SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2401.00788 · 2024

Astraios: Parameter-Efficient Instruction Tuning Code Large Language Models

arXiv · PDF · Open in the Atlas

Code that ran

7 of 7 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 6 failures of our instrument; 0 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2401.00788.json · badge: /api/ran/2401.00788.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
bigcode-project/astraios canonical (classifier label)
classifier basis: classifier_2025
3 ran · 0 no recorded run
copy not recorded community
not in the paper's declared repositories
4 ran · 0 no recorded run
FunctionStatusWhere it lives
chars_token_ratio Ran
ran, honoured the contract we drafted
bigcode-project/astraios/training/finetune.py
code served (permissive licence) · get_code("96de41c883c409e1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np
from transformers import AutoTokenizer

# Create a mock tokenizer class that satisfies the function's needs
class MockTokenizer:
    def __init__(self, is_fast=True):
        self.is_fast = is_fast
    
    def __call__(self, text):
        # Return an object with a tokens() method
        return MockTokenizerOutput(text)
    
    def tokenize(self, text):
        # For non-fast tokenizers, split by whitespace
        return text.split()

class MockTokenizerOutput:
    def __init__(self, text):
        self.text = text
    
    def tokens(self):
        # Simple tokenization: split by whitespace
        return self.text.split()

# Create a mock dataset
class MockDataset:
    def __init__(self, examples):
        self.examples = examples
    
    def __iter__(self):
        return iter(self.examples)

# Create sample data
dataset = MockDataset([
    {
        "question": "What is machine learning?",
        "answer": "Machine learning is a subset of artificial intelligence."
    },
    {
        "question": "What is deep learning?",
        "answer": "Deep learning uses neural networks with multiple layers."
    },
    {
        "question": "What is Python?",
        "answer": "Python is a popular programming language."
    },
    {
        "question": "What is transformers?",
        "answer": "Transformers are a type of neural network architecture."
    }
])

# Create tokenizer
tokenizer = MockTokenizer(is_fast=True)

ARGS = {
    "dataset": dataset,
    "tokenizer": tokenizer,
    "input_column_name": "question",
    "output_column_name": "answer",
    "nb_examples": 4
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch, tqdm, transformers — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
generate_prompt Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from xichen-fy/Fira
pointer only · get_code("56e674752db00649")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

generate_prompt Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from agi-edgerunners/llm-adapters
pointer only · get_code("0dbc7fc6c518d0fb")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

generate_prompt Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from agi-edgerunners/llm-adapters
pointer only · get_code("1c6b8be890b6118a")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

load_data Ran
ran; our contract draft was wrong, not the code
this paper's copy was not recorded; identical code first harvested from agi-edgerunners/llm-adapters
pointer only · get_code("56a9466ff35b35bb")
How it ran

The body of this sample is withheld: this paper's copy of the code was not recorded (its link names no repository or source path), so no licence was read for it. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

prepare_sample_text Ran
ran; our contract draft was wrong, not the code
bigcode-project/astraios/training/finetune.py
code served (permissive licence) · get_code("1163ae6f8c7ed82a")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import numpy as np

# Create a minimal example dictionary that has the required fields
example = {
    'question': 'What is the capital of France?',
    'answer': 'Paris'
}

# Define arguments for the function
ARGS = {
    'example': example,
    'input_column_name': 'question',
    'output_column_name': 'answer'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy, torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
print_trainable_parameters Ran
ran; our contract draft was wrong, not the code
bigcode-project/astraios/training/finetune.py
code served (permissive licence) · get_code("a97c88f3a8082897")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned str. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
import torch.nn as nn

torch.manual_seed(0)

class SimpleModel(nn.Module):
    def __init__(self):
        super().__init__()
        self.linear1 = nn.Linear(4, 3)
        self.linear2 = nn.Linear(3, 2)
        # Freeze some parameters to have non-trainable params
        self.linear2.weight.requires_grad = False
    
    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

model = SimpleModel()

ARGS = {
    'model': model
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type str — a str output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a str output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block

Also present in other papers' code

Of 7 samples lifted from this paper's linked repositories, 7 carry a function identity and 0 do not. Of the 7 distinct functions with an identity, 5 are also present in other papers whose code we hold (4 with at least one paper dated earlier, 0 the same month, 5 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

FunctionAlso present inEarliest-datedOthers named
generate_prompt 15 other papers
2 dated earlier, 12 later, 1 same year, order not determinable
2304.01933 (2023-04)
LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models
2307.08072 (2023-07)
Do Emergent Abilities Exist in Quantized Large Language Models: An Empirical Study
2024.naacl-long.132 (2024 (year only))
LLM-based Medical Assistant Personalization with Short-and Long-Term Memory Coordination
and 12 more not named here
load_data 10 other papers
1 dated earlier, 9 later
2304.01933 (2023-04)
LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models
2402.09353 (2024-02)
DoRA: Weight-Decomposed Low-Rank Adaptation
2406.11909 (2024-06)
Mixture-of-Subspaces in Low-Rank Adaptation
and 7 more not named here
generate_prompt 9 other papers
1 dated earlier, 8 later
2304.01933 (2023-04)
LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models
2402.09353 (2024-02)
DoRA: Weight-Decomposed Low-Rank Adaptation
2406.11909 (2024-06)
Mixture-of-Subspaces in Low-Rank Adaptation
and 6 more not named here
generate_prompt 5 other papers
1 dated earlier, 4 later
2304.01933 (2023-04)
LLM-Adapters: An Adapter Family for Parameter-Efficient Fine-Tuning of Large Language Models
2402.09353 (2024-02)
DoRA: Weight-Decomposed Low-Rank Adaptation
2409.01035 (2024-09)
Task-Specific Directions: Definition, Exploration, and Utilization in Parameter Efficient Fine-Tuning
and 2 more not named here
print_trainable_parameters 1 other paper
1 later
2403.13592 (2024-03)
Llama meets EU: Investigating the European Political Spectrum through the Lens of LLMs

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The high cost of full-parameter fine-tuning (FFT) of Large Language Models (LLMs) has led to a series of parameter-efficient fine-tuning (PEFT) methods. However, it remains unclear which methods provide the best cost-performance trade-off at different model scales. We introduce Astraios, a suite of 28 instruction-tuned OctoCoder models using 7 tuning methods and 4 model sizes up to 16 billion parameters. Through investigations across 5 tasks and 8 different datasets encompassing both code comprehension and code generation tasks, we find that FFT generally leads to the best downstream performance across all scales, and PEFT methods differ significantly in their efficacy based on the model scale. LoRA usually offers the most favorable trade-off between cost and performance. Further investigation into the effects of these methods on both model robustness and code security reveals that larger models tend to demonstrate reduced robustness and less security. At last, we explore the relationships among updated parameters, cross-entropy loss, and task performance. We find that the tuning effectiveness observed in small models generalizes well to larger models, and the validation loss in instruction tuning can be a reliable indicator of overall downstream performance.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2401.00788")
get_code_for_paper("2401.00788")
have("2401.00788")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2401.00788.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2401.00788.svg)](https://syntology.ai/paper/2401.00788)

Connect an agent — have() is free.