Hugging Face's logo

SmallDoge
/
Doge-40M-MoE

Text Generation
Transformers
Safetensors
doge
conversational
Model card Files
xet
Community
JingzeShi commited on
Commit
9fe67a9
·
verified ·
1 Parent(s): 3a5380c

Upload DogeForCausalLM

Browse files
Files changed (3) hide show
  1. config.json +37 -34
  2. generation_config.json +7 -4
  3. model.safetensors +2 -2
config.json CHANGED
@@ -1,34 +1,37 @@
1
- {
2
- "architectures": [
3
- "DogeForCausalLM"
4
- ],
5
- "attention_bias": false,
6
- "attention_dropout": 0.0,
7
- "hidden_act": "silu",
8
- "hidden_dropout": 0.0,
9
- "hidden_size": 256,
10
- "initializer_range": 0.02,
11
- "intermediate_size": 768,
12
- "is_moe": true,
13
- "keep_window_size": 2048,
14
- "max_position_embeddings": 2048,
15
- "mlp_bias": false,
16
- "model_type": "doge",
17
- "norm_topk_prob": false,
18
- "num_attention_heads": 2,
19
- "num_experts": 1024,
20
- "num_experts_per_tok": 8,
21
- "num_hidden_layers": 8,
22
- "num_key_value_heads": 1,
23
- "output_router_logits": true,
24
- "rms_norm_eps": 1e-06,
25
- "rope_scaling": null,
26
- "rope_theta": 10000.0,
27
- "router_aux_loss_coef": 0.001,
28
- "sliding_window": null,
29
- "tie_word_embeddings": true,
30
- "torch_dtype": "bfloat16",
31
- "transformers_version": "4.54.0.dev0",
32
- "use_cache": true,
33
- "vocab_size": 128256
34
- }
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DogeForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 128000,
8
+ "eos_token_id": 128001,
9
+ "hidden_act": "silu",
10
+ "hidden_dropout": 0.0,
11
+ "hidden_size": 256,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 1024,
14
+ "is_moe": true,
15
+ "keep_window_size": 4096,
16
+ "max_position_embeddings": 4096,
17
+ "mlp_bias": false,
18
+ "model_type": "doge",
19
+ "norm_topk_prob": false,
20
+ "num_attention_heads": 2,
21
+ "num_experts": 1024,
22
+ "num_experts_per_tok": 8,
23
+ "num_hidden_layers": 8,
24
+ "num_key_value_heads": 1,
25
+ "output_router_logits": false,
26
+ "pad_token_id": 128001,
27
+ "rms_norm_eps": 1e-06,
28
+ "rope_scaling": null,
29
+ "rope_theta": 40000.0,
30
+ "router_aux_loss_coef": 0.001,
31
+ "sliding_window": null,
32
+ "tie_word_embeddings": true,
33
+ "torch_dtype": "bfloat16",
34
+ "transformers_version": "4.55.0",
35
+ "use_cache": true,
36
+ "vocab_size": 128256
37
+ }
generation_config.json CHANGED
@@ -1,4 +1,7 @@
1
- {
2
- "_from_model_config": true,
3
- "transformers_version": "4.54.0.dev0"
4
- }
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 128000,
4
+ "eos_token_id": 128001,
5
+ "pad_token_id": 128001,
6
+ "transformers_version": "4.55.0"
7
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f672b60a5a0fd68b76383d5f9f0cc30911c6388d723a769415780f9bfdea394a
3
- size 86939504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2a9aa93b11ae0470af3f218b25c003fd82d2b1fd252bd8372ad7fc9cb769b3e3
3
+ size 90085256