arch-space commited on
Commit
0454e9f
·
verified ·
1 Parent(s): 72c531b

Add files using upload-large-folder tool

Browse files
README.md ADDED
@@ -0,0 +1,49 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ pipeline_tag: text-generation
4
+ tags:
5
+ - olmo3
6
+ - baseline
7
+ - safetensors
8
+ - sliding-window-attention
9
+ ---
10
+
11
+ # OLMo 3 1B Baseline — Stage 3 Long-context Training
12
+
13
+ This repository contains the pure OLMo 3 1B baseline checkpoint from
14
+ `o3b1b-s3-longmino50b-s65536-g64-m1-ga1-tp1-cp8-dp64-hsdp32-b2-lr2p5e4-w200-save1000-identity-512npu-share-20260802-v1` at iteration `11921`. SiameseNorm and Depth-Attention
15
+ are disabled.
16
+
17
+ - Training sequence length: 65,536
18
+ - Model context capacity: 65,536
19
+ - Sliding-window size: 4,096
20
+ - Attention pattern: `[SWA, SWA, SWA, Full]`
21
+ - Vocabulary: 100,278 real tokens; 74 Megatron padding rows removed
22
+
23
+ Stage 3/4 apply YaRN only to Full-Attention layers. OLMo 3 SWA layers use the
24
+ original RoPE and retain their 4,096-token local window.
25
+
26
+ ## Loading
27
+
28
+ `transformers>=4.57.6,<5` is required.
29
+
30
+ ```python
31
+ import torch
32
+ from transformers import AutoModelForCausalLM, AutoTokenizer
33
+
34
+ repo_id = "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention--baseline-stage3"
35
+ tokenizer = AutoTokenizer.from_pretrained(
36
+ repo_id,
37
+ use_fast=True,
38
+ fix_mistral_regex=False,
39
+ )
40
+ model = AutoModelForCausalLM.from_pretrained(
41
+ repo_id,
42
+ dtype=torch.bfloat16,
43
+ attn_implementation="sdpa",
44
+ )
45
+ ```
46
+
47
+ `fix_mistral_regex=False` preserves the tokenizer behavior used for training.
48
+ The checkpoint uses the official Transformers `Olmo3ForCausalLM`
49
+ implementation and does not require remote code.
SHA256SUMS ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ ee4e133cd814cf50530202f800a9d79df885dfe97fa5e540e4c9ca431b12d1e6 README.md
2
+ 0dbc55ce09a6ce495d6889efa2d31ef01cd14a1f4b70262fe4612b94ec57f99c config.json
3
+ 775219d4e613a55e6f052b0e92ee1e0880eb3dd0a8a00cb690e435f2628c82be conversion_manifest.json
4
+ dbd4aa5dc6f7f1bdc0c181d9e5582f6571e1d2a22878dfcd98149324b9f5c4a2 generation_config.json
5
+ 257cd56d6752f2a494e69907f8e4aabfd653a496df5ee6c6be38648e8a437237 hf_validation_report.json
6
+ d9bfc690d349b8c2a7ea1172481013b6b09d64e9b73056a037037e2ecb0aee96 model.safetensors
7
+ 4cb2b52960bababa8ec27153831e405ef7811e3fad1226106e09c28e715cfb21 special_tokens_map.json
8
+ 73fd5254624f39a88e3faac6a8e11300fc3c735ed37880d4f4f08db898eaecca tokenizer.json
9
+ df7285f1e002db7ce39f3da5b85abed05bbf8949e3f576051e60b51c73b64fdf tokenizer_config.json
10
+ 9e14712c91b37c7aab74b1306baa46ac342d620637a4b44523cdc3aec7d24195 vocab.json
config.json ADDED
@@ -0,0 +1,66 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "",
3
+ "architectures": [
4
+ "Olmo3ForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": null,
9
+ "eos_token_id": 100257,
10
+ "hidden_act": "silu",
11
+ "hidden_dropout": 0.0,
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "layer_types": [
16
+ "sliding_attention",
17
+ "sliding_attention",
18
+ "sliding_attention",
19
+ "full_attention",
20
+ "sliding_attention",
21
+ "sliding_attention",
22
+ "sliding_attention",
23
+ "full_attention",
24
+ "sliding_attention",
25
+ "sliding_attention",
26
+ "sliding_attention",
27
+ "full_attention",
28
+ "sliding_attention",
29
+ "sliding_attention",
30
+ "sliding_attention",
31
+ "full_attention"
32
+ ],
33
+ "max_position_embeddings": 65536,
34
+ "max_sequence_length": 65536,
35
+ "mlp_bias": false,
36
+ "model_type": "olmo3",
37
+ "num_attention_heads": 16,
38
+ "num_hidden_layers": 16,
39
+ "num_key_value_heads": 16,
40
+ "olmo3_weight_decay": true,
41
+ "pad_token_id": 100277,
42
+ "padded_vocab_size": 100352,
43
+ "qk_norm": true,
44
+ "qk_norm_mode": "full_projection",
45
+ "rms_norm_eps": 1e-06,
46
+ "rope_full_precision": true,
47
+ "rope_scaling": {
48
+ "beta_fast": 32.0,
49
+ "beta_slow": 1.0,
50
+ "factor": 8.0,
51
+ "original_max_position_embeddings": 8192,
52
+ "rope_type": "yarn"
53
+ },
54
+ "rope_theta": 500000.0,
55
+ "scale_embeds": false,
56
+ "scale_output_layer_init": false,
57
+ "sliding_window": 4096,
58
+ "tie_word_embeddings": false,
59
+ "torch_dtype": "bfloat16",
60
+ "transformers_version": "4.57.6",
61
+ "true_vocab_size": 100278,
62
+ "truncated_normal_factor": 3.0,
63
+ "use_cache": true,
64
+ "vocab_size": 100278,
65
+ "vocab_z_loss_coeff": 1e-05
66
+ }
conversion_manifest.json ADDED
@@ -0,0 +1,228 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "schema": "olmo3.base.hf_export/v1",
3
+ "source": {
4
+ "checkpoint_metadata_sha256": "2d103e0f458a8975b3e53fca66e47d6e216daba7bcee75a6fc8d8fe7e0176170",
5
+ "iteration": 11921,
6
+ "model_schema_sha256": "5be7d972eb3294c31abe5a6ea18bb1b2c34e2ca8225a32a9bf1673fe5d67e4fd",
7
+ "parameter_count_with_padded_vocab": 1484916736,
8
+ "run_id": "o3b1b-s3-longmino50b-s65536-g64-m1-ga1-tp1-cp8-dp64-hsdp32-b2-lr2p5e4-w200-save1000-identity-512npu-share-20260802-v1"
9
+ },
10
+ "stage": "stage3",
11
+ "target": {
12
+ "format": "huggingface_olmo3_safetensors",
13
+ "padded_vocab_size": 100352,
14
+ "parameter_count": 1484613632,
15
+ "tensor_count": 179,
16
+ "tensor_sha256": {
17
+ "lm_head.weight": "ef19ff51170da8c44a868436cb5c84dfc4d0e1e1705ef46bb1ef88f9bbd7f41c",
18
+ "model.embed_tokens.weight": "54d285c0f8552eb4fd2962d630980850d782b5771cee35b098dc124b6efd2fa1",
19
+ "model.layers.0.mlp.down_proj.weight": "bfe3c2baa3424681c4a5f695ad7ee5071dbb904b047222b9d6c44733b0cbb986",
20
+ "model.layers.0.mlp.gate_proj.weight": "4c17f918d297d6595ad88cbad9e554764c1380524bddee2e0f08c3af2afd290e",
21
+ "model.layers.0.mlp.up_proj.weight": "274c159094190410de770361c903f97921746070701458458b5d1f085e8bf3d6",
22
+ "model.layers.0.post_attention_layernorm.weight": "e4278adc48f68b30763ff03cb223668859f26d2595f1af92f0c083129722730a",
23
+ "model.layers.0.post_feedforward_layernorm.weight": "b6b47569a419830e2a76646ec5d872e2526ee48777fc0e3a0d708935cf8e1a98",
24
+ "model.layers.0.self_attn.k_norm.weight": "f988861de9b493d43a22a0b76b889c9eb82c2a1e33e23d23bec74389eea0d7af",
25
+ "model.layers.0.self_attn.k_proj.weight": "e0960d9d68492eb9e72d1e47ded96316cab7089f46d23e41d72031205dc3ab55",
26
+ "model.layers.0.self_attn.o_proj.weight": "9285f495d9e39dca672b22e21678dad19a9b5e3de971bf9a85348c4e36f3e00e",
27
+ "model.layers.0.self_attn.q_norm.weight": "9d755e66a004320d02a65c43cc0ced1f611792b7333d4add4ad1678acaa6d194",
28
+ "model.layers.0.self_attn.q_proj.weight": "1df02bc323456107b003cf4e4196a918a052c47b7d99430a7178eaaada6c8159",
29
+ "model.layers.0.self_attn.v_proj.weight": "f613954e4bf1e88ab746c50d5493d07d1b130492afada166123058fd00833d9c",
30
+ "model.layers.1.mlp.down_proj.weight": "ae212cb7ded77e03f48abe1f3111e4b723dde6423eeddc7dede0c993426f7df0",
31
+ "model.layers.1.mlp.gate_proj.weight": "5f16cd585b94b3e9b00c9511252f1eb30e5761180793d0aa847f9f89e9c9acdb",
32
+ "model.layers.1.mlp.up_proj.weight": "0f470dc101378f4083f8406def9a5a4fd82e5c2da6edf721da3b39b4ca8bf42d",
33
+ "model.layers.1.post_attention_layernorm.weight": "5a69cb34daff2b5eec9850506e2ab5a6e29e7a886e1c79c8e28a5c4d8e05250f",
34
+ "model.layers.1.post_feedforward_layernorm.weight": "069920d566b4d9c181f1e260ddb1960e3bf2d712e41af8125f5e8a1b5a830c3d",
35
+ "model.layers.1.self_attn.k_norm.weight": "011cd8b1cec5326aa66c732f52178f4f5b953f49a3c390bff0fdd91a86a6f125",
36
+ "model.layers.1.self_attn.k_proj.weight": "1bd148c8d26cd005ff5221b09eec73d82a9c27d2616ea56be1950dcb2c9f5e7a",
37
+ "model.layers.1.self_attn.o_proj.weight": "c051f06d2d3e4db77d04f51cd53a09d2632690d98016891aba7a45573eb3c939",
38
+ "model.layers.1.self_attn.q_norm.weight": "0d9b99239c8a43072cb4ef815e369fd24956de2bb66dfa21128c8015c4f15a42",
39
+ "model.layers.1.self_attn.q_proj.weight": "883f29ef60291b4081c1c9bd325ee4718ee46d334b665e1af77f0c59d997b0c0",
40
+ "model.layers.1.self_attn.v_proj.weight": "59594c26b8477b24bbbbeda9a665bfb5430481bcd144ff11395d6ec71582b45f",
41
+ "model.layers.10.mlp.down_proj.weight": "2e35c3bdd31acacc82d6560c0b9ca9b50224e2c5174114a127fe2760ea956ba2",
42
+ "model.layers.10.mlp.gate_proj.weight": "387347e75b194a814d412b2ab4e183376790146008a1bae195712500f3159133",
43
+ "model.layers.10.mlp.up_proj.weight": "aa09ca26a3cdeaf5797fc768e17422a34ab164e42ce1c59ef28b5cae57ade346",
44
+ "model.layers.10.post_attention_layernorm.weight": "5aa9aad641b9310c5f7e9330cb5a5c11fbdc8ab23fde594f511e9c2e426c75d3",
45
+ "model.layers.10.post_feedforward_layernorm.weight": "d5728e5738e28e2168e5dd2901726bd1450c6425c65689d827579bc7a1167598",
46
+ "model.layers.10.self_attn.k_norm.weight": "f2352a4dc02ac1d3535a7d3d67df4191f9ef9d22b8003f084f7ac88297723a88",
47
+ "model.layers.10.self_attn.k_proj.weight": "cf0efaf4570e6c3bb3275268eb3532b67a34daf0db2ddd9b1f57181e258eb530",
48
+ "model.layers.10.self_attn.o_proj.weight": "81a1148f47c90a4a079720927d77b402c1bbe4568702d879365b2e476d5f4757",
49
+ "model.layers.10.self_attn.q_norm.weight": "76fef254d7ed98567b15fd814060ed6b5852e5a79236b7b0a43f28fa5be8f0a5",
50
+ "model.layers.10.self_attn.q_proj.weight": "c83383f5dbdf02a2a5a498719c6c8c7edba81db02d441c013c54378011e5d41f",
51
+ "model.layers.10.self_attn.v_proj.weight": "0e0b2f3acd6321838326fc41c989e0a53c64fb0480a62ca32bd487f46d0789a7",
52
+ "model.layers.11.mlp.down_proj.weight": "36bdffa9a296ea97af3096fea0a723fcf2c3ac9c8066bc334142e709f6acea5b",
53
+ "model.layers.11.mlp.gate_proj.weight": "b29ff2ec3c5cf1992b82c58e8e87e28131422da74e2dd863b1dd2502e8959628",
54
+ "model.layers.11.mlp.up_proj.weight": "5a9f291a60fc48cf8e9aebd1225f2864383cf6e202ded18bb4dba81d42faa4e0",
55
+ "model.layers.11.post_attention_layernorm.weight": "8b047461b2b973824f547fc8e007dfa250309930e83873f10513eb17403433e7",
56
+ "model.layers.11.post_feedforward_layernorm.weight": "e1c7e6245d26fb74bb3f3b873fc894b874843cfbde1602dd65baab127d4b3023",
57
+ "model.layers.11.self_attn.k_norm.weight": "66ef1073370a8385b4420f03fe437a54830f8d6521e1fb172ebd15d387104ca2",
58
+ "model.layers.11.self_attn.k_proj.weight": "fa1ded0889d209136a2bab4e58cdd8eefb88eb4e22de506b2ea7af328f74796c",
59
+ "model.layers.11.self_attn.o_proj.weight": "7dc0396d1cbd4f47383f3167c4816a4e22000f827de2750b68cac45901ffb54a",
60
+ "model.layers.11.self_attn.q_norm.weight": "264f7b7d43c4adbfd9bd7bb097e5946c2016a0b21a7cb17329ce6fbefc695b2b",
61
+ "model.layers.11.self_attn.q_proj.weight": "25677df90834ccafeb12d99e31e85b5a5918cf6dc326383264cd398409aaf721",
62
+ "model.layers.11.self_attn.v_proj.weight": "51216bfb33305b1e06a0036e57aeadd1b2adc2726627e8e4d413ac775fb1c1ba",
63
+ "model.layers.12.mlp.down_proj.weight": "21c704bd2ce3ddfd851d222a37909e89125ebf85c779047067325fd6b84ade06",
64
+ "model.layers.12.mlp.gate_proj.weight": "230468322195dc5c170064a918eb8314248ca14b0c2ec2f64d2c0410022e0e16",
65
+ "model.layers.12.mlp.up_proj.weight": "b9904897873143291c94ca61e3719cc2726dd8b982cd2f58de05930c32d1d771",
66
+ "model.layers.12.post_attention_layernorm.weight": "32be81386b67c9dae15fe9f5d178c4a61d16f21729c41846ae44d2ff09d6e12c",
67
+ "model.layers.12.post_feedforward_layernorm.weight": "fb349fc01d692e194f7a0a8a52362e78e071fc9c4317df910f9e7777ec1ab2ae",
68
+ "model.layers.12.self_attn.k_norm.weight": "a960a70b3f19eb1189daa71dc5c021612622837e7b650c60bb5d006542b9d6a3",
69
+ "model.layers.12.self_attn.k_proj.weight": "d6b17c7a46b9ecb802448db19830a8b739fe5014e37fa4399e713c022a1fd33f",
70
+ "model.layers.12.self_attn.o_proj.weight": "9de55dd081b4db1d6996be30072f39aae652a95338c228a08c6ac8872452d029",
71
+ "model.layers.12.self_attn.q_norm.weight": "605724fd61e4befe74059960517e55d350eec598ec9d6094f9480d781631aad6",
72
+ "model.layers.12.self_attn.q_proj.weight": "e64344d62f77ea85bc66a573188beb4e937ab63166483e1046b4330c013f949e",
73
+ "model.layers.12.self_attn.v_proj.weight": "a9a8ca1bbfa81a82bbf8821ae99ab3c5672b596bc539a194a83f076f7ca4995e",
74
+ "model.layers.13.mlp.down_proj.weight": "d6c9ea767adec370e2e674e14b36758d87598c11fa407d96b00a54baa5a11da4",
75
+ "model.layers.13.mlp.gate_proj.weight": "b3a2d1da4f76615e689bf87cec3c27c50ab86420064ff864ace312df02bec208",
76
+ "model.layers.13.mlp.up_proj.weight": "04d4b9f59c6843bb25b9611f4a3e13bc35e6f7bc7bb2405d0b1f7a6f0c20bf70",
77
+ "model.layers.13.post_attention_layernorm.weight": "01875197c31754ebed008ae14454bb82e43bb72cce7a28d420b14fbdb5403152",
78
+ "model.layers.13.post_feedforward_layernorm.weight": "846d9493ecec82454946f19085702a62594a4600ae052c391cffa5ece2e276cd",
79
+ "model.layers.13.self_attn.k_norm.weight": "f6c7ec12f7752711d842c43c37b607d65554af8d979b4fd7106262a588adb5b1",
80
+ "model.layers.13.self_attn.k_proj.weight": "9a3bd6d0c0340ce6976b6e3f4b417d5d5f284acb7e1b6dbf7cc8f3ddf8ef4da1",
81
+ "model.layers.13.self_attn.o_proj.weight": "ff63f5f0fb1be5b7acbee253223579600e29ec6130ab1275bce3651d5f2388dd",
82
+ "model.layers.13.self_attn.q_norm.weight": "b04be1b7cdd8e6c607f8524823e89440e2f9f347f87fd9d9df9552f5fcff004e",
83
+ "model.layers.13.self_attn.q_proj.weight": "52cf83d8d377feaeafea311fc84a3c0980f072e1453d969c2773dbc3acb4570f",
84
+ "model.layers.13.self_attn.v_proj.weight": "e9b41f910039ff2a54a40d79c42a2f261e2ae77ec3155b60a96931de45598393",
85
+ "model.layers.14.mlp.down_proj.weight": "aa35156c245f49763833d510bb69f37b9e39332b880b77d1b61be33a71251516",
86
+ "model.layers.14.mlp.gate_proj.weight": "b331c26e3e11544dbd28c967c7d130d574e468e8ce1a4ab45b215f65ac669d07",
87
+ "model.layers.14.mlp.up_proj.weight": "af94cf91571f0baa6343d7b65cf8954c43ffc734c07fa9a6504f11db8bff3bca",
88
+ "model.layers.14.post_attention_layernorm.weight": "35b64ed61ac1d7dd266b779740510b2dc22d8fc88bb3b8b2ef66bf6ee24a3791",
89
+ "model.layers.14.post_feedforward_layernorm.weight": "7b71d96322ea75c365a32e1323dcd37f68f6a9b1bc3cc59d12dc49fc146966e8",
90
+ "model.layers.14.self_attn.k_norm.weight": "dcedf325edf1a1091202d06cbf3dd594bae0aaadd7be0351750fb17a71694c1a",
91
+ "model.layers.14.self_attn.k_proj.weight": "2e77c833fb4be7c06cf526662420678f92c5be13ad40254c5d37bed4b5054c9f",
92
+ "model.layers.14.self_attn.o_proj.weight": "f4ec3ca489accad525800b36c22bdee45a222b17f472582e6d77b730e5da68c7",
93
+ "model.layers.14.self_attn.q_norm.weight": "09666af05129b860ebd655bc2413435fd3921604f54bf3c6c77a4fcd8943b76a",
94
+ "model.layers.14.self_attn.q_proj.weight": "4193167794617f953ad3da26f58298c203fa984c60074f0fa2ad3635f69d0ac5",
95
+ "model.layers.14.self_attn.v_proj.weight": "17770979b23d87c23800d39ae02da8359860f9b24aed2769fa2a2d2997e7dae9",
96
+ "model.layers.15.mlp.down_proj.weight": "675de1a43d96272f2c9d8e100a887915a977500d7ad92928ed8c2283716219a7",
97
+ "model.layers.15.mlp.gate_proj.weight": "240f02583cea0325401dbf1e88cd540c57de1dcb24d6f1f78d6362e4cd28620c",
98
+ "model.layers.15.mlp.up_proj.weight": "8814adbf8933f5e35cc9e4fd521c6df37a278a67d1a3b38a598f7e25753848db",
99
+ "model.layers.15.post_attention_layernorm.weight": "55c62ecf24a47b9ceb1743a390384427c94d87b8eb6bda60455cef09667089f7",
100
+ "model.layers.15.post_feedforward_layernorm.weight": "894a1a57b7809b1939c5dfe6936b66171206a1fc12daf6370fc3040faf341be6",
101
+ "model.layers.15.self_attn.k_norm.weight": "98e118f92ce93b5c1eac96b4f10b0860344a6062a730bc99cccf67e3a86dcec8",
102
+ "model.layers.15.self_attn.k_proj.weight": "629e26f4c7425d2a008c00034b22d99ade7284a5ef90be4604aeb6fb72ed456a",
103
+ "model.layers.15.self_attn.o_proj.weight": "474d4d2edd0681ecbec92d6000e3cc33144f579be82faf39e0ca510bf5e8cb25",
104
+ "model.layers.15.self_attn.q_norm.weight": "8a1755099c3864084919fae9751bc48b1c11a4690ff2d2ee8c2eaa6880c8c106",
105
+ "model.layers.15.self_attn.q_proj.weight": "eb8fbbd594982cb93dc5a49274cba319cd27deb26e4020ac6f247d3ea01e6d7d",
106
+ "model.layers.15.self_attn.v_proj.weight": "d3d0cbf0153cb4a48a0367076c7b98be5e695c6f9a5b5ed683adc8743415ce7d",
107
+ "model.layers.2.mlp.down_proj.weight": "e4dc31393618e2270a42b2775e7435c90f25d4d598933408470b1438941fdec9",
108
+ "model.layers.2.mlp.gate_proj.weight": "b93081b660493d8d0995ce9fe52c0365b2440e38b26fe08d83dd73b5d1d8bce1",
109
+ "model.layers.2.mlp.up_proj.weight": "b940b4b2b8491d9939f15b49b2175595470245b11fd46c5e752ed089f6fa3696",
110
+ "model.layers.2.post_attention_layernorm.weight": "61f85f750cc7cf4600e8b4904765a490a7c95d17a9fcd8396e2ddbcdb1548dac",
111
+ "model.layers.2.post_feedforward_layernorm.weight": "3b3ba36d09454f022f93a7045f9c500eda5eae6f37c240a413ac6a8d1c842e11",
112
+ "model.layers.2.self_attn.k_norm.weight": "fb5db64845eb2cab1a42e2d28378d42a3da0105960b9788f16a5a71fdacfc5cc",
113
+ "model.layers.2.self_attn.k_proj.weight": "ea88b110bbaec369e18136762d09643d883df219a1f2e16de46c4043fb5c225a",
114
+ "model.layers.2.self_attn.o_proj.weight": "ea5ca5d346fa2fc85e6a47162e0831653faba5edc1e0cc86e10206206f483bf2",
115
+ "model.layers.2.self_attn.q_norm.weight": "8cf1a1cee8543327b2d1db1a453ba95a63f90107620219018fb42bf01ee35433",
116
+ "model.layers.2.self_attn.q_proj.weight": "14e68d5d04ec04a6736d236186a937264dca4828c0fb6cf5b3c97d2d07fe4e17",
117
+ "model.layers.2.self_attn.v_proj.weight": "92260dbb982387951aa5fe739d5fb60d2b07de583e075e50988f9d46b2ebfa59",
118
+ "model.layers.3.mlp.down_proj.weight": "693e660799c106d02b2ec4244f227d391e466587b908f2c54113e6aba023629f",
119
+ "model.layers.3.mlp.gate_proj.weight": "4c4a359cf11fe6e9630e440ae67da38ec81106b3c7eeee91764b6e156eff5b06",
120
+ "model.layers.3.mlp.up_proj.weight": "9347ee667d79fad5e213f768bea01ea1c1ca116b90950b7c2fd13ec8f277b4e2",
121
+ "model.layers.3.post_attention_layernorm.weight": "b9f61f956d3aca7edb03e390d43f164bca6ae3908658d12deac463d8b447aa69",
122
+ "model.layers.3.post_feedforward_layernorm.weight": "d5cabd987350910373167e81e9c06a75323ad0fe809c74a3b758eb6000409038",
123
+ "model.layers.3.self_attn.k_norm.weight": "54b53acb61ca55ba454b7de2ff2ece7a09fccd1ce7659bd64b41eba6757668f4",
124
+ "model.layers.3.self_attn.k_proj.weight": "1c0ae1df934f01ab826c44e4283984c04785c8bed21bb85f6103a7f9c0429928",
125
+ "model.layers.3.self_attn.o_proj.weight": "5f90f6d2ea704741aacd9915b042e057a5e0b90125f17ef15baf331dad4d6c5a",
126
+ "model.layers.3.self_attn.q_norm.weight": "8c7314b26096b6f3c01250eb2f4da723d287497fd8d43fafe31f56ba350be142",
127
+ "model.layers.3.self_attn.q_proj.weight": "be9cda007dd260ae4c3b40122ffa7a5df125b7355b5cbcf2fa750de0340b0820",
128
+ "model.layers.3.self_attn.v_proj.weight": "edc0b0be931bda6fd42f9540b905dac9a5f309d7d5e14505bf18b9b2d4738dc1",
129
+ "model.layers.4.mlp.down_proj.weight": "864d89d80ad408a38974a6ba29e905dea938baf22fc3f82943f2c1d7318317b4",
130
+ "model.layers.4.mlp.gate_proj.weight": "b514e882d243d33edba9b125514dcca97ee7c66d83ab6d31676bba9c4021b8cf",
131
+ "model.layers.4.mlp.up_proj.weight": "50f256d041c0047ed3f6ef3519d6a19621e374fa3f93463f6386abc56f2d6352",
132
+ "model.layers.4.post_attention_layernorm.weight": "a6a87147721295ad86872fdfd3608ca91a14943edaf4c9adc7555c15f68b142d",
133
+ "model.layers.4.post_feedforward_layernorm.weight": "b614fb164fd0f1f4d303e72596a38752a25101cc97e10e81142e83e6b57b0584",
134
+ "model.layers.4.self_attn.k_norm.weight": "f8ccbd3ca15376dd13f51ffee808232a806ca014d21a287a7057c365bd45759b",
135
+ "model.layers.4.self_attn.k_proj.weight": "f06df78a077e905a71ecebada43125512cd11891598ce6aa5685002f5893000f",
136
+ "model.layers.4.self_attn.o_proj.weight": "f64fa5947c59b931475e3a510093350980dec4ed89b33d8e929e4bd5c5abc22b",
137
+ "model.layers.4.self_attn.q_norm.weight": "c4a23e2a0305a80eeab43fda77a7b9f2e77c2e8e1e87f0172e19e606a4574fb0",
138
+ "model.layers.4.self_attn.q_proj.weight": "f4afd6114ad333b3fcd0827d82884e3bf0619e17b86e4e463437960b6f8da8ef",
139
+ "model.layers.4.self_attn.v_proj.weight": "84ba7470708c6978eddb0c787b3e14e725f0460257d7dfcba06c28db77832c1c",
140
+ "model.layers.5.mlp.down_proj.weight": "582ca0d8a3c763324d38e4fe6b51d80aee8dc077a871a365814a71bda8cb0bc8",
141
+ "model.layers.5.mlp.gate_proj.weight": "10a78f7fc488eda9118091bf3b75afed468a6c98bf6a1ea2b8a153d1964b1069",
142
+ "model.layers.5.mlp.up_proj.weight": "f2e7d6d51a64073cfb747801fc0c793c37d31b4813e9b37b7c31007285556b33",
143
+ "model.layers.5.post_attention_layernorm.weight": "60d68199f4f2708c02c2f8d3d4bb13317c923740c1538c90c27bd88068566d2b",
144
+ "model.layers.5.post_feedforward_layernorm.weight": "ad8c77c518f2f5f224b35b924c948894821bde4cdd97cf81e13114ab95aa62cf",
145
+ "model.layers.5.self_attn.k_norm.weight": "e73632ed78390be86af3066e7c7087787392675321b7d3cbf63560feea2b4a1c",
146
+ "model.layers.5.self_attn.k_proj.weight": "9c98b3959ca0ec98959efd372381e45ea2d8427d5b449f33de9c2bc9240e8b95",
147
+ "model.layers.5.self_attn.o_proj.weight": "c5442e61d93947147754574cd192243a73d44ed68540061f5ac2eeba0ad810c1",
148
+ "model.layers.5.self_attn.q_norm.weight": "86dccbcc9a1b484223deea4b4ce4f02ceba1280cbf91b6e7f16b3468d1fdc50a",
149
+ "model.layers.5.self_attn.q_proj.weight": "e6128c7dbfd5c7a6646bd7d638d3c29f1f00edb99217f0d621eea74a1387a539",
150
+ "model.layers.5.self_attn.v_proj.weight": "8143fe0c8b69d1debdf8502b9b0863869b9e55cf188397a69189156d2b821ee8",
151
+ "model.layers.6.mlp.down_proj.weight": "933a695f60bba4d7fc1328911e4d392d25fd514a81bc5a0bfd7bcaa511dccc46",
152
+ "model.layers.6.mlp.gate_proj.weight": "f17b9133d594704554289b7667b98d60578075523b780bff49c4f7348a3d001e",
153
+ "model.layers.6.mlp.up_proj.weight": "ecd96261115c511b52818d03a1527f50980eaf2c116102f10e7a21d0806dfe6a",
154
+ "model.layers.6.post_attention_layernorm.weight": "851da8289f5954d41c74265b8f2ab31dbd11265c9e8f6e781982e1b0e241507e",
155
+ "model.layers.6.post_feedforward_layernorm.weight": "da4ee009a77ecf4a95201482a4419ec1a8d89229300d956ed32412a64c0fe01b",
156
+ "model.layers.6.self_attn.k_norm.weight": "159e8018eccf4c6356e379aa47e8422cd1568034eedb9fc73a99e13b7a454fd6",
157
+ "model.layers.6.self_attn.k_proj.weight": "5af407f32d446a91610506372821007629abcd0768dd1c2896ddd6464de8b3c3",
158
+ "model.layers.6.self_attn.o_proj.weight": "170195da04cd868451341361aecd20bf3e3ccb4c07af5cf50a3604a0110498a7",
159
+ "model.layers.6.self_attn.q_norm.weight": "4b93fe6794bf9a9b1f8110f2c9c95ca4b91d795119bcdbf47dd55e8fdcd03742",
160
+ "model.layers.6.self_attn.q_proj.weight": "5e8ce5439157ad66ae15f5d4352f4a29ea5aaf1d61d9a8989766270c0d09031c",
161
+ "model.layers.6.self_attn.v_proj.weight": "df6cfdccb1697f5e67b7a17f8688690d79ac04a0c3ff0970efb28992d6c120cc",
162
+ "model.layers.7.mlp.down_proj.weight": "6a3c1869029bbe7069173b817abbd0f533329288c244b2de3a8aa5d428aa1042",
163
+ "model.layers.7.mlp.gate_proj.weight": "16b1d853c61edf714e33d88bdadfa00f973910583d9d2bb130331e99351375ba",
164
+ "model.layers.7.mlp.up_proj.weight": "28532869ff12483e7866b422a6f15f8876e5ef91256dd1894ec4cdda5e01242a",
165
+ "model.layers.7.post_attention_layernorm.weight": "7f39cd35e7a4e3a35a5a48b942bfd698ad69478d221928026add10c5815f8836",
166
+ "model.layers.7.post_feedforward_layernorm.weight": "6f51f9542dbf8d6fd14a34150e2a0feda26c39d221f0208a93005136b65dc05b",
167
+ "model.layers.7.self_attn.k_norm.weight": "84f8eccfa6c4c8cf068bb0bfcc41275e7847448b82cdbb539eb73e215fb4813c",
168
+ "model.layers.7.self_attn.k_proj.weight": "b92de79b633e55972d58d629e3a84374dc6237be2593c267e1f750e148bb364b",
169
+ "model.layers.7.self_attn.o_proj.weight": "c3be8e40db0116ec137360386aafb1f6766d686be28152eef7e266fa65470939",
170
+ "model.layers.7.self_attn.q_norm.weight": "cbf4ca950e9bb14af1bf9d8b0029e5488f2d2a31ef33f5da43e37f80241bcbe7",
171
+ "model.layers.7.self_attn.q_proj.weight": "178ba7defa194e021303907ba799a6d473bd5794cf7e1befd94d26b825ceb504",
172
+ "model.layers.7.self_attn.v_proj.weight": "ecee970ed8169baecb57d0bf5cdb092e8839115f0e16c15b7e20cf8ab23dd98e",
173
+ "model.layers.8.mlp.down_proj.weight": "e43c4b5bfd85906926577dfe2a99d0a3a87eabb0303aa976d01d2e0b9869bb50",
174
+ "model.layers.8.mlp.gate_proj.weight": "ca28b1507902fedfc9139cf705a35671b4cf7f12ae553d4051db7234f0e185e0",
175
+ "model.layers.8.mlp.up_proj.weight": "6ba496db6082d95e8db38022cba79d9e6ed13bedc433f3be0e8c2d59b29b79e7",
176
+ "model.layers.8.post_attention_layernorm.weight": "eb77a257685c71c5629ac05ea1d562bc3f0de39be4d8e29202b376a9671819a4",
177
+ "model.layers.8.post_feedforward_layernorm.weight": "8132a40fc6302cc9b036661ca542f4fe53886d140270a6549958a478bc964a4c",
178
+ "model.layers.8.self_attn.k_norm.weight": "b4a745b952cd29c82c1e93a5a7e0fca0b5d6a139a79cb66bd732867cdca25ac7",
179
+ "model.layers.8.self_attn.k_proj.weight": "c15b8deb955a1d399db80caa4f86595423e03059771ba344d52382a057d6c379",
180
+ "model.layers.8.self_attn.o_proj.weight": "67b6887dba4c8acb6bc7dc73b0013ffb40e064902aff389b4a16007ae7ccc1f3",
181
+ "model.layers.8.self_attn.q_norm.weight": "1559429e81d29d29f9c562c04353afdbff586bb9ea64c2828026285a2f85b31a",
182
+ "model.layers.8.self_attn.q_proj.weight": "c52be8459340f86366dab0b48e16ce78d999ec2c625d777d0cb8cbb96d5e5873",
183
+ "model.layers.8.self_attn.v_proj.weight": "e2ca14b1dfaba9a7bf8c8eb67e8fd31539b16a5720eaa1063601527453d4c0f4",
184
+ "model.layers.9.mlp.down_proj.weight": "afe6b4074995ede19920cff75e60d2bc7aec434271754b5c9ed75d5eb03af842",
185
+ "model.layers.9.mlp.gate_proj.weight": "b9d2d50ff9509e67de634c44be2beeaeaa91457c60c1d1964f19035cda20bc5c",
186
+ "model.layers.9.mlp.up_proj.weight": "a98763c7a46f2a443645ad3c59c3c2efd877fdcaa5c713b3f28f499f21775479",
187
+ "model.layers.9.post_attention_layernorm.weight": "9a503aa458321a7d8b918a5b4f7d169468c1573f9769b685518191ccee30a0e3",
188
+ "model.layers.9.post_feedforward_layernorm.weight": "36099f76b86a8089b2e4bdafcef1ce5f3bb78adeabf0cf1e526e1e035f74eb09",
189
+ "model.layers.9.self_attn.k_norm.weight": "7fbf77d05014dde3002bb85dbac184fd6711f1b700aa43478e6044730318f316",
190
+ "model.layers.9.self_attn.k_proj.weight": "c6dd530eec028a8e43d88d432676ee396ec92660d7dcc03ca4e208eadd17aee8",
191
+ "model.layers.9.self_attn.o_proj.weight": "00b3506275e7eac378a4d905481a7c22aafb1cbbb296995529899f63851e5a6c",
192
+ "model.layers.9.self_attn.q_norm.weight": "196d884a4c7e2c0e301af23f32d278f291a8c9d155255f8e228e1aec13f945b3",
193
+ "model.layers.9.self_attn.q_proj.weight": "0741fa267bf572642c896732c58a5d807490a8b43477f1fdd7672c5983c7f374",
194
+ "model.layers.9.self_attn.v_proj.weight": "e449d8b916c0a955fbb3afa7cea3d3b777605542072f21826df30baf075764d2",
195
+ "model.norm.weight": "1f4405c246cb379ccea30048a4f45ef8890eb12055644596c106edb8f617f7ce"
196
+ },
197
+ "tokenizer_files": [
198
+ "tokenizer.json",
199
+ "tokenizer_config.json",
200
+ "special_tokens_map.json",
201
+ "vocab.json"
202
+ ],
203
+ "transformers": ">=4.57.6,<5",
204
+ "true_vocab_size": 100278,
205
+ "vocab_crop": {
206
+ "embedding": {
207
+ "exported_rows": 100278,
208
+ "removed_rows": 74,
209
+ "removed_tail_sha256": "9478ec4d7366618d530922bea53340f7ff004f464119c3635b5a0a59457b3818",
210
+ "source_rows": 100352
211
+ },
212
+ "lm_head": {
213
+ "exported_rows": 100278,
214
+ "removed_rows": 74,
215
+ "removed_tail_sha256": "ee80a918fe064a7ee878670699d9ef30fa8b8a5fa5468a5d06e5fe6f61d57e19",
216
+ "source_rows": 100352
217
+ }
218
+ }
219
+ },
220
+ "validation": {
221
+ "fc1_split_repack": "bitwise",
222
+ "finite_model_tensors": true,
223
+ "qkv_split_repack": "bitwise",
224
+ "safetensors_reload": "bitwise",
225
+ "source_key_coverage": "exact",
226
+ "source_shapes": "exact"
227
+ }
228
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": null,
4
+ "eos_token_id": 100257,
5
+ "pad_token_id": 100277,
6
+ "transformers_version": "4.57.6"
7
+ }
hf_validation_report.json ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "forward": {
3
+ "cache_no_cache_max_abs": 0.0,
4
+ "logits_finite": true,
5
+ "logits_shape": [
6
+ 1,
7
+ 4,
8
+ 100278
9
+ ],
10
+ "post_8192_yarn_finite": true
11
+ },
12
+ "parameter_count": 1484613632,
13
+ "schema": "olmo3.base.hf_validation/v1",
14
+ "stage": "stage3",
15
+ "status": "pass",
16
+ "tensor_count": 179,
17
+ "tokenizer_size": 100278,
18
+ "torch_version": "2.6.0+cpu",
19
+ "transformers_version": "4.57.6"
20
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9bfc690d349b8c2a7ea1172481013b6b09d64e9b73056a037037e2ecb0aee96
3
+ size 2969248016
special_tokens_map.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "<|endoftext|>",
3
+ "pad_token": "<|pad|>",
4
+ "unk_token": {
5
+ "content": "<|endoftext|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false
10
+ }
11
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "100256": {
5
+ "content": "<|extra_id_0|>",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": false
11
+ },
12
+ "100257": {
13
+ "content": "<|endoftext|>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "100258": {
21
+ "content": "<|fim_prefix|>",
22
+ "lstrip": false,
23
+ "normalized": false,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "100259": {
29
+ "content": "<|fim_middle|>",
30
+ "lstrip": false,
31
+ "normalized": false,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "100260": {
37
+ "content": "<|fim_suffix|>",
38
+ "lstrip": false,
39
+ "normalized": false,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ },
44
+ "100261": {
45
+ "content": "|||PHONE_NUMBER|||",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false,
50
+ "special": false
51
+ },
52
+ "100262": {
53
+ "content": "|||EMAIL_ADDRESS|||",
54
+ "lstrip": false,
55
+ "normalized": false,
56
+ "rstrip": false,
57
+ "single_word": false,
58
+ "special": false
59
+ },
60
+ "100263": {
61
+ "content": "|||IP_ADDRESS|||",
62
+ "lstrip": false,
63
+ "normalized": false,
64
+ "rstrip": false,
65
+ "single_word": false,
66
+ "special": false
67
+ },
68
+ "100264": {
69
+ "content": "<|im_start|>",
70
+ "lstrip": false,
71
+ "normalized": false,
72
+ "rstrip": false,
73
+ "single_word": false,
74
+ "special": true
75
+ },
76
+ "100265": {
77
+ "content": "<|im_end|>",
78
+ "lstrip": false,
79
+ "normalized": false,
80
+ "rstrip": false,
81
+ "single_word": false,
82
+ "special": true
83
+ },
84
+ "100266": {
85
+ "content": "<|extra_id_1|>",
86
+ "lstrip": false,
87
+ "normalized": false,
88
+ "rstrip": false,
89
+ "single_word": false,
90
+ "special": false
91
+ },
92
+ "100267": {
93
+ "content": "<|extra_id_2|>",
94
+ "lstrip": false,
95
+ "normalized": false,
96
+ "rstrip": false,
97
+ "single_word": false,
98
+ "special": false
99
+ },
100
+ "100268": {
101
+ "content": "<|extra_id_3|>",
102
+ "lstrip": false,
103
+ "normalized": false,
104
+ "rstrip": false,
105
+ "single_word": false,
106
+ "special": false
107
+ },
108
+ "100269": {
109
+ "content": "<|extra_id_4|>",
110
+ "lstrip": false,
111
+ "normalized": false,
112
+ "rstrip": false,
113
+ "single_word": false,
114
+ "special": false
115
+ },
116
+ "100270": {
117
+ "content": "<|extra_id_5|>",
118
+ "lstrip": false,
119
+ "normalized": false,
120
+ "rstrip": false,
121
+ "single_word": false,
122
+ "special": false
123
+ },
124
+ "100271": {
125
+ "content": "<|extra_id_6|>",
126
+ "lstrip": false,
127
+ "normalized": false,
128
+ "rstrip": false,
129
+ "single_word": false,
130
+ "special": false
131
+ },
132
+ "100272": {
133
+ "content": "<|extra_id_7|>",
134
+ "lstrip": false,
135
+ "normalized": false,
136
+ "rstrip": false,
137
+ "single_word": false,
138
+ "special": false
139
+ },
140
+ "100273": {
141
+ "content": "<|extra_id_8|>",
142
+ "lstrip": false,
143
+ "normalized": false,
144
+ "rstrip": false,
145
+ "single_word": false,
146
+ "special": false
147
+ },
148
+ "100274": {
149
+ "content": "<|extra_id_9|>",
150
+ "lstrip": false,
151
+ "normalized": false,
152
+ "rstrip": false,
153
+ "single_word": false,
154
+ "special": false
155
+ },
156
+ "100275": {
157
+ "content": "<|extra_id_10|>",
158
+ "lstrip": false,
159
+ "normalized": false,
160
+ "rstrip": false,
161
+ "single_word": false,
162
+ "special": false
163
+ },
164
+ "100276": {
165
+ "content": "<|endofprompt|>",
166
+ "lstrip": false,
167
+ "normalized": false,
168
+ "rstrip": false,
169
+ "single_word": false,
170
+ "special": true
171
+ },
172
+ "100277": {
173
+ "content": "<|pad|>",
174
+ "lstrip": false,
175
+ "normalized": false,
176
+ "rstrip": false,
177
+ "single_word": false,
178
+ "special": true
179
+ }
180
+ },
181
+ "bos_token": null,
182
+ "clean_up_tokenization_spaces": false,
183
+ "eos_token": "<|endoftext|>",
184
+ "extra_special_tokens": {},
185
+ "fix_mistral_regex": false,
186
+ "model_max_length": 65536,
187
+ "pad_token": "<|pad|>",
188
+ "tokenizer_class": "GPT2Tokenizer",
189
+ "unk_token": "<|endoftext|>"
190
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff