A high-throughput and memory-efficient inference and serving engine for LLMs
[Model] Remove XverseForCausalLM (#45638)
Signed-off-by: Xianbao QIAN <xianbao.qian@gmail.com> Co-authored-by: Claude <noreply@anthropic.com>
Tiezhen WANG committed
Jun 15, 2026 at 13:09 UTC
7df4fe1bd78517d6e0f3d73487a60cb53cdb51c7
4 files changed
+1
-15
docs/models/supported_models.md
-1
@@ -488,7 +488,6 @@ th {
488
| `TeleChat2ForCausalLM` | TeleChat2 | `Tele-AI/TeleChat2-3B`, `Tele-AI/TeleChat2-7B`, `Tele-AI/TeleChat2-35B`, etc. | ✅︎ | ✅︎ |
489
| `TeleChat3ForCausalLM` | TeleChat3 | `Tele-AI/TeleChat3-36B-Thinking`, `Tele-AI/TeleChat3-Coder-36B-Thinking`, etc. | ✅︎ | ✅︎ |
490
| `TeleFLMForCausalLM` | TeleFLM | `CofeAI/FLM-2-52B-Instruct-2407`, `CofeAI/Tele-FLM`, etc. | ✅︎ | ✅︎ |
491
-| `XverseForCausalLM` | XVERSE | `xverse/XVERSE-7B-Chat`, `xverse/XVERSE-13B-Chat`, `xverse/XVERSE-65B-Chat`, etc. | ✅︎ | ✅︎ |
491
| `MiniMaxM1ForCausalLM` | MiniMax-Text | `MiniMaxAI/MiniMax-M1-40k`, `MiniMaxAI/MiniMax-M1-80k`, etc. | | |
492
| `MiniMaxText01ForCausalLM` | MiniMax-Text | `MiniMaxAI/MiniMax-Text-01`, etc. | | |
493
| `Zamba2ForCausalLM` | Zamba2 | `Zyphra/Zamba2-7B-instruct`, `Zyphra/Zamba2-2.7B-instruct`, `Zyphra/Zamba2-1.2B-instruct`, etc. | | |
tests/distributed/test_pipeline_parallel.py
-3
@@ -155,9 +155,6 @@ TEXT_GENERATION_MODELS = {
155
"stabilityai/stablelm-3b-4e1t": PPTestSettings.fast(),
156
"bigcode/starcoder2-3b": PPTestSettings.fast(),
157
"upstage/solar-pro-preview-instruct": PPTestSettings.fast(load_format="dummy"),
158
- # FIXME: Cannot load tokenizer in latest transformers version.
159
- # Need to use tokenizer from `meta-llama/Llama-2-7b-chat-hf`
160
- # "xverse/XVERSE-7B-Chat": PPTestSettings.fast(),
158
# [Encoder-only]
159
# TODO: Implement PP
160
# "facebook/bart-base": PPTestSettings.fast(),
tests/models/registry.py
-10
@@ -564,16 +564,6 @@ _TEXT_GENERATION_EXAMPLE_MODELS = {
564
"TeleFLMForCausalLM": _HfExamplesInfo(
565
"CofeAI/FLM-2-52B-Instruct-2407", trust_remote_code=True
566
),
567
- "XverseForCausalLM": _HfExamplesInfo(
568
- "xverse/XVERSE-7B-Chat",
569
- tokenizer="meta-llama/Llama-2-7b",
570
- trust_remote_code=True,
571
- max_transformers_version="4.57",
572
- transformers_version_reason={
573
- "vllm": "XVERSE tokenizer is incompatible with transformers v5 "
574
- "(add_prefix_space / prepend_scheme mismatch).",
575
- },
576
- ),
567
"Zamba2ForCausalLM": _HfExamplesInfo("Zyphra/Zamba2-7B-instruct"),
568
"MiMoForCausalLM": _HfExamplesInfo("XiaomiMiMo/MiMo-7B-RL", trust_remote_code=True),
569
"MiMoV2FlashForCausalLM": _HfExamplesInfo(
vllm/model_executor/models/registry.py
+1
-1
@@ -214,7 +214,6 @@ _TEXT_GENERATION_MODELS = {
214
"TeleChat2ForCausalLM": ("telechat2", "TeleChat2ForCausalLM"),
215
"TeleChat3ForCausalLM": ("llama", "LlamaForCausalLM"),
216
"TeleFLMForCausalLM": ("teleflm", "TeleFLMForCausalLM"),
217
- "XverseForCausalLM": ("llama", "LlamaForCausalLM"),
217
"Zamba2ForCausalLM": ("zamba2", "Zamba2ForCausalLM"),
218
}
219
@@ -723,6 +722,7 @@ _PREVIOUSLY_SUPPORTED_MODELS = {
722
# have been removed for V0 deprecation.
723
"DonutForConditionalGeneration": "0.10.2",
724
"MllamaForConditionalGeneration": "0.10.2",
725
+ "XverseForCausalLM": "0.23.0",
726
}
727
728
_OOT_SUPPORTED_MODELS = {