intel / intel/auto-round

[Bug]: MiniMax-M2.7_mxfp4_mixed_fp8kv_LLMC vllm eval failed

Open
#2,369 0 comments 0 reactions 1 assignee Claimed by @xin3he View on GitHub
bug
Dominant language
Python
Stars
1.6k
Forks
175
Avg merge
1d 18h
Merged PRs (30d)
99

Description

### Problem Description

MiniMax-M2.7_mxfp4_mixed_fp8kv_LLMC vllm eval failed

https://inteltf-jenk.sh.intel.com/job/Intel-GPU-key-model-test/1183/artifact/evaluation_MiniMax-M2.7_mxfp4_mixed.log

### Reproduction Steps

cd /software/data/jenkins/CS09_B200/workspace/Intel-GPU-key-model-test/neural-compressor/examples/pytorch/nlp/huggingface_models/language-modeling/quantization/auto_round/minimax_m2.7
bash run_evalscope.sh --model /home/hshen/suyue/saved_models/MiniMax-M2.7_mxfp4_mixed_fp8kv_LLMC/MiniMax-M2.7-mxfp-w8g32 --tp 2 --tasks gpqa_diamond,aime25,gsm8k,piqa

### Environment Information

_No response_

### Error Logs

```shell
AttributeError: Layer model.layers.0.mlp.experts has no parameter 'w13_weight_packed' for checkpoint weight 'model.layers.0.mlp.experts.0.w1.weight_packed'
```

### Additional Context

_No response_

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.