File size: 3,229 Bytes
0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f 0818b0a ab7710f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 |
---
library_name: transformers
tags:
- small-lm
- math
- code
- reasoning
- slm
- french
- merge
license: apache-2.0
base_model:
- SamsungSAILMontreal/Qwen3-4B-Instruct-2507
- SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Math
- SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Code
- SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Fr
language:
- fr
- en
---
# Qwen3-4B-Instruct-2507-Math-Fr
This model is obtained by merging [SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Math](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Math),
[SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Code](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Code) and
[SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Fr](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Fr).
The model is used in the experiments described in https://bknyaz.github.io/blog/2026/meta-merge/.
Single A100 was used for merging and evaluation.
The following versions were used for merge/eval:
- python >= 3.10
- torch : 2.9.0+cu128
- lm_eval : 0.4.9.1
- vllm : 0.11.1
- transformers : 4.57.6
- datasets : 3.2.0
- numpy : 2.2.6
## Merging
Merging was done using parameter averaging implemented in [merge_qwen.py](https://github.com/SamsungSAILMontreal/nino/blob/main/merge_qwen.py).
## Evaluation
Evaluation was done with lm_eval on the test split of [gsm8k](https://huggingface.co/datasets/openai/gsm8k),
[french_bench (avg score)](https://github.com/EleutherAI/lm-evaluation-harness/tree/main/lm_eval/tasks/french_bench),
[gsm8k-fr](https://huggingface.co/datasets/cmh/gsm8k_fr) and humaneval (instruct):
```bash
python -m lm_eval --model vllm --model_args pretrained=${model},tensor_parallel_size=1,dtype=auto,gpu_memory_utilization=0.9,data_parallel_size=1 \
--tasks gsm8k,french_bench,gsm8k-fr,humaneval_instruct --batch_size 1 --apply_chat_template=True --confirm_run_unsafe_code --trust_remote_code
```
To evaluate on gsm8k-fr you can use our fork https://github.com/bknyaz/lm-evaluation-harness/tree/main/lm_eval/tasks/gsm8k.
### Results
| Model | gsm8k | french | gsm8k-fr | humaneval_instruct | avg |
|-------------------------------------|--------|--------|----------|--------------------|------|
| Qwen3-4B-Instruct-2507 | 80.4 | 43.1 | 66.0 | 90.2 | 69.9 |
| Qwen3-4B-Instruct-2507-Math | 76.8 | 43.0 | 65.3 | 72.0 | 64.3 |
| Qwen3-4B-Instruct-2507-Fr | 72.3 | 45.7 | 60.7 | 74.4 | 63.3 |
| Qwen3-4B-Instruct-2507-Code | 72.5 | 45.4 | 53.0 | 76.2 | 61.8 |
| Qwen3-4B-Instruct-2507-Math-Code-Fr | 82.9 | 45.8 | 69.8 | 79.9 | 69.6 |
## License
Please refer to the license of the base models [SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Math](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Math),
[SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Code](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Code) and
[SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Fr](https://huggingface.co/SamsungSAILMontreal/Qwen3-4B-Instruct-2507-Fr). |