RichardErkhov commited on
Commit
6addbf7
·
verified ·
1 Parent(s): 7d22698

uploaded model

Browse files
Files changed (1) hide show
  1. config.json +46 -0
config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "tiny-random-Phi3ForCausalLM",
3
+ "architectures": [
4
+ "Phi3ForCausalLM"
5
+ ],
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 1,
8
+ "embd_pdrop": 0.0,
9
+ "eos_token_id": 32000,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 32,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 64,
14
+ "max_position_embeddings": 4096,
15
+ "model_type": "phi3",
16
+ "num_attention_heads": 4,
17
+ "num_hidden_layers": 2,
18
+ "num_key_value_heads": 4,
19
+ "original_max_position_embeddings": 4096,
20
+ "pad_token_id": 32000,
21
+ "quantization_config": {
22
+ "_load_in_4bit": true,
23
+ "_load_in_8bit": false,
24
+ "bnb_4bit_compute_dtype": "float32",
25
+ "bnb_4bit_quant_storage": "uint8",
26
+ "bnb_4bit_quant_type": "fp4",
27
+ "bnb_4bit_use_double_quant": false,
28
+ "llm_int8_enable_fp32_cpu_offload": false,
29
+ "llm_int8_has_fp16_weight": false,
30
+ "llm_int8_skip_modules": null,
31
+ "llm_int8_threshold": 6.0,
32
+ "load_in_4bit": true,
33
+ "load_in_8bit": false,
34
+ "quant_method": "bitsandbytes"
35
+ },
36
+ "resid_pdrop": 0.0,
37
+ "rms_norm_eps": 1e-05,
38
+ "rope_scaling": null,
39
+ "rope_theta": 10000.0,
40
+ "sliding_window": 2047,
41
+ "tie_word_embeddings": false,
42
+ "torch_dtype": "float16",
43
+ "transformers_version": "4.44.2",
44
+ "use_cache": true,
45
+ "vocab_size": 32064
46
+ }