teszenofficial commited on
Commit
c19d26b
·
verified ·
1 Parent(s): 106454f

Upload 3 files

Browse files
Files changed (3) hide show
  1. MTP2_9_MODEL.pt +3 -0
  2. MTP2_9_config.json +36 -0
  3. MTP2_9_tokenizer.model +3 -0
MTP2_9_MODEL.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:60efb36ba134f1d710dda5758645cc79a332e85498e76c59e2a4478037311d18
3
+ size 96473509
MTP2_9_config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab_size": 8000,
3
+ "block_size": 384,
4
+ "n_layer": 8,
5
+ "n_head": 7,
6
+ "n_embd": 448,
7
+ "dropout": 0.2,
8
+ "tokenizer_type": "sentencepiece_bpe",
9
+ "tokenizer_file": "MTP2_9_tokenizer.model",
10
+ "special_tokens": {
11
+ "pad_id": 0,
12
+ "bos_id": 1,
13
+ "eos_id": 2,
14
+ "unk_id": 3
15
+ },
16
+ "generation_defaults": {
17
+ "max_new_tokens": 200,
18
+ "temperature": 0.8,
19
+ "top_k": 40,
20
+ "top_p": 0.9,
21
+ "repetition_penalty": 1.2,
22
+ "no_repeat_ngram_size": 3
23
+ },
24
+ "meta": {
25
+ "model_name": "MTP-2.9",
26
+ "architecture": "gpt_rope_swiglu_rmsnorm",
27
+ "trained_examples": 7136,
28
+ "augmented_examples": 21512,
29
+ "corpus_fragments": 318,
30
+ "total_sequences": 21864,
31
+ "best_val_loss": 2.4462781286239625,
32
+ "stopped_early": false,
33
+ "label_smoothing": 0.1,
34
+ "created": "2026-07-15 07:09:14"
35
+ }
36
+ }
MTP2_9_tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fdf0142c925cbf9d7df287779d28341705a94e8906604a69d305585d12ed1601
3
+ size 371798