statezeropy commited on
Commit
717c396
1 Parent(s): 4d3f58c

Upload GPTNeoXForCausalLM

Browse files
Files changed (3) hide show
  1. config.json +43 -0
  2. generation_config.json +6 -0
  3. model.safetensors +3 -0
config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "beomi/KoAlpaca-Polyglot-5.8B",
3
+ "architectures": [
4
+ "GPTNeoXForCausalLM"
5
+ ],
6
+ "attention_bias": true,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "classifier_dropout": 0.1,
10
+ "eos_token_id": 0,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout": 0.0,
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 16384,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 2048,
18
+ "model_type": "gpt_neox",
19
+ "num_attention_heads": 16,
20
+ "num_hidden_layers": 28,
21
+ "num_steps": "global_step320000",
22
+ "quantization_config": {
23
+ "bnb_4bit_compute_dtype": "bfloat16",
24
+ "bnb_4bit_quant_type": "nf4",
25
+ "bnb_4bit_use_double_quant": true,
26
+ "llm_int8_enable_fp32_cpu_offload": false,
27
+ "llm_int8_has_fp16_weight": false,
28
+ "llm_int8_skip_modules": null,
29
+ "llm_int8_threshold": 6.0,
30
+ "load_in_4bit": true,
31
+ "load_in_8bit": false,
32
+ "quant_method": "bitsandbytes"
33
+ },
34
+ "rope_scaling": null,
35
+ "rotary_emb_base": 10000,
36
+ "rotary_pct": 0.25,
37
+ "tie_word_embeddings": false,
38
+ "torch_dtype": "float16",
39
+ "transformers_version": "4.37.0.dev0",
40
+ "use_cache": true,
41
+ "use_parallel_residual": true,
42
+ "vocab_size": 30080
43
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 0,
4
+ "eos_token_id": 2,
5
+ "transformers_version": "4.37.0.dev0"
6
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01193a2fab720a5fc34272ffe13c1419e72fdacebaac0ace8f03e33662a18044
3
+ size 3404110067