Downloads · 30 days
31
16% of all-time downloads
ibnsina-3b-checkpoints
ibnsina-llm/ibnsina-3b-checkpoints
ibnsina-3b-checkpoints is a machine learning model from ibnsina-llm. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
These are not the files you want for running the model. For that, use ibnsina-llm/ibnsina-3b (GGUF for llama.cpp / ollama / LM Studio: ollama run ibnsina/ibnsina-3b).
All-time downloads
190
Public
Repo size
130 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.pt96.7 GB · 75%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- Sep 10, 2026
- Updated
- Sep 10, 2026
- SHA
- 6b08a61c
Model card
From the Hugging Face model README
ibnsina-3b-checkpoints — raw training checkpoints of IbnSina-3B
These are not the files you want for running the model. For that, use ibnsina-llm/ibnsina-3b (GGUF for llama.cpp / ollama / LM Studio: ollama run ibnsina/ibnsina-3b).
This repository holds the raw nanochat-format checkpoints (.pt, PyTorch pickles of tensors) of the IbnSina-3B pretraining and fine-tuning runs, published for reproducibility, research and fine-tuning from the base. Architecture: Qwen3-style (40 layers, d=2560, 20 heads / 4 KV heads, QK-norm), 3.0 B parameters, 2048 context, tokenizer v2_32k_llama (32,768 BPE; shared with IbnSina-1.5B). Loading requires nanochat with the IbnSina patches (training/nanochat_patches/, NANOCHAT_ARCH=qwen3) from github.com/ibnsina-llm.
| path | what it is | notes |
|---|---|---|
base/big3b_190000/model_190000.pt + meta_190000.json | final pretraining checkpoint (step 190,000; 99.6 B tokens; val bpb 0.4759) | the model every SFT below starts from |
base/big3b_190000/optim/optim_190000_rank{0..7}.pt | Muon/AdamW optimizer state, 8 rank shards | for warm restarts |
base/tokenizer_TRAINING.tar.gz | the exact tokenizer files used in training (tokenizer.pkl md5 7e262b94, token_bytes.pt) | the canonical tokenizer; do not mix with the pilot's v1_32k |
sft/v2_published/ | chat SFT v2 (sft/v2 + MMLU auxiliary, 85 steps) — the model published on 2026-09-09 08:20 JST | evals in results_big3b.json |
sft/v21h_shipped/ | chat SFT v2.1 — the currently released weights (v2 data + 4,499 synthetic NLI/paraphrase/format rows + identity rows; 8×H100 run) | results_big3b_v21h.json, diag_cls_big3b_v21h.json |
sft/v21_seed2_unpublished/ | alternate SFT v2.1 run — NOT the release. Same data and recipe on one A100 (micro-batch 2); it did not reproduce the entailment gain (30.0 % vs 39.9 %). Kept as evidence that the v2.1 recipe is unstable | results_big3b_v21.json, diag_cls_big3b_v21_seed2.json |
release/ibnsina-3b/, release/ibnsina-3b-v21h/, release/ibnsina-3b-v21/ | the GGUF release bundles built from the three SFT checkpoints (v2 with F16 — do not use F16 on CUDA, it overflows; v2.1h = the public repo; v2.1 seed2 unpublished) | |
MANIFEST.sha256 | sha256 of every file | verify after download |
Evaluation numbers, the classification diagnostic and the v2.1 non-replication are documented on the model card and in the forthcoming family technical report. Licence: Apache-2.0 (weights and code). Author: Sina Meraji · ORCID 0009-0002-8028-1932.
- License
- apache-2.0
- Languages
- fa, en
- Created
- Sep 10, 2026
- Updated
- Sep 10, 2026