MoreThought commited on
Commit
d557ada
·
verified ·
1 Parent(s): 2f431e7

AresMicro-10k checkpoint

Browse files
Files changed (5) hide show
  1. AresMicro-10k-fp16.gguf +0 -0
  2. README.md +1 -1
  3. checkpoint.pkl +2 -2
  4. config.json +1 -1
  5. model_fp16.npz +2 -2
AresMicro-10k-fp16.gguf CHANGED
Binary files a/AresMicro-10k-fp16.gguf and b/AresMicro-10k-fp16.gguf differ
 
README.md CHANGED
@@ -19,7 +19,7 @@ tags:
19
  - Shared block loops: 50
20
  - Hidden: 32
21
  - Heads: 4
22
- - MLP: 52
23
  - PT target: 5,000,000,000
24
  - RL target: 5,000,000,000
25
  - Benchmark file: purearith-1-answer-only.jsonl
 
19
  - Shared block loops: 50
20
  - Hidden: 32
21
  - Heads: 4
22
+ - MLP: 64
23
  - PT target: 5,000,000,000
24
  - RL target: 5,000,000,000
25
  - Benchmark file: purearith-1-answer-only.jsonl
checkpoint.pkl CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5d0707aff5c1b07581152ec2e70c274c14f61f89a19c8f2f4e4db9212b1fb8cf
3
- size 120228
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62d7392f0b52a072945e35ce2e7d4d095e21f6be078625850524c6a048bb79d6
3
+ size 134052
config.json CHANGED
@@ -10,7 +10,7 @@
10
  "num_hidden_layers": 1,
11
  "num_loops": 50,
12
  "shared_loop_weights": true,
13
- "intermediate_size": 52,
14
  "dropout": 0.0,
15
  "torch_dtype": "bfloat16",
16
  "tokenizer_class": "AresMicro10kTokenizer",
 
10
  "num_hidden_layers": 1,
11
  "num_loops": 50,
12
  "shared_loop_weights": true,
13
+ "intermediate_size": 64,
14
  "dropout": 0.0,
15
  "torch_dtype": "bfloat16",
16
  "tokenizer_class": "AresMicro10kTokenizer",
model_fp16.npz CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f3b85caca2a11de11ede16e92aa343688ea9be3bfb8bae60968159e1084cd6e0
3
- size 20527
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ccca632ea172ed75f564dd27002f0e998d0e6f73210e3c79f9770471643538c6
3
+ size 22574