File size: 1,394 Bytes
5afa639 5b29e4c 5afa639 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 | ---
library_name: transformers
pipeline_tag: text-generation
tags:
- olmo3
- baseline
- safetensors
- sliding-window-attention
---
# OLMo 3 1B Baseline — Stage 1 Pretraining
This repository contains the pure OLMo 3 1B baseline checkpoint from
`o3b1b-s8192-g8192-m2-ga4-tp1-cp1-dp1024-i1024-b4-lr1e3-fresh-1024npu-share-20260726-v1` at iteration `89407`. SiameseNorm and Depth-Attention
are disabled.
- Training sequence length: 8,192
- Model context capacity: 8,192
- Sliding-window size: 4,096
- Attention pattern: `[SWA, SWA, SWA, Full]`
- Vocabulary: 100,278 real tokens; 74 Megatron padding rows removed
Stage 3/4 apply YaRN only to Full-Attention layers. OLMo 3 SWA layers use the
original RoPE and retain their 4,096-token local window.
## Loading
`transformers>=4.57.6,<5` is required.
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
repo_id = "ArchSpace-Collection/OLMo3-1B-SiameseNorm-DepthAttention-baseline-stage1"
tokenizer = AutoTokenizer.from_pretrained(
repo_id,
use_fast=True,
fix_mistral_regex=False,
)
model = AutoModelForCausalLM.from_pretrained(
repo_id,
dtype=torch.bfloat16,
attn_implementation="sdpa",
)
```
`fix_mistral_regex=False` preserves the tokenizer behavior used for training.
The checkpoint uses the official Transformers `Olmo3ForCausalLM`
implementation and does not require remote code.
|