self.labels not set in eval_forward() for huggingfacemodel if self.use_logits=False

[andydelworth]

** Environment **

Collecting system information...
---------------------------------
System Environment Report        
Created: 2024-09-20 23:30:14 UTC
---------------------------------

PyTorch information
-------------------
PyTorch version: 2.3.1+cu121
Is debug build: False
CUDA used to build PyTorch: 12.1
ROCM used to build PyTorch: N/A

OS: Ubuntu 22.04.4 LTS (x86_64)
GCC version: (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
Clang version: Could not collect
CMake version: version 3.30.0
Libc version: glibc-2.35

Python version: 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0] (64-bit runtime)
Python platform: Linux-6.5.0-26-generic-x86_64-with-glibc2.35
Is CUDA available: True
CUDA runtime version: 12.5.82
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration: 
GPU 0: NVIDIA A100 80GB PCIe
GPU 1: NVIDIA A100 80GB PCIe
GPU 2: NVIDIA A100 80GB PCIe
GPU 3: NVIDIA A100 80GB PCIe
GPU 4: NVIDIA A100 80GB PCIe
GPU 5: NVIDIA A100 80GB PCIe
GPU 6: NVIDIA A100 80GB PCIe
GPU 7: NVIDIA A100 80GB PCIe

Nvidia driver version: 535.154.05
cuDNN version: Probably one of the following:
/usr/lib/x86_64-linux-gnu/libcudnn.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_adv.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_cnn.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_precompiled.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_engines_runtime_compiled.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_graph.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_heuristic.so.9.2.1
/usr/lib/x86_64-linux-gnu/libcudnn_ops.so.9.2.1
HIP runtime version: N/A
MIOpen runtime version: N/A
Is XNNPACK available: True

CPU:
Architecture:                       x86_64
CPU op-mode(s):                     32-bit, 64-bit
Address sizes:                      48 bits physical, 48 bits virtual
Byte Order:                         Little Endian
CPU(s):                             128
On-line CPU(s) list:                0-127
Vendor ID:                          AuthenticAMD
Model name:                         AMD EPYC 7513 32-Core Processor
CPU family:                         25
Model:                              1
Thread(s) per core:                 2
Core(s) per socket:                 32
Socket(s):                          2
Stepping:                           1
Frequency boost:                    enabled
CPU max MHz:                        3681.6399
CPU min MHz:                        1500.0000
BogoMIPS:                           5200.27
Flags:                              fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 fma cx16 pcid sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 invpcid_single hw_pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 invpcid cqm rdt_a rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoinvd amd_ppin brs arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca
Virtualization:                     AMD-V
L1d cache:                          2 MiB (64 instances)
L1i cache:                          2 MiB (64 instances)
L2 cache:                           32 MiB (64 instances)
L3 cache:                           256 MiB (8 instances)
NUMA node(s):                       2
NUMA node0 CPU(s):                  0-31,64-95
NUMA node1 CPU(s):                  32-63,96-127
Vulnerability Gather data sampling: Not affected
Vulnerability Itlb multihit:        Not affected
Vulnerability L1tf:                 Not affected
Vulnerability Mds:                  Not affected
Vulnerability Meltdown:             Not affected
Vulnerability Mmio stale data:      Not affected
Vulnerability Retbleed:             Not affected
Vulnerability Spec rstack overflow: Mitigation; Safe RET
Vulnerability Spec store bypass:    Mitigation; Speculative Store Bypass disabled via prctl
Vulnerability Spectre v1:           Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:           Mitigation; Retpolines, IBPB conditional, IBRS_FW, STIBP always-on, RSB filling, PBRSB-eIBRS Not affected
Vulnerability Srbds:                Not affected
Vulnerability Tsx async abort:      Not affected

Versions of relevant libraries:
[pip3] lion-pytorch==0.2.2
[pip3] numpy==1.24.4
[pip3] onnx==1.16.0
[pip3] optree==0.12.1
[pip3] pytorch-ranger==0.1.1
[pip3] pytorch-triton==3.0.0+989adb9a2
[pip3] torch==2.3.1
[pip3] torch-optimizer==0.3.0
[pip3] torch-tensorrt==2.5.0a0
[pip3] torchmetrics==1.4.0.post0
[pip3] torchvision==0.18.1
[pip3] triton==2.3.1
[conda] Could not collect


Composer information
--------------------
Composer Version: 0.24.1
Composer Commit Hash: None
CPU Model: AMD EPYC 7513 32-Core Processor
CPU Count: 64
Number of Nodes: 1
GPU Model: NVIDIA A100 80GB PCIe
GPUs per Node: 1
GPU Count: 1
CUDA Device Count: 8

** To reproduce

Steps to reproduce the behavior:

Since I am working out of an internal repo, it's not straightforward for me to provide this - please do lmk if you think it will be helpful.

Expected behavior

https://github.com/mosaicml/composer/blob/main/composer/models/huggingface.py#L129

This line mentions that self.labels should be set in eval_forward(), if it exists. However, we see that self.labels is set in the if and elif branches of the eval_forward() method, but not in the else branch shown here:

https://github.com/mosaicml/composer/blob/main/composer/models/huggingface.py#L567

This causes down stream metrics to throw errors, since self.labels remains as None.

Additional context

Please let me know if I should open a PR to fix, or if there is something that I am not understanding correctly here - thank you all!

[dakinggg]

hmm looking at the history of the code, I think the expectation was that if use_logits=False, self.labels would not be set, because the metric is going to be taken directly from the output class of the HF model (e.g. cross entropy loss, rather than recomputing externally to the hf model). That being said, I don't think we have any code that does this anymore, and happy to accept a PR setting self.labels in the else branch.