"RuntimeError: weight lm_head.weight does not exist" quantizing Llama-3.2-11B-Vision-Instruct

[akowalsk]

System Info

Running official docker image: ghcr.io/huggingface/text-generation-inference:2.4.0

os: Linux 5.15.0-124-generic #134-Ubuntu SMP Fri Sep 27 20:20:17 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux

nvidia-smi:
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.90.07 Driver Version: 550.90.07 CUDA Version: 12.4 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 3090 On | 00000000:01:00.0 Off | N/A |
| 0% 27C P8 23W / 350W | 2MiB / 24576MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
| 1 NVIDIA GeForce RTX 3090 On | 00000000:21:00.0 Off | N/A |
| 0% 28C P8 21W / 350W | 2MiB / 24576MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
| 2 NVIDIA GeForce RTX 3090 On | 00000000:4B:00.0 Off | N/A |
| 0% 28C P8 21W / 350W | 2MiB / 24576MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
| 3 NVIDIA GeForce RTX 3090 On | 00000000:4C:00.0 Off | N/A |
| 0% 27C P8 19W / 350W | 2MiB / 24576MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

Information

• Docker
• The CLI directly

Tasks

• An officially supported command
• My own modifications

Reproduction

I'm attempting to quantize the Llama 3.2 Vision model and get the error "RuntimeError: weight lm_head.weight does not exist"

I'm using the following command:

docker run --gpus all --shm-size 1g -e HF_TOKEN=REDACTED -v $(pwd):/data --entrypoint='' ghcr.io/huggingface/text-generation-inference:2.4.0 text-generation-server quantize meta-llama/Llama-3.2-11B-Vision-Instruct /data/Llama-3.2-11B-Vision-Instruct-GPTQ-INT4

I have attached the full output.

tgi_quantize_error.txt

Expected behavior

I would like the quantization process to succeed. I couldn't find any specific reference to whether multi-modal models work with GPTQ quantization or not.