LittleKedi-tiny-wide-sparse-base

LittleKedi-tiny-wide-sparse-base is a small DeepSeek-V3-style Mixture-of-Experts language model trained from scratch. It has 3.8M parameters in total, of which 1.6M are active per token (0.52M excluding the embedding table). It was pretrained on a 1,000M-token mix (FineWeb-Edu, Cosmopedia (Stanford), Cosmopedia (OpenStax), Cosmopedia (WikiHow), Cosmopedia (Khan Academy), FineMath 4+, SmolTalk, github-code-clean).

Part of the LittleKedi family. Not affiliated with DeepSeek.

This is the base model, before any instruction tuning. It continues text; it doesn't follow instructions. It's an educational and research artifact: it writes fluent, on-topic English but knows few facts and confidently makes things up. Don't use it for anything that matters.

The LittleKedi family

LittleKedi is a family of small DeepSeek-V3-style Mixture-of-Experts language models trained from scratch. Every size comes in three variants that differ in how sparse they are: how many routed experts they have, and what share of their parameters a token actually passes through.

More experts means more total capacity to store knowledge, but each expert sees fewer tokens (tokens × top-k ÷ experts per layer), so sparser variants need more training data before their experts are well trained. The variants of a size share a tokenizer and dataset, so they can be compared directly. Shapes change as the family scales; the table below is generated from the current presets.

This card is for tiny-wide-sparse, the wide-sparse tiny model: 1.6M of its 3.8M parameters (42%) are active per token. Its siblings are tiny and tiny-sparse.

"Active" counts everything a token passes through: embeddings, attention, the dense layer, shared experts and its top-k routed experts. "Active non-embedding" leaves out the embedding table (tied with the output head), which costs memory but almost no compute; it's the fair number for comparing compute between models.

Architecture

Training

Evaluation

Validation loss over training: 5.59 (16M tok) → 4.07 (164M tok) → 3.98 (295M tok) → 3.93 (442M tok) → 3.91 (590M tok) → 3.89 (737M tok) → 3.84 (868M tok) → 3.74 (1,000M tok)

Zero-shot benchmarks (evals.py)

Each answer choice is scored by the log-probability of its text after the prompt, using lm-evaluation-harness prompts (SciQ includes its supporting passage). acc_norm divides by the answer's length in characters; acc doesn't. Splits: SciQ test, ARC-Easy test, HellaSwag validation.

MMLU (mmlu_eval.py)

Biology & medicine (cloze): 27.8% on 2,089 questions, +2.8 points against 25% chance (about 2.9 standard errors, significant). This pools the 10 life-science and health subjects; it isn't an official MMLU category.

In the letter format the model compares " A"–" D", with same-subject examples added while they fit the context. In the cloze format each answer's text is scored by log-probability per byte; at this size the cloze results are the meaningful ones. Chance is 25%.

Compared with its siblings and other small models

LittleKedi columns are measured with this repo's evals.py and mmlu_eval.py on the checkpoints named. Validation losses are only comparable between models that share a tokenizer and validation set. The other columns are published figures: Pythia-70M from EleutherAI's evaluations as quoted on the Wisp-15M card, GPT-2 and Supra-50M from the Supra-50M-Base card. Their prompts may differ from ours, so treat gaps of a few points as noise.

What it can and can't do

This is the sparsest tiny model: the same 0.5M active non-embedding parameters per token as tiny, but with 64 narrower experts and ~3× tiny's expert capacity. It models text measurably better than tiny (validation loss 3.736 against 3.869 on the same data, matching tiny's loss with roughly a third of the tokens), but in use it's a model of the same kind: it has learned the form of English much better than its content. It's a research artifact for studying sparsity in small MoE models, not a source of information.

What it does

• ✅ Writes grammatical, fluent sentences for a paragraph or two, in the register of educational web text and textbooks.
• ✅ Stays roughly on topic, and sometimes gets the broad frame right where tiny didn't: "The French Revolution began in" → "…the late 18th century" (with sampling; tiny said 1906). Treat this as an occasional improvement, not a reliable one: the same prompt decoded greedily gives "the late 19th century".
• ✅ Picks up document structure from its data mix: markdown headings and # math: blocks after maths prompts, bullet lists, dialogue turns after User: / Assistant: text.
• ✅ Benchmark signal above chance, on a par with tiny (SciQ 53.9%, ARC-Easy 32.4%, MMLU biology & medicine 28% in the cloze format). Its lower loss hasn't turned into higher multiple-choice accuracy at this size.

What it can't do

• ❌ Facts. Names and dates are plausible in kind but usually wrong: "Albert Einstein was" → "a professor of medicine at the University of California"; "DNA is made of" → "a coin, a piece of paper". Treat every name, date and number it writes as made up.
• ❌ Arithmetic. "2 + 2 =" → "6", "7 + 6 =" → "6".
• ❌ Code. Given a Python function signature it produces docstring fragments and whitespace.
• ❌ Instructions or questions. It's a base model: it continues text rather than answering it.
• ❌ Long-range coherence. Topics drift after a sentence or two ("Photosynthesis is the process by which a person's body is able to move in a controlled way").

Decoding

Greedy decoding falls into loops even faster than tiny's ("the heart, the heart, the heart…"). Use sampling with temperature ≈ 0.7, top-p ≈ 0.9 and a repetition penalty of about 1.15.

Usage

PyTorch (needs the littlekedi package from the training code):

import json, torch
from safetensors.torch import load_file
from littlekedi import LittleKediModel, ModelConfig
from littlekedi.tokenizer import BPETokenizer

model = LittleKediModel(ModelConfig.from_dict(json.load(open("config.json")))).eval()
model.load_state_dict(load_file("model.safetensors"))
tok = BPETokenizer.from_file("tokenizer.json")
ids = torch.tensor([[tok.eot_id] + tok.encode("Photosynthesis is")])
print(tok.decode(model.generate(ids, 60, temperature=0.7, top_p=0.9, repetition_penalty=1.15)[0].tolist()))

Or interactively: python sample.py --ckpt <checkpoint>.pt. Base models are released as PyTorch weights only; GGUF builds come with the instruct models.

Files

Acknowledgements

• The architecture follows DeepSeek-V3 (DeepSeek-AI, 2024): MLA, DeepSeekMoE and auxiliary-loss-free balancing.
• Muon (Keller Jordan et al., 2024) with Moonlight's update scaling (Liu et al., 2025).
• Pretraining data: FineWeb-Edu (ODC-BY 1.0).
• Pretraining data: Cosmopedia (Apache 2.0).
• Pretraining data: FineMath 4+ (ODC-BY 1.0).
• Pretraining data: SmolTalk (Apache 2.0 for its new subsets, with the datasets it includes under their own licenses).
• Pretraining data: github-code-clean (Apache 2.0, with each file under its own open-source license).

Card generated 2026-10-10 by prepare_release.py from tiny-wide-sparse.pt (step 15257).