Modelpublic

CodeXomics/CodeXomics-ToolAgent-4B-v1

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face

0likes49downloads

Model Card

CodeXomics-ToolAgent-4B-v1

CodeXomics-ToolAgent-4B-v1 (internally qwen3.5:4b-codexomics-tools-v5) is a 4.2B-parameter tool-calling model fine-tuned from Qwen/Qwen3.5-4B for the CodeXomics genomics workbench. It performs native function calling against the CodeXomics tool registry (file loading, navigation, sequence analysis, annotation, track control, export, BLAST, primer design, database/protein retrieval, task management, and UI control).

CodeXomics is an AI-native genome browser: a cross-platform desktop application in which conversational AI agents drive genome visualization and run real biological analyses, with built-in tool execution, a dynamic tool registry, plugin development, and Model Context Protocol (MCP) integration. Source code and documentation: github.com/Scilence2022/CodeXomics and scilence2022.github.io/CodeXomics.

Fine-tuning

—Method: QLoRA (rank 16, scale 32.0, dropout 0.05, 4 layers) with MLX-LM 0.31.2 / MLX 0.32.0
—Trainable parameters: 4.058M (0.096%)
—Optimizer: AdamW, learning rate 1.0e-5, effective batch size 4, 200 iterations
—Maximum sequence length: 3,072 tokens; prompt masking enabled
—Hardware: Apple M3 Max; peak memory 191 GB (including swap)
—Checkpoint selection: iteration 75 (validation loss 0.020); test loss 0.074 (perplexity 1.077)
—Training data: CodeXomics-ToolCalling-v1 (373/123/30 train/validation/test examples)

Evaluation

On the CodeXomics Benchmark (172 automatic tests: 143 single-operation, 29 multi-step), evaluated in the real application loop with task-completion scoring plus execution evidence. Both the fine-tuned model and the un-fine-tuned Qwen3.5-4B baseline were evaluated in three independent complete sessions; results were identical across sessions for both models.

Suite	Qwen3.5-4B (base)	CodeXomics-ToolAgent-4B-v1
Simple	139/143	143/143
Complex	26/29	29/29
Total	165/172 (95.9%)	172/172 (100%)

Fine-tuning improved the overall accuracy by 7 tests (+4 simple, +3 complex). Inference settings: temperature 0, thinking enabled.

Inference speed (mean ± SD over three independent runs; offline harness, 172 tests, Ollama Q4KM on Apple M3 Max):

Metric	Qwen3.5-4B (base)	CodeXomics-ToolAgent-4B-v1
Average latency per test (s)	12.0 ± 0.1	10.7 ± 0.9
Simple suite (s)	9.2 ± 0.1	8.6 ± 0.7
Complex suite (s)	25.8 ± 0.3	21.5 ± 1.9
Average latency per tool call (s)	7.9 ± 0.1	7.3 ± 0.6
Generation throughput (tokens/s)	33.9 ± 0.4	33.9 ± 2.8
Generated tokens per test	408 (identical across runs)	363 (identical across runs)
Prompt tokens per test	9,004 (identical across runs)	8,777 (identical across runs)

[image]

Figure 1. Two-panel comparison between the un-fine-tuned Qwen3.5-4B baseline and CodeXomics-ToolAgent-4B-v1. Left: CodeXomics Benchmark pass rates (simple/complex/total; three sessions each, identical results). Right: average inference latency per test with error bars (mean ± SD over three runs; Ollama Q4KM, Apple M3 Max). Latency values are the mean of the per-suite rows above; error bars show the run-to-run SD.

Usage

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("CodeXomics-ToolAgent-4B-v1")
tokenizer = AutoTokenizer.from_pretrained("CodeXomics-ToolAgent-4B-v1")

For deployment in CodeXomics, the model is served through Ollama as qwen3.5:4b-codexomics-tools-v5 (Q4KM, 2.7 GB) with native tool calling and thinking enabled.

Limitations

—The model is specialized for CodeXomics genomic workflows; generalization to other tool-calling domains was not evaluated.

Citation

bibtex

@software{codexomics-toolagent-v1,
  title = {CodeXomics-ToolAgent-4B-v1},
  author = {Song, Lifu},
  year = {2026},
  license = {Apache-2.0},
  publisher = {Hugging Face},
  base_model = {Qwen/Qwen3.5-4B}
}