Per Egil Kummervold commited on
Commit
ddcb9c0
·
1 Parent(s): 742aa05

First submit

Browse files
Files changed (2) hide show
  1. config.json +7 -6
  2. generation_config.json +3 -3
config.json CHANGED
@@ -1,4 +1,5 @@
1
  {
 
2
  "architectures": [
3
  "LlamaForCausalLM"
4
  ],
@@ -6,16 +7,16 @@
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 128000,
8
  "eos_token_id": 128001,
9
- "head_dim": 64,
10
  "hidden_act": "silu",
11
- "hidden_size": 2048,
12
  "initializer_range": 0.02,
13
  "intermediate_size": 8192,
14
  "max_position_embeddings": 131072,
15
  "mlp_bias": false,
16
  "model_type": "llama",
17
- "num_attention_heads": 32,
18
- "num_hidden_layers": 16,
19
  "num_key_value_heads": 8,
20
  "pretraining_tp": 1,
21
  "rms_norm_eps": 1e-05,
@@ -28,8 +29,8 @@
28
  },
29
  "rope_theta": 500000.0,
30
  "tie_word_embeddings": true,
31
- "torch_dtype": "bfloat16",
32
- "transformers_version": "4.45.0.dev0",
33
  "use_cache": true,
34
  "vocab_size": 128256
35
  }
 
1
  {
2
+ "_name_or_path": "meta-llama/Llama-3.2-3B",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
 
7
  "attention_dropout": 0.0,
8
  "bos_token_id": 128000,
9
  "eos_token_id": 128001,
10
+ "head_dim": 128,
11
  "hidden_act": "silu",
12
+ "hidden_size": 3072,
13
  "initializer_range": 0.02,
14
  "intermediate_size": 8192,
15
  "max_position_embeddings": 131072,
16
  "mlp_bias": false,
17
  "model_type": "llama",
18
+ "num_attention_heads": 24,
19
+ "num_hidden_layers": 28,
20
  "num_key_value_heads": 8,
21
  "pretraining_tp": 1,
22
  "rms_norm_eps": 1e-05,
 
29
  },
30
  "rope_theta": 500000.0,
31
  "tie_word_embeddings": true,
32
+ "torch_dtype": "float32",
33
+ "transformers_version": "4.48.0",
34
  "use_cache": true,
35
  "vocab_size": 128256
36
  }
generation_config.json CHANGED
@@ -1,9 +1,9 @@
1
  {
2
  "_from_model_config": true,
3
  "bos_token_id": 128000,
4
- "eos_token_id": 128001,
5
- "transformers_version": "4.45.0.dev0",
6
  "do_sample": true,
 
7
  "temperature": 0.6,
8
- "top_p": 0.9
 
9
  }
 
1
  {
2
  "_from_model_config": true,
3
  "bos_token_id": 128000,
 
 
4
  "do_sample": true,
5
+ "eos_token_id": 128001,
6
  "temperature": 0.6,
7
+ "top_p": 0.9,
8
+ "transformers_version": "4.48.0"
9
  }