metadata

license: apache-2.0
datasets:
  - bigcode/the-stack
  - bigcode/the-stack-v2
  - bigcode/starcoderdata
  - bigcode/commitpack
library_name: transformers
tags:
  - code
base_model:
  - JetBrains/Mellum-4b-base
model-index:
  - name: Mellum-4b-sft-all
    results:
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2823
            verified: false
          - name: EM ≤ 8k
            type: exact_match
            value: 0.287
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python, 2k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2638
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python, 4k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.293
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python, 8k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.3042
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python, 12k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2685
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_python_v1.1
          name: RepoBench 1.1 (Python, 16k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2818
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2867
            verified: false
          - name: EM ≤ 8k
            type: exact_match
            value: 0.3023
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java, 2k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2883
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java, 4k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.3228
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java, 8k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2958
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java, 12k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2447
            verified: false
      - task:
          type: text-generation
        dataset:
          type: tianyang/repobench_java_v1.1
          name: RepoBench 1.1 (Java, 16k)
        metrics:
          - name: EM
            type: exact_match
            value: 0.2821
            verified: false
      - task:
          type: text-generation
        dataset:
          type: gonglinyuan/safim
          name: SAFIM
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.5285
            verified: false
      - task:
          type: text-generation
        dataset:
          type: gonglinyuan/safim
          name: SAFIM (API)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.6548
            verified: false
      - task:
          type: text-generation
        dataset:
          type: gonglinyuan/safim
          name: SAFIM (Block)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.4005
            verified: false
      - task:
          type: text-generation
        dataset:
          type: gonglinyuan/safim
          name: SAFIM (Control)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.5303
            verified: false
      - task:
          type: text-generation
        dataset:
          type: loubnabnl/humaneval_infilling
          name: HumanEval Infilling (Single-Line)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.8083
            verified: false
      - task:
          type: text-generation
        dataset:
          type: loubnabnl/humaneval_infilling
          name: HumanEval Infilling (Multi-Line)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.4819
            verified: false
      - task:
          type: text-generation
        dataset:
          type: loubnabnl/humaneval_infilling
          name: HumanEval Infilling (Random Span)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.372
            verified: false
      - task:
          type: text-generation
        dataset:
          type: loubnabnl/humaneval_infilling
          name: HumanEval Infilling (Random Span Light)
        metrics:
          - name: pass@1
            type: pass@1
            value: 0.4024
            verified: false

Model Description

Mellum-4b-sft-all is a fine-tuned version of JetBrains' first open-source large language model (LLM) optimized for code-related tasks.

Pre-trained on over 4 trillion tokens with a context window of 8192 tokens across multiple programming languages, and then fine-tuned, Mellum-4b-sft-all is tailored for context-aware code completion tasks. It was fine-tuned on a diverse set of languages, including Batchfile, C, C#, CMake, C++, CSS, Cython, Dockerfile, F#, Go, Groovy, HCL, HTML (and variants like Django, EEx, ERB, and PHP templates), Java, JSP, JavaScript, JSX, Kotlin, Less, Makefile, Objective-C++, PHP, PowerShell, Python, R, RHTML, Ruby, Rust, Sass, Scala, SCSS, Shell, SQL, Swift, TOML, TypeScript, Visual Basic, Vue, and YAML. The model follows a LLaMA-style architecture with 4 billion parameters, making it efficient for both cloud inference (e.g., via vLLM) and local deployment (e.g., using llama.cpp or Ollama).

Mellum was trained using Automatic Mixed Precision (AMP) with bf16 precision. The uploaded version on Hugging Face retains the bf16 format for public use.

Designed for integration into professional developer tooling (e.g., intelligent code suggestions in IDEs), AI-powered coding assistants, and research on code understanding and generation, Mellum is also well-suited for educational applications and fine-tuning experiments.

Limitations

Biases: May reflect biases present in public codebases. For example it will likely produce code which is similar in style to the open-source repositories.
Security: Code suggestions should not be assumed to be secure or free of vulnerabilities.

Sample Usage

Here is an example of how to run and sample from the model with additional files context and fill in the middle.

Fill in the middle with additional files as context generation

import json
from transformers import AutoTokenizer, AutoModelForCausalLM

example = """
<filename>Utils.kt
package utils

fun multiply(x: Int, y: Int): Int {
    return x * y
}

<filename>Config.kt
package config

object Config {
    const val DEBUG = true
    const val MAX_VALUE = 100
}

<filename>Example.kt
<fim_suffix>
fun main() {
    val result = calculateSum(5, 10)
    println(result)
}
<fim_prefix>fun calculateSum(a: Int, b: Int): Int {
<fim_middle>
"""

tokenizer = AutoTokenizer.from_pretrained('JetBrains/Mellum-4b-sft-all')
model = AutoModelForCausalLM.from_pretrained('JetBrains/Mellum-4b-sft-all')
encoded_input = tokenizer(example, return_tensors='pt', return_token_type_ids=False)
out = model.generate(
    **encoded_input,
    max_new_tokens=100,
)

Benchmarks

We are providing scores for Mellum‑4b‑sft‑all to give users an estimate of the model’s potential capabilities.

RepoBench 1.1

Type: single‑line Languages: Python and Java Metric: Exact Match (EM), %

Since Mellum has a maximum context window of 8 k, we report both the average over all evaluated context lengths (2 k, 4 k, 8 k, 12 k and 16 k) and the average over the lengths within its supported range (≤ 8 k).

Python subset

Model	2 k	4 k	8 k	12 k	16 k	Avg	Avg ≤ 8 k
Mellum‑4b‑sft‑all	26.38%	29.30%	30.42%	26.85%	28.18%	28.23%	28.70%

Java subset

Model	2 k	4 k	8 k	12 k	16 k	Avg	Avg ≤ 8 k
Mellum‑4b‑sft‑all	28.83%	32.28%	29.58%	24.47%	28.21%	28.67%	30.23%

Syntax‑Aware Fill‑in‑the‑Middle (SAFIM)

Type: mix of multi‑line and single‑line Languages: multi‑language Metric: pass@1, %

Model	Algorithmic	Control	API	Average
Mellum‑4b‑sft‑all	40.05%	53.03%	65.48%	52.85%

HumanEval Infilling

Type: single‑line and multi‑line
Languages: Python
Metric: pass@1, %

Model	Single‑Line	Multi‑Line	Random Span
Mellum‑4b‑sft‑all	80.83%	48.19%	37.20%

Citation

If you use this model, please cite:

@misc{Mellum-4b-base,
  title     = {Mellum-4b-base},
  author    = {Pavlichenko, Nikita and Nazarov, Iurii and Dolgov, Ivan and Garanina, Ekaterina and Lasocki, Karol and Reshetnikova, Julia and Boitsov, Sergei and Bondyrev, Ivan and Karaeva, Dariia and Sheptyakov, Maksim and Ustalov, Dmitry and Mukhin, Artem and Proshev, Semyon and Abramov, Nikita and Kolomyttseva, Olga and Lysaniuk, Kseniia and Zavidnyi, Ilia and Semenkin, Anton and Tankov, Vladislav and Sazanovich, Uladzislau},
  year      = {2025},
}

Contact

For questions, collaborations and requests reach us out via [email protected]