
  12. exp10.pt +3 -0
  13. exp11.pt +3 -0
  14. exp12.pt +3 -0
  15. exp13.pt +3 -0
  16. exp14.pt +3 -0
  17. exp15.pt +3 -0
  18. exp16.pt +3 -0
  19. exp17.pt +3 -0
  20. exp18.pt +3 -0
  21. exp19.pt +3 -0
  22. exp20.pt +3 -0
  23. exp21.pt +3 -0
  24. exp22.pt +3 -0
  25. exp23.pt +3 -0
  26. noexp.pt +3 -0
  27. sizes.pt +3 -0
config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_router_probs": false,
3
+ "architectures": [
4
+ "SwitchTransformersForConditionalGeneration"
5
+ ],
6
+ "d_ff": 4096,
7
+ "d_kv": 64,
8
+ "d_model": 1024,
9
+ "decoder_sparse_step": 2,
10
+ "dense_act_fn": "relu",
11
+ "dropout_rate": 0.1,
12
+ "encoder_sparse_step": 2,
13
+ "eos_token_id": 1,
14
+ "expert_capacity": 1024,
15
+ "feed_forward_proj": "relu",
16
+ "initializer_factor": 1.0,
17
+ "is_encoder_decoder": true,
18
+ "is_gated_act": false,
19
+ "layer_norm_epsilon": 1e-06,
20
+ "model_type": "switch_transformers",
21
+ "num_decoder_layers": 24,
22
+ "num_experts": 128,
23
+ "num_heads": 16,
24
+ "num_layers": 24,
25
+ "num_sparse_decoder_layers": 12,
26
+ "num_sparse_encoder_layers": 12,
27
+ "pad_token_id": 0,
28
+ "relative_attention_max_distance": 128,
29
+ "relative_attention_num_buckets": 32,
30
+ "router_aux_loss_coef": 0.001,
31
+ "router_bias": false,
32
+ "router_dtype": "float32",
33
+ "router_ignore_padding_tokens": false,
34
+ "router_jitter_noise": 0,
35
+ "router_type": "tokens_masked",
36
+ "router_z_loss_coef": 0.001,
37
+ "torch_dtype": "bfloat16",
38
+ "transformers_version": "4.28.0.dev0",
39
+ "use_cache": false,
40
+ "vocab_size": 32128
41
+ }
exp00.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dbabe7948e8c1ec63c763d0913f79acf4287305e409e8eabdd433565aeff1db
3
+ size 120733461
exp01.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7838b419802b53da2fac0365391619951985e85ab21beda5899f25a7aaef9dfb
3
+ size 114955349
exp02.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:158c70d2c5cbd3d38096fc98b65a20f8aadb02ecad40b47109a52460cb24f038
3
+ size 113555797
exp03.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f13213d041e2d392e18ef4db3c4d6001421030c81aeb9b8dbba328be9997a643
3
+ size 115741845
exp04.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6ec980e1087804d9ae1102abffa3ac8e7a406427abfe68644e8444171bdf61d
3
+ size 112386837
exp05.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:16ecd33ade38af93799e0de313c4ff8e41b3f9362542c37fbe3fa14bd503fedb
3
+ size 113111829
exp06.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a35f371a3067daaecf8f06e8968f9df8ed6f4d9e1ef22055c9a153d69931e32f
3
+ size 113632149
exp07.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:446e8e97737fd3feb7af8b71298298dc5a295ac700e85d3599ff662df67b8d92
3
+ size 113846421
exp08.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:867552450dc56082f4e1b31e26e650171eab690e6d3b53bd909504403e15236a
3
+ size 114498133
exp09.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c3c7085aa42e6500bcf34b22fd06d23b12afc22fdab396d0a9c1f9be5cc513dd
3
+ size 114390037
exp10.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11ba602874dd6732dd93b46dadeff39040c754cd00ccd57a3bf285a49194e3a1
3
+ size 117299093
exp11.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3465a36c18cf0fb68026d9b611062c047be0d89e8a3af2fc083ae6ffb9718847
3
+ size 117312149
exp12.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c6ca1e637f87b9e8dc14aae3f767c5293f5d695275d33d8093f0a60e5a7f2bd
3
+ size 114166165
exp13.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a6bbf58d465ac9d90576870b8bf24f4d2eeadf49a076d380e9e6a7e6b516ff6
3
+ size 124696149
exp14.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7808a4494561851f84e3d1d2586814d19c642cc85cceadadf6c301cfad812571
3
+ size 120507861
exp15.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa92e00d24a17219233c024d40e2c1989871844d34c3dbf78f02e15d5b82b1c0
3
+ size 116771669
exp16.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ded7535a66c2cf267ffc87b40b1ab3567c7331dc81e768b5461b0830a1b65f8
3
+ size 118875989
exp17.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:247ef84448d9d4c128a7d578c2f6332ba9bd81446f6776233952a5d0f5c91eb3
3
+ size 120403605
exp18.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5182a89a334198e1c709dac0155d6aa9288fa359e1ec9c998b3fd0c291e86a8e
3
+ size 120391189
exp19.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f015a0a1843d7654f5804b533e5e3bce30fa28b0eaf4d4780b6755ce23d3f4d2
3
+ size 120465749
exp20.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c0ccee11dc80d0811cec1faeaeb0ad850ce14a1715b2f1b15a51b74bc107006
3
+ size 120861461
exp21.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6597d1d5d12fc14318047493fbca60c6860fe478349566523395ed42633dcfc0
3
+ size 120224341
exp22.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1ed1555791d8379278182a4180888af3e6070a2caa47fadcd2030e1789cbc13
3
+ size 119406165
exp23.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89a6ab25653edb978724af0c2e607038aa4f8279f1784e52ae4feb7a774d784b
3
+ size 117228245
noexp.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3e9dbe5a22e6b3ccade4820605aa43e414b5ebf61a0454c4645b3a74bc735eda
3
+ size 1146084069
sizes.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:410486944077a8cee924e2bef3f542c93aa2a234ea302c20b8968e03ad8a4d83
3
+ size 391147