vasimakram01 commited on
Commit
384941f
1 Parent(s): b96fb94

Create config.json

Browse files
Files changed (1) hide show
  1. config.json +38 -0
config.json ADDED
@@ -0,0 +1,38 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ { "_name_or_path": "tiiuae/falcon-7b",
2
+ "alibi": false,
3
+ "apply_residual_connection_post_layernorm": false,
4
+ "architectures": [ "RWForCausalLM" ],
5
+ "attention_dropout": 0.0,
6
+ "auto_map": {
7
+ "AutoConfig": "tiiuae/falcon-7b--configuration_RW.RWConfig",
8
+ "AutoModel": "tiiuae/falcon-7b--modelling_RW.RWModel",
9
+ "AutoModelForCausalLM": "tiiuae/falcon-7b--modelling_RW.RWForCausalLM",
10
+ "AutoModelForQuestionAnswering": "tiiuae/falcon-7b--modelling_RW.RWForQuestionAnswering",
11
+ "AutoModelForSequenceClassification": "tiiuae/falcon-7b--modelling_RW.RWForSequenceClassification",
12
+ "AutoModelForTokenClassification": "tiiuae/falcon-7b--modelling_RW.RWForTokenClassification" },
13
+ "bias": false,
14
+ "bos_token_id": 11,
15
+ "eos_token_id": 11,
16
+ "hidden_dropout": 0.0,
17
+ "hidden_size": 4544,
18
+ "initializer_range": 0.02,
19
+ "layer_norm_epsilon": 1e-05,
20
+ "model_type": "RefinedWebModel",
21
+ "multi_query": true,
22
+ "n_head": 71,
23
+ "n_layer": 32,
24
+ "parallel_attn": true,
25
+ "quantization_config": {
26
+ "bnb_4bit_compute_dtype": "bfloat16",
27
+ "bnb_4bit_quant_type": "nf4",
28
+ "bnb_4bit_use_double_quant": true,
29
+ "llm_int8_enable_fp32_cpu_offload": false,
30
+ "llm_int8_has_fp16_weight": false,
31
+ "llm_int8_skip_modules": null,
32
+ "llm_int8_threshold": 6.0,
33
+ "load_in_4bit": true,
34
+ "load_in_8bit": false },
35
+ "torch_dtype": "bfloat16",
36
+ "transformers_version": "4.30.0.dev0",
37
+ "use_cache": true,
38
+ "vocab_size": 65024 }