mpasila commited on
Commit
147d8b1
·
verified ·
1 Parent(s): 38e2e05

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +5 -1
config.json CHANGED
@@ -6,22 +6,26 @@
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 1,
8
  "eos_token_id": 2,
 
9
  "hidden_act": "silu",
10
  "hidden_size": 4096,
11
  "initializer_range": 0.02,
12
  "intermediate_size": 12288,
13
  "max_position_embeddings": 4096,
 
14
  "model_type": "llama",
15
  "num_attention_heads": 32,
16
  "num_hidden_layers": 42,
17
  "num_key_value_heads": 8,
 
18
  "pretraining_tp": 1,
19
  "rms_norm_eps": 1e-05,
20
  "rope_scaling": null,
21
  "rope_theta": 10000.0,
22
  "tie_word_embeddings": false,
23
  "torch_dtype": "bfloat16",
24
- "transformers_version": "4.40.1",
 
25
  "use_cache": true,
26
  "vocab_size": 128000
27
  }
 
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 1,
8
  "eos_token_id": 2,
9
+ "head_dim": 128,
10
  "hidden_act": "silu",
11
  "hidden_size": 4096,
12
  "initializer_range": 0.02,
13
  "intermediate_size": 12288,
14
  "max_position_embeddings": 4096,
15
+ "mlp_bias": false,
16
  "model_type": "llama",
17
  "num_attention_heads": 32,
18
  "num_hidden_layers": 42,
19
  "num_key_value_heads": 8,
20
+ "pad_token_id": 0,
21
  "pretraining_tp": 1,
22
  "rms_norm_eps": 1e-05,
23
  "rope_scaling": null,
24
  "rope_theta": 10000.0,
25
  "tie_word_embeddings": false,
26
  "torch_dtype": "bfloat16",
27
+ "transformers_version": "4.51.3",
28
+ "unsloth_version": "2025.4.3",
29
  "use_cache": true,
30
  "vocab_size": 128000
31
  }