Hugging Face's logo

minpeter
/
tiny-ko-random

Text Generation
Transformers
Safetensors
llama
text-generation-inference
Model card Files
xet
Community
minpeter commited on
Commit
12a2439
·
verified ·
1 Parent(s): 0554c8c

Upload MistralForCausalLM

Browse files
Files changed (2) hide show
  1. config.json +3 -6
  2. model.safetensors +1 -1
config.json CHANGED
@@ -1,8 +1,7 @@
1
  {
2
  "architectures": [
3
- "LlamaForCausalLM"
4
  ],
5
- "attention_bias": false,
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 1,
8
  "eos_token_id": 32001,
@@ -12,16 +11,14 @@
12
  "initializer_range": 0.02,
13
  "intermediate_size": 1920,
14
  "max_position_embeddings": 8192,
15
- "mlp_bias": false,
16
- "model_type": "llama",
17
  "num_attention_heads": 6,
18
  "num_hidden_layers": 32,
19
  "num_key_value_heads": 2,
20
  "pad_token_id": 32003,
21
- "pretraining_tp": 1,
22
  "rms_norm_eps": 1e-06,
23
- "rope_scaling": null,
24
  "rope_theta": 10000.0,
 
25
  "tie_word_embeddings": true,
26
  "torch_dtype": "float32",
27
  "transformers_version": "4.51.3",
 
1
  {
2
  "architectures": [
3
+ "MistralForCausalLM"
4
  ],
 
5
  "attention_dropout": 0.0,
6
  "bos_token_id": 1,
7
  "eos_token_id": 32001,
 
11
  "initializer_range": 0.02,
12
  "intermediate_size": 1920,
13
  "max_position_embeddings": 8192,
14
+ "model_type": "mistral",
 
15
  "num_attention_heads": 6,
16
  "num_hidden_layers": 32,
17
  "num_key_value_heads": 2,
18
  "pad_token_id": 32003,
 
19
  "rms_norm_eps": 1e-06,
 
20
  "rope_theta": 10000.0,
21
+ "sliding_window": 4096,
22
  "tie_word_embeddings": true,
23
  "torch_dtype": "float32",
24
  "transformers_version": "4.51.3",
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:21afd18bda646d6b654228622b2a0e03309ca5c2e0047f8a5e57854e8b63d437
3
  size 494169296
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:caba12d551bed5628f3926faf2ff23966de3fc986ac0faba3055aab214859abb
3
  size 494169296