Qwen/Qwen2-1.5B-Instruct
| benchmark | category | score | context | conf | evidence |
|---|---|---|---|---|---|
| C-Eval | knowledge | 37.2 | C-Eval | medium | | C-Eval | 37.2 | **45.2** | 55.3 | **63.8** | |
| MMLU | knowledge | 35.0 | MMLU | medium | | MMLU | 35.0 | **37.9** | 43.7 | **52.4** | |
| IFEval | instruction-following | 14.6 | IFEval (Prompt Strict-Acc.) | medium | | IFEval (Prompt Strict-Acc.) | 14.6 | **20.0** | 16.8 | **29.0** | |
| GSM8K | math | 11.3 | GSM8K | medium | | GSM8K | 11.3 | **40.1** | 35.3 | **61.6** | |
| HumanEval | coding | 9.1 | HumanEval | medium | | HumanEval | 9.1 | **17.1** | 25.0 | **37.8** | |
| title | author | status | comments | type |
|---|---|---|---|---|
| Qwen2 1.5B — the predecessor that proved small models work | 3morixd | open | 0 | disc |
| Listed on OpenModelMap | duola15 | open | 0 | disc |
| TemporalMesh Transformer: 29.4 PPL at 48% compute — beats Mamba, new open-source architecture | vigneshwar234 | open | 0 | disc |
| Add EvalEval community eval results | EvalEvalBot | open | 0 | disc |
| Update README.md | cherry0328 | open | 0 | disc |
| Add Space in chat template. Not sure, if correct. | ThiloteE | open | 0 | disc |
| Error. Crash. "The attention mask is not set and cannot be inferred from input | MartialTerran | open | 0 | disc |
| Qwen2 sample model.py does not work. | MartialTerran | closed | 0 | disc |
| Add base_model metadata | davanstrien | open | 0 | disc |
| [AUTOMATED] Model Memory Requirements | model-sizer-bot | open | 0 | disc |
| Issue with RuntimeError when Loading Fine-Tuned qwen2-1.5-instruct Model | Ganz00 | open | 0 | disc |
| lm_head.weight is missing | DeepMount00 | open | 0 | disc |
| model multi-turn? | NickyNicky | open | 0 | disc |
| Upload ONNX weights | Xenova | open | 0 | disc |