jukofyork commited on
Commit
40951bb
·
verified ·
1 Parent(s): e4a4015

Upload 6 files

Browse files
Files changed (2) hide show
  1. config.json +2 -3
  2. model.safetensors +1 -1
config.json CHANGED
@@ -1,5 +1,4 @@
1
  {
2
- "_name_or_path": "DeepSeek-V3-0324-CODER-DRAFT-0.6B",
3
  "architectures": [
4
  "Qwen2ForCausalLM"
5
  ],
@@ -10,7 +9,7 @@
10
  "hidden_size": 896,
11
  "initializer_range": 0.02,
12
  "intermediate_size": 4864,
13
- "max_position_embeddings": 163840,
14
  "max_window_layers": 21,
15
  "model_type": "qwen2",
16
  "num_attention_heads": 14,
@@ -19,7 +18,7 @@
19
  "pad_token_id": 1,
20
  "rms_norm_eps": 1e-06,
21
  "rope_scaling": {
22
- "factor": 4.0,
23
  "original_max_position_embeddings": 32768,
24
  "type": "yarn"
25
  },
 
1
  {
 
2
  "architectures": [
3
  "Qwen2ForCausalLM"
4
  ],
 
9
  "hidden_size": 896,
10
  "initializer_range": 0.02,
11
  "intermediate_size": 4864,
12
+ "max_position_embeddings": 65536,
13
  "max_window_layers": 21,
14
  "model_type": "qwen2",
15
  "num_attention_heads": 14,
 
18
  "pad_token_id": 1,
19
  "rms_norm_eps": 1e-06,
20
  "rope_scaling": {
21
+ "factor": 2.0,
22
  "original_max_position_embeddings": 32768,
23
  "type": "yarn"
24
  },
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3827b3badd63324ecd8b2986b9f6739f6f0d40b23a1f62e282db65b2666787c8
3
  size 1179168272
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac8908a40463523e49cf50f8faf9e5526b4302cf41f42f70f8f86bb4451b063b
3
  size 1179168272