jgalego commited on
Commit
6ae67af
·
verified ·
1 Parent(s): 0e03331

Add olmo-tiny

Browse files
Files changed (4) hide show
  1. chipformer.json +74 -0
  2. config.json +55 -0
  3. generation_config.json +6 -0
  4. model.safetensors +3 -0
chipformer.json ADDED
@@ -0,0 +1,74 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "vocab": 230324,
3
+ "pad": 230324,
4
+ "controls": {
5
+ "NT": 230316,
6
+ "ST0": 230317,
7
+ "ST1": 230318,
8
+ "SS0": 230319,
9
+ "SS1": 230320,
10
+ "HALT": 230321,
11
+ "GO": 230322,
12
+ "SC": 230323
13
+ },
14
+ "families": {
15
+ "V": [
16
+ 0,
17
+ 32768
18
+ ],
19
+ "F0": [
20
+ 32768,
21
+ 32768
22
+ ],
23
+ "F1": [
24
+ 65536,
25
+ 32768
26
+ ],
27
+ "F2": [
28
+ 98304,
29
+ 4
30
+ ],
31
+ "J": [
32
+ 98308,
33
+ 65536
34
+ ],
35
+ "R": [
36
+ 163844,
37
+ 65536
38
+ ],
39
+ "BYTE": [
40
+ 229380,
41
+ 512
42
+ ],
43
+ "FLT": [
44
+ 229892,
45
+ 8
46
+ ],
47
+ "HEX": [
48
+ 229900,
49
+ 16
50
+ ],
51
+ "QS": [
52
+ 229916,
53
+ 16
54
+ ],
55
+ "QD": [
56
+ 229932,
57
+ 128
58
+ ],
59
+ "BD": [
60
+ 230060,
61
+ 256
62
+ ]
63
+ },
64
+ "chip_layers": 11,
65
+ "circuit": {
66
+ "layers": 11,
67
+ "dims": 2679,
68
+ "heads": 144,
69
+ "ffn_units": 42014,
70
+ "max_units": 7876,
71
+ "max_heads": 30
72
+ },
73
+ "hidden": 2816
74
+ }
config.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "OlmoHybridForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 230324,
8
+ "dtype": "bfloat16",
9
+ "eos_token_id": 230321,
10
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2816,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
15
+ "layer_types": [
16
+ "linear_attention",
17
+ "linear_attention",
18
+ "linear_attention",
19
+ "linear_attention",
20
+ "linear_attention",
21
+ "linear_attention",
22
+ "linear_attention",
23
+ "linear_attention",
24
+ "linear_attention",
25
+ "linear_attention",
26
+ "linear_attention",
27
+ "full_attention"
28
+ ],
29
+ "linear_a_log_max": 16.0,
30
+ "linear_a_log_min": 0.0,
31
+ "linear_allow_neg_eigval": true,
32
+ "linear_conv_kernel_dim": 4,
33
+ "linear_dt_init_floor": 0.0001,
34
+ "linear_dt_max": 0.1,
35
+ "linear_dt_min": 0.001,
36
+ "linear_key_head_dim": 96,
37
+ "linear_num_key_heads": 30,
38
+ "linear_num_value_heads": 30,
39
+ "linear_value_head_dim": 192,
40
+ "max_position_embeddings": 32768,
41
+ "model_type": "olmo_hybrid",
42
+ "num_attention_heads": 1,
43
+ "num_hidden_layers": 12,
44
+ "num_key_value_heads": 1,
45
+ "pad_token_id": 230324,
46
+ "rms_norm_eps": 1e-06,
47
+ "rope_parameters": {
48
+ "rope_theta": null,
49
+ "rope_type": "default"
50
+ },
51
+ "tie_word_embeddings": false,
52
+ "transformers_version": "5.19.0",
53
+ "use_cache": true,
54
+ "vocab_size": 230336
55
+ }
generation_config.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": false,
3
+ "eos_token_id": 230321,
4
+ "pad_token_id": 230324,
5
+ "transformers_version": "5.19.0"
6
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11ae8f6e0f92fb27c8a8fb7cb4d46ef1687eb4c3e0518dd7bc9aabdf93a5ccb0
3
+ size 6260116672