akumar23 commited on
Commit
97f0f03
1 Parent(s): 85ff63c

Upload config.json

Browse files
Files changed (1) hide show
  1. config.json +50 -0
config.json ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "vilsonrodrigues/falcon-7b-instruct-sharded",
3
+ "alibi": false,
4
+ "apply_residual_connection_post_layernorm": false,
5
+ "architectures": [
6
+ "FalconForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "auto_map": {
10
+ "AutoConfig": "vilsonrodrigues/falcon-7b-instruct-sharded--configuration_falcon.FalconConfig",
11
+ "AutoModel": "vilsonrodrigues/falcon-7b-instruct-sharded--modeling_falcon.FalconModel",
12
+ "AutoModelForCausalLM": "vilsonrodrigues/falcon-7b-instruct-sharded--modeling_falcon.FalconForCausalLM",
13
+ "AutoModelForQuestionAnswering": "vilsonrodrigues/falcon-7b-instruct-sharded--modeling_falcon.FalconForQuestionAnswering",
14
+ "AutoModelForSequenceClassification": "vilsonrodrigues/falcon-7b-instruct-sharded--modeling_falcon.FalconForSequenceClassification",
15
+ "AutoModelForTokenClassification": "vilsonrodrigues/falcon-7b-instruct-sharded--modeling_falcon.FalconForTokenClassification"
16
+ },
17
+ "bias": false,
18
+ "bos_token_id": 11,
19
+ "eos_token_id": 11,
20
+ "hidden_dropout": 0.0,
21
+ "hidden_size": 4544,
22
+ "initializer_range": 0.02,
23
+ "layer_norm_epsilon": 1e-05,
24
+ "model_type": "falcon",
25
+ "multi_query": true,
26
+ "new_decoder_architecture": false,
27
+ "num_attention_heads": 71,
28
+ "num_hidden_layers": 32,
29
+ "num_kv_heads": 71,
30
+ "parallel_attn": true,
31
+ "quantization_config": {
32
+ "_load_in_4bit": true,
33
+ "_load_in_8bit": false,
34
+ "bnb_4bit_compute_dtype": "bfloat16",
35
+ "bnb_4bit_quant_storage": "uint8",
36
+ "bnb_4bit_quant_type": "nf4",
37
+ "bnb_4bit_use_double_quant": true,
38
+ "llm_int8_enable_fp32_cpu_offload": false,
39
+ "llm_int8_has_fp16_weight": false,
40
+ "llm_int8_skip_modules": null,
41
+ "llm_int8_threshold": 6.0,
42
+ "load_in_4bit": true,
43
+ "load_in_8bit": false,
44
+ "quant_method": "bitsandbytes"
45
+ },
46
+ "torch_dtype": "bfloat16",
47
+ "transformers_version": "4.39.3",
48
+ "use_cache": true,
49
+ "vocab_size": 65024
50
+ }