[Bug]: stream_options.include_usage being retrieved on every chunk

[erkintelnyx]

Your current environment

PyTorch version: 2.5.0.dev20240726+rocm6.1
Is debug build: False
CUDA used to build PyTorch: N/A
ROCM used to build PyTorch: 6.1.40091-a8dbc0c19

OS: Ubuntu 20.04.6 LTS (x86_64)
GCC version: (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
Clang version: 17.0.0 (https://github.com/RadeonOpenCompute/llvm-project roc-6.1.2 24193 669db884972e769450470020c06a6f132a8a065b)
CMake version: version 3.30.2
Libc version: glibc-2.31

Python version: 3.9.19 (main, May  6 2024, 19:43:03)  [GCC 11.2.0] (64-bit runtime)
Python platform: Linux-5.15.0-117-generic-x86_64-with-glibc2.31
Is CUDA available: True
CUDA runtime version: Could not collect
CUDA_MODULE_LOADING set to: LAZY
GPU models and configuration: AMD Instinct MI100 (gfx908:sramecc+:xnack-)
Nvidia driver version: Could not collect
cuDNN version: Could not collect
HIP runtime version: 6.1.40093
MIOpen runtime version: 3.1.0
Is XNNPACK available: True

CPU:
Architecture:                         x86_64
CPU op-mode(s):                       32-bit, 64-bit
Byte Order:                           Little Endian
Address sizes:                        48 bits physical, 48 bits virtual
CPU(s):                               16
On-line CPU(s) list:                  0-15
Thread(s) per core:                   1
Core(s) per socket:                   16
Socket(s):                            1
NUMA node(s):                         1
Vendor ID:                            AuthenticAMD
CPU family:                           25
Model:                                1
Model name:                           AMD EPYC 7313 16-Core Processor
Stepping:                             1
CPU MHz:                              2999.998
BogoMIPS:                             5999.99
Virtualization:                       AMD-V
Hypervisor vendor:                    KVM
Virtualization type:                  full
L1d cache:                            1 MiB
L1i cache:                            1 MiB
L2 cache:                             8 MiB
L3 cache:                             16 MiB
NUMA node0 CPU(s):                    0-15
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 Reg file data sampling: Not affected
Vulnerability Retbleed:               Not affected
Vulnerability Spec rstack overflow:   Mitigation; safe RET, no microcode
Vulnerability Spec store bypass:      Mitigation; Speculative Store Bypass disabled via prctl and seccomp
Vulnerability Spectre v1:             Mitigation; usercopy/swapgs barriers and __user pointer sanitization
Vulnerability Spectre v2:             Mitigation; Retpolines; IBPB conditional; IBRS_FW; STIBP disabled; RSB filling; PBRSB-eIBRS Not affected; BHI Not affected
Vulnerability Srbds:                  Not affected
Vulnerability Tsx async abort:        Not affected
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 rep_good nopl cpuid extd_apicid tsc_known_freq pni pclmulqdq ssse3 fma cx16 pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy svm cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw perfctr_core invpcid_single ssbd ibrs ibpb stibp vmmcall fsgsbase tsc_adjust bmi1 avx2 smep bmi2 invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves clzero xsaveerptr wbnoinvd arat npt nrip_save umip pku ospke vaes vpclmulqdq rdpid arch_capabilities

Versions of relevant libraries:
[pip3] mypy==1.7.0
[pip3] mypy-extensions==1.0.0
[pip3] numpy==1.26.4
[pip3] optree==0.9.1
[pip3] pytorch-triton-rocm==3.0.0+21eae954ef
[pip3] pyzmq==26.1.0
[pip3] torch==2.5.0.dev20240726+rocm6.1
[pip3] torchvision==0.20.0.dev20240726+rocm6.1
[pip3] transformers==4.43.2
[pip3] triton==2.1.0
[conda] No relevant packages
ROCM Version: 6.1.40093-bd86f1708
Neuron SDK Version: N/A
vLLM Version: 0.5.4
vLLM Build Flags:
CUDA Archs: Not Set; ROCm: Disabled; Neuron: Disabled
GPU Topology:
Could not collect

🐛 Describe the bug

This could be the intended behaviour as well.

When stream_options.include_usage is set to True usage stats comes with every chunk. That was not the case for version 0.5.3 or current OpenAI API.

Llama3 8b on 0.5.4 with include_usage:

ChatCompletionChunk(id='chat-2487a8a959e2480081550c5c25694def', choices=[Choice(delta=ChoiceDelta(content=None, function_call=None, refusal=None, role='assistant', tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723032007, model='meta-llama/Meta-Llama-3-8B-Instruct', object='chat.completion.chunk', service_tier=None, system_fingerprint=None, usage=CompletionUsage(completion_tokens=0, prompt_tokens=16, total_tokens=16))


ChatCompletionChunk(id='chat-2487a8a959e2480081550c5c25694def', choices=[Choice(delta=ChoiceDelta(content='!', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723032007, model='meta-llama/Meta-Llama-3-8B-Instruct', object='chat.completion.chunk', service_tier=None, system_fingerprint=None, usage=CompletionUsage(completion_tokens=1, prompt_tokens=16, total_tokens=17))


ChatCompletionChunk(id='chat-2487a8a959e2480081550c5c25694def', choices=[Choice(delta=ChoiceDelta(content=' glare', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723032007, model='meta-llama/Meta-Llama-3-8B-Instruct', object='chat.completion.chunk', service_tier=None, system_fingerprint=None, usage=CompletionUsage(completion_tokens=2, prompt_tokens=16, total_tokens=18))


ChatCompletionChunk(id='chat-2487a8a959e2480081550c5c25694def', choices=[Choice(delta=ChoiceDelta(content=' десяти', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason='length', index=0, logprobs=None, stop_reason=None)], created=1723032007, model='meta-llama/Meta-Llama-3-8B-Instruct', object='chat.completion.chunk', service_tier=None, system_fingerprint=None, usage=CompletionUsage(completion_tokens=3, prompt_tokens=16, total_tokens=19))


ChatCompletionChunk(id='chat-2487a8a959e2480081550c5c25694def', choices=[], created=1723032007, model='meta-llama/Meta-Llama-3-8B-Instruct', object='chat.completion.chunk', service_tier=None, system_fingerprint=None, usage=CompletionUsage(completion_tokens=3, prompt_tokens=16, total_tokens=19))

GPT-4 request with include_usage:

ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[Choice(delta=ChoiceDelta(content='', function_call=None, refusal=None, role='assistant', tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=None)


ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[Choice(delta=ChoiceDelta(content='Sure', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=None)


ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[Choice(delta=ChoiceDelta(content='!', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=None)


ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[Choice(delta=ChoiceDelta(content=' Did', function_call=None, refusal=None, role=None, tool_calls=None), finish_reason=None, index=0, logprobs=None)], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=None)


ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[Choice(delta=ChoiceDelta(content=None, function_call=None, refusal=None, role=None, tool_calls=None), finish_reason='length', index=0, logprobs=None)], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=None)


ChatCompletionChunk(id='chatcmpl-9tZb6z7iaeWpNHIOk7vHP1PxCCpUU', choices=[], created=1723031904, model='gpt-4o-2024-05-13', object='chat.completion.chunk', service_tier=None, system_fingerprint='fp_3aa7262c27', usage=CompletionUsage(completion_tokens=3, prompt_tokens=13, total_tokens=16))

[robertgshaw2-redhat]

• Introduced here: [Feature]: return Usage info for streaming request for each chunk in ChatCompletion #6540

cc @tdoublep

[tomeras91]

• Introduced here: [Feature]: return Usage info for streaming request for each chunk in ChatCompletion #6540

cc @tdoublep

I think this can be fixed with changing the default of StreamOptions.continuous_usage_stats to False no?

[tdoublep]

Yes, we discussed whether it should enabled by default here: #5742 (comment)

I'm fine with disabling it by default btw

[DarkLight1337]

Let's change the default to False so it's in line with OpenAI's implementation by default.

[tpfau]

Is there any eta, when the default for this will be changed?
Or is there some easy way to change this via environment variables // runtime arguments for the OpenAI server.

[samos123]

This will be fixed in the upcoming release. There doesn't seem to be a way to tweak this through environment variables.