Hugging Face's logo

11-47
/
Nano.Deep.Reasoner.11m-HyperMini

Transformers
Safetensors
nano_deep_reasoner_hypermini
causal-lm
decoder-only
reasoning
deep-reasoning
recurrent-transformer
adaptive-computation
chain-of-thought
adaptive-reasoning
Model card Files
xet
 Community
Plans11 commited on
Commit
0e15fbd
·
verified ·
1 Parent(s): 3d578b2

HyperMini Session 3 complete | 20,000 examples | loss=1.343789

Browse files
Files changed (7) hide show
  1. README.md +6 -6
  2. example_ledger.json +0 -0
  3. model.safetensors +1 -1
  4. optimizer.pt +2 -2
  5. rng_state.pt +1 -1
  6. tokenizer_config.json +1 -0
  7. training_state.json +0 -0
README.md CHANGED
@@ -59,7 +59,7 @@ Dataset:
59
 
60
  Plans11/Organized_PreTrain_1k_Context
61
 
62
- Each session contains up to 20,000 NEW examples.
63
 
64
  Examples are protected by SHA-256 hashes.
65
 
@@ -87,16 +87,16 @@ before resume.
87
  ## Current state
88
 
89
  Completed sessions:
90
- 2
91
 
92
  Unique examples reserved/trained:
93
- 20,000
94
 
95
  Unique completed examples:
96
- 40,000
97
 
98
  Global optimizer steps:
99
- 626
100
 
101
  Last session loss:
102
- 2.140315593892336
 
59
 
60
  Plans11/Organized_PreTrain_1k_Context
61
 
62
+ Each session contains up to 200,000 NEW examples.
63
 
64
  Examples are protected by SHA-256 hashes.
65
 
 
87
  ## Current state
88
 
89
  Completed sessions:
90
+ 3
91
 
92
  Unique examples reserved/trained:
93
+ 0
94
 
95
  Unique completed examples:
96
+ 60,000
97
 
98
  Global optimizer steps:
99
+ 939
100
 
101
  Last session loss:
102
+ 1.3437887360095977
example_ledger.json CHANGED
The diff for this file is too large to render. See raw diff
 
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cb66b65f180c06b99e704a79946345d7edf6b776c9c0e6f78f07cbfa7d05593e
3
  size 59742900
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45e25b7cb4a59e783e4d8e37d345051909703bc97fe63e3909c24f51ee40a092
3
  size 59742900
optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4336bdf5d3c09c632c3acd21d06590e31dca3788d42e666d6915e0d425582cb9
3
- size 88811659
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3b9358427a9c8b4715941c79894f9f76ca531fd2155d4248a98cc8797ded0eda
3
+ size 88812619
rng_state.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:681ab3e07914b997afa64fd531f8576d52944fab321630a3464a27fcccc14d82
3
  size 10931
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3379799a4824e1efe76695d57b9d5339e7a7e76e0f36ac561d1ed73d67a443f
3
  size 10931
tokenizer_config.json CHANGED
@@ -2,6 +2,7 @@
2
  "backend": "tokenizers",
3
  "bos_token": "<bos>",
4
  "eos_token": "<eos>",
 
5
  "model_max_length": 1000000000000000019884624838656,
6
  "pad_token": "<pad>",
7
  "tokenizer_class": "TokenizersBackend",
 
2
  "backend": "tokenizers",
3
  "bos_token": "<bos>",
4
  "eos_token": "<eos>",
5
+ "is_local": true,
6
  "model_max_length": 1000000000000000019884624838656,
7
  "pad_token": "<pad>",
8
  "tokenizer_class": "TokenizersBackend",
training_state.json CHANGED
The diff for this file is too large to render. See raw diff