A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 3, 15:15 UTC
[Deprecation] Remove old locations of `get_tokenizer` and `resolve_hf_chat_template` (#35024)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Cyrus Leung committed
May 15, 2026 at 15:13 UTC
2676ab1e0b93c3d75829119fdc5bbdbc1b8193f0
17 files changed
+15
-64
.buildkite/lm-eval-harness/run-lm-eval-chartqa-vllm-vlm-baseline.sh
+1
-1
@@ -2,7 +2,7 @@
2
# We can use this script to compute baseline accuracy on chartqa for vllm.
3
#
4
# Make sure you have lm-eval-harness installed:
5
-# pip install "lm-eval[api]>=0.4.11"
5
+# pip install "lm-eval[api]>=0.4.12"
6
7
usage() {
8
echo``
.buildkite/lm-eval-harness/run-lm-eval-gsm-hf-baseline.sh
+1
-1
@@ -2,7 +2,7 @@
2
# We can use this script to compute baseline accuracy on GSM for transformers.
3
#
4
# Make sure you have lm-eval-harness installed:
5
-# pip install "lm-eval[api]>=0.4.11"
5
+# pip install "lm-eval[api]>=0.4.12"
6
7
usage() {
8
echo``
.buildkite/lm-eval-harness/run-lm-eval-gsm-vllm-baseline.sh
+1
-1
@@ -3,7 +3,7 @@
3
# We use this for fp8, which HF does not support.
4
#
5
# Make sure you have lm-eval-harness installed:
6
-# pip install "lm-eval[api]>=0.4.11"
6
+# pip install "lm-eval[api]>=0.4.12"
7
8
usage() {
9
echo``
.buildkite/lm-eval-harness/run-lm-eval-mmlupro-vllm-baseline.sh
+1
-1
@@ -3,7 +3,7 @@
3
# We use this for fp8, which HF does not support.
4
#
5
# Make sure you have lm-eval-harness installed:
6
-# pip install "lm-eval[api]>=0.4.11"
6
+# pip install "lm-eval[api]>=0.4.12"
7
8
usage() {
9
echo``
.buildkite/scripts/hardware_ci/run-tpu-v1-test-part2.sh
+1
-1
@@ -61,7 +61,7 @@ echo "Results will be stored in: $RESULTS_DIR"
61
echo "--- Installing Python dependencies ---"
62
python3 -m pip install --progress-bar off git+https://github.com/thuml/depyf.git \
63
&& python3 -m pip install --progress-bar off pytest pytest-asyncio tpu-info \
64
- && python3 -m pip install --progress-bar off "lm-eval[api]>=0.4.11" \
64
+ && python3 -m pip install --progress-bar off "lm-eval[api]>=0.4.12" \
65
&& python3 -m pip install --progress-bar off hf-transfer tblib==3.1.0
66
echo "--- Python dependencies installed ---"
67
.buildkite/scripts/hardware_ci/run-tpu-v1-test.sh
+1
-1
@@ -61,7 +61,7 @@ echo "Results will be stored in: $RESULTS_DIR"
61
echo "--- Installing Python dependencies ---"
62
python3 -m pip install --progress-bar off git+https://github.com/thuml/depyf.git \
63
&& python3 -m pip install --progress-bar off pytest pytest-asyncio tpu-info \
64
- && python3 -m pip install --progress-bar off "lm-eval[api]>=0.4.11" \
64
+ && python3 -m pip install --progress-bar off "lm-eval[api]>=0.4.12" \
65
&& python3 -m pip install --progress-bar off hf-transfer tblib==3.1.0
66
echo "--- Python dependencies installed ---"
67
docs/features/quantization/fp8.md
+1
-1
@@ -84,7 +84,7 @@ Since simple RTN does not require data for weight quantization and the activatio
84
Install `vllm` and `lm-evaluation-harness` for evaluation:
85
86
```bash
87
-pip install vllm "lm-eval[api]>=0.4.11"
87
+pip install vllm "lm-eval[api]>=0.4.12"
88
```
89
90
Load and run the model in `vllm`:
docs/features/quantization/int4.md
+1
-1
@@ -18,7 +18,7 @@ pip install llmcompressor
18
Additionally, install `vllm` and `lm-evaluation-harness` for evaluation:
19
20
```bash
21
-pip install vllm "lm-eval[api]>=0.4.11"
21
+pip install vllm "lm-eval[api]>=0.4.12"
22
```
23
24
## Quantization Process
docs/features/quantization/int8.md
+1
-1
@@ -23,7 +23,7 @@ pip install llmcompressor
23
Additionally, install `vllm` and `lm-evaluation-harness` for evaluation:
24
25
```bash
26
-pip install vllm "lm-eval[api]>=0.4.11"
26
+pip install vllm "lm-eval[api]>=0.4.12"
27
```
28
29
## Quantization Process
docs/features/quantization/quark.md
+1
-1
@@ -20,7 +20,7 @@ for more installation details.
20
Additionally, install `vllm` and `lm-evaluation-harness` for evaluation:
21
22
```bash
23
-pip install vllm "lm-eval[api]>=0.4.11"
23
+pip install vllm "lm-eval[api]>=0.4.12"
24
```
25
26
## Quantization Process
requirements/test/cuda.in
+1
-1
@@ -36,7 +36,7 @@ num2words # required for smolvlm test
36
open_clip_torch==2.32.0 # Required for nemotron_vl test, Nemotron Parse in test_common.py
37
opencv-python-headless >= 4.13.0 # required for video test
38
datamodel_code_generator # required for minicpm3 test
39
-lm-eval[api]>=0.4.11 # required for model evaluation test
39
+lm-eval[api]>=0.4.12 # required for model evaluation test
40
mteb[bm25s]>=2, <3 # required for mteb test
41
transformers==5.5.3
42
tokenizers==0.22.2
requirements/test/cuda.txt
+1
-6
@@ -41,7 +41,6 @@ attrs==24.2.0
41
# via
42
# aiohttp
43
# hypothesis
44
- # jsonlines
44
# jsonschema
45
# pytest-subtests
46
# referencing
@@ -357,8 +356,6 @@ joblib==1.4.2
356
# librosa
357
# nltk
358
# scikit-learn
360
-jsonlines==4.0.0
361
- # via lm-eval
359
jsonpointer==3.0.0
360
# via jsonschema
361
jsonschema==4.23.0
@@ -387,7 +384,7 @@ librosa==0.10.2.post1
384
# via -r requirements/test/cuda.in
385
llvmlite==0.47.0
386
# via numba
390
-lm-eval==0.4.11
387
+lm-eval==0.4.12
388
# via -r requirements/test/cuda.in
389
lxml==5.3.0
390
# via
@@ -1141,5 +1138,3 @@ yarl==1.17.1
1138
# schemathesis
1139
zipp==3.23.0
1140
# via importlib-metadata
1144
-zstandard==0.23.0
1145
- # via lm-eval
requirements/test/nightly-torch.txt
+1
-1
@@ -27,7 +27,7 @@ mistral_common[image,audio] >= 1.11.2 # required for voxtral test
27
num2words # required for smolvlm test
28
opencv-python-headless >= 4.13.0 # required for video test
29
datamodel_code_generator # required for minicpm3 test
30
-lm-eval[api]>=0.4.11 # required for model evaluation test
30
+lm-eval[api]>=0.4.12 # required for model evaluation test
31
mteb[bm25s]>=2, <3 # required for mteb test
32
transformers==5.5.3
33
tokenizers==0.22.2
requirements/test/rocm.in
+1
-1
@@ -35,7 +35,7 @@ num2words # required for smolvlm test
35
open_clip_torch==2.32.0 # Required for nemotron_vl test, Nemotron Parse in test_common.py
36
opencv-python-headless>=4.13.0 # required for video test
37
datamodel_code_generator # required for minicpm3 test
38
-lm-eval[api]>=0.4.11 # required for model evaluation test
38
+lm-eval[api]>=0.4.12 # required for model evaluation test
39
mteb[bm25s]>=2, <3 # required for mteb test
40
transformers==5.5.3
41
tokenizers==0.22.2
requirements/test/rocm.txt
+1
-6
@@ -55,7 +55,6 @@ astor==0.8.1
55
attrs==26.1.0
56
# via
57
# aiohttp
58
- # jsonlines
58
# jsonschema
59
# pytest-subtests
60
# referencing
@@ -433,8 +432,6 @@ joblib==1.5.3
432
# librosa
433
# nltk
434
# scikit-learn
436
-jsonlines==4.0.0
437
- # via lm-eval
435
jsonpointer==3.1.0
436
# via jsonschema
437
jsonschema==4.26.0
@@ -472,7 +469,7 @@ llguidance==1.7.5
469
# -r requirements/test/../common.txt
470
llvmlite==0.47.0
471
# via numba
475
-lm-eval==0.4.11
472
+lm-eval==0.4.12
473
# via -r requirements/test/rocm.in
474
lm-format-enforcer==0.11.3
475
# via
@@ -1362,8 +1359,6 @@ yarl==1.23.0
1359
# schemathesis
1360
zipp==3.23.0
1361
# via importlib-metadata
1365
-zstandard==0.25.0
1366
- # via lm-eval
1362
1363
# The following packages were excluded from the output:
1364
# torch
vllm/entrypoints/chat_utils.py
-18
@@ -3,7 +3,6 @@
3
4
import asyncio
5
import json
6
-import warnings
6
from abc import ABC, abstractmethod
7
from collections import Counter, defaultdict
8
from collections.abc import Awaitable, Callable, Iterable
@@ -73,23 +72,6 @@ else:
72
logger = init_logger(__name__)
73
74
76
-def __getattr__(name: str):
77
- if name == "resolve_hf_chat_template":
78
- from vllm.renderers.hf import resolve_chat_template
79
-
80
- warnings.warn(
81
- "`vllm.entrypoints.chat_utils.resolve_hf_chat_template` has been moved to "
82
- "`vllm.renderers.hf.resolve_chat_template`. "
83
- "The old name will be removed in v0.16.",
84
- DeprecationWarning,
85
- stacklevel=2,
86
- )
87
-
88
- return resolve_chat_template
89
-
90
- raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
91
-
92
-
75
class ChatTemplateResolutionError(ValueError):
76
"""Raised when chat template resolution fails.
77
vllm/transformers_utils/tokenizer.py
deleted
-21
@@ -1,21 +0,0 @@
1
-# SPDX-License-Identifier: Apache-2.0
2
-# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
3
-import warnings
4
-
5
-
6
-def __getattr__(name: str):
7
- # Keep until lm-eval is updated
8
- if name == "get_tokenizer":
9
- from vllm.tokenizers import get_tokenizer
10
-
11
- warnings.warn(
12
- "`vllm.transformers_utils.tokenizer.get_tokenizer` "
13
- "has been moved to `vllm.tokenizers.get_tokenizer`. "
14
- "The old name will be removed in a future version.",
15
- DeprecationWarning,
16
- stacklevel=2,
17
- )
18
-
19
- return get_tokenizer
20
-
21
- raise AttributeError(f"module {__name__!r} has no attribute {name!r}")