luccafong commited on
Commit
3c6dd49
·
verified ·
1 Parent(s): 78f400f

Upload config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. config.json +70 -74
config.json CHANGED
@@ -1,76 +1,72 @@
1
  {
2
-
3
- "model_type": "deepseek_mtp",
4
- "model": {
5
- "model_type": "deepseek_v3",
6
- "architectures": [
7
- "DeepseekV3ForCausalLM"
8
- ],
9
- "attention_bias": false,
10
- "attention_dropout": 0.0,
11
- "auto_map": {
12
- "AutoConfig": "configuration_deepseek.DeepseekV3Config",
13
- "AutoModel": "modeling_deepseek.DeepseekV3Model",
14
- "AutoModelForCausalLM": "modeling_deepseek.DeepseekV3ForCausalLM"
15
- },
16
- "aux_loss_alpha": 0.001,
17
- "bos_token_id": 100000,
18
- "eos_token_id": 100001,
19
- "ep_size": 1,
20
- "first_k_dense_replace": 1,
21
- "hidden_act": "silu",
22
- "hidden_size": 2560,
23
- "initializer_range": 0.02,
24
- "intermediate_size": 12288,
25
- "kv_lora_rank": 512,
26
- "lm_head": false,
27
- "max_position_embeddings": 4096,
28
- "moe_intermediate_size": 1536,
29
- "moe_layer_freq": 1,
30
- "n_group": 1,
31
- "n_routed_experts": 72,
32
- "n_shared_experts": 2,
33
- "norm_topk_prob": true,
34
- "num_attention_heads": 32,
35
- "num_experts_per_tok": 6,
36
- "num_hidden_layers": 30,
37
- "num_key_value_heads": 32,
38
- "num_nextn_predict_layers": 3,
39
- "pretraining_tp": 1,
40
- "q_lora_rank": null,
41
- "qk_nope_head_dim": 128,
42
- "qk_rope_head_dim": 64,
43
- "quantization_config": {
44
- "activation_scheme": "dynamic",
45
- "fmt": "e4m3",
46
- "quant_method": "fp8",
47
- "weight_block_size": [
48
- 128,
49
- 128
50
- ]
51
- },
52
- "rm_head": false,
53
- "rms_norm_eps": 1e-06,
54
- "rope_scaling": {
55
- "beta_fast": 32,
56
- "beta_slow": 1,
57
- "factor": 40,
58
- "mscale": 1.0,
59
- "mscale_all_dim": 1.0,
60
- "original_max_position_embeddings": 4096,
61
- "type": "yarn"
62
- },
63
- "rope_theta": 10000,
64
- "routed_scaling_factor": 2.0,
65
- "scoring_func": "sigmoid",
66
- "seq_aux": true,
67
- "tie_word_embeddings": false,
68
- "topk_group": 1,
69
- "topk_method": "noaux_tc",
70
- "torch_dtype": "bfloat16",
71
- "transformers_version": "4.33.1",
72
- "use_cache": true,
73
- "v_head_dim": 128,
74
- "vocab_size": 129280
75
- }
76
  }
 
1
  {
2
+ "model_type": "deepseek_v3",
3
+ "architectures": [
4
+ "DeepseekV3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "auto_map": {
9
+ "AutoConfig": "configuration_deepseek.DeepseekV3Config",
10
+ "AutoModel": "modeling_deepseek.DeepseekV3Model",
11
+ "AutoModelForCausalLM": "modeling_deepseek.DeepseekV3ForCausalLM"
12
+ },
13
+ "aux_loss_alpha": 0.001,
14
+ "bos_token_id": 100000,
15
+ "eos_token_id": 100001,
16
+ "ep_size": 1,
17
+ "first_k_dense_replace": 1,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 2560,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 12288,
22
+ "kv_lora_rank": 512,
23
+ "lm_head": false,
24
+ "max_position_embeddings": 4096,
25
+ "moe_intermediate_size": 1536,
26
+ "moe_layer_freq": 1,
27
+ "n_group": 1,
28
+ "n_routed_experts": 72,
29
+ "n_shared_experts": 2,
30
+ "norm_topk_prob": true,
31
+ "num_attention_heads": 32,
32
+ "num_experts_per_tok": 6,
33
+ "num_hidden_layers": 30,
34
+ "num_key_value_heads": 32,
35
+ "num_nextn_predict_layers": 3,
36
+ "pretraining_tp": 1,
37
+ "q_lora_rank": null,
38
+ "qk_nope_head_dim": 128,
39
+ "qk_rope_head_dim": 64,
40
+ "quantization_config": {
41
+ "activation_scheme": "dynamic",
42
+ "fmt": "e4m3",
43
+ "quant_method": "fp8",
44
+ "weight_block_size": [
45
+ 128,
46
+ 128
47
+ ]
48
+ },
49
+ "rm_head": false,
50
+ "rms_norm_eps": 1e-06,
51
+ "rope_scaling": {
52
+ "beta_fast": 32,
53
+ "beta_slow": 1,
54
+ "factor": 40,
55
+ "mscale": 1.0,
56
+ "mscale_all_dim": 1.0,
57
+ "original_max_position_embeddings": 4096,
58
+ "type": "yarn"
59
+ },
60
+ "rope_theta": 10000,
61
+ "routed_scaling_factor": 2.0,
62
+ "scoring_func": "sigmoid",
63
+ "seq_aux": true,
64
+ "tie_word_embeddings": false,
65
+ "topk_group": 1,
66
+ "topk_method": "noaux_tc",
67
+ "torch_dtype": "bfloat16",
68
+ "transformers_version": "4.33.1",
69
+ "use_cache": true,
70
+ "v_head_dim": 128,
71
+ "vocab_size": 129280
 
 
 
 
72
  }