Happyb commited on
Commit
03d205d
1 Parent(s): 53eea9a

Upload config.json

Browse files
Files changed (1) hide show
  1. config.json +3 -4
config.json CHANGED
@@ -1,13 +1,12 @@
1
  {
2
- "_name_or_path": "/scratch/gpfs/awettig/wura/checkpoints/lcft_Llama-3.1-8B_datasets_wura_unimax_1.0+edu_bsz512_steps2400_lr1e-5_warmup0.1",
3
  "architectures": [
4
- "LlamaForSequenceClassification"
5
  ],
6
  "attention_bias": false,
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": 4096,
13
  "initializer_range": 0.02,
@@ -31,7 +30,7 @@
31
  "rope_theta": 500000.0,
32
  "tie_word_embeddings": false,
33
  "torch_dtype": "float32",
34
- "transformers_version": "4.47.0",
35
  "use_cache": true,
36
  "vocab_size": 128256
37
  }
 
1
  {
2
+ "_name_or_path": "/home/awettig/pli/models/meta-llama/Llama-3.1-8B",
3
  "architectures": [
4
+ "LlamaForCausalLM"
5
  ],
6
  "attention_bias": false,
7
  "attention_dropout": 0.0,
8
  "bos_token_id": 128000,
9
  "eos_token_id": 128001,
 
10
  "hidden_act": "silu",
11
  "hidden_size": 4096,
12
  "initializer_range": 0.02,
 
30
  "rope_theta": 500000.0,
31
  "tie_word_embeddings": false,
32
  "torch_dtype": "float32",
33
+ "transformers_version": "4.44.2",
34
  "use_cache": true,
35
  "vocab_size": 128256
36
  }