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 15, 12:15 UTC
Remove Falcon3 2x7B from CI (#17404)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Harry Mellor committed
Apr 29, 2025 at 20:52 UTC
0350809f3a84d9bbc3586d3dcca98ba1400660c5
2 files changed
+4
-5
tests/models/decoder_only/language/test_models.py
+3
-4
@@ -29,7 +29,7 @@ AITER_MODEL_LIST = [
29
"openbmb/MiniCPM3-4B",
30
"Qwen/Qwen-7B-Chat",
31
"Qwen/Qwen2.5-0.5B-Instruct",
32
- "ehristoforu/Falcon3-MoE-2x7B-Insruct",
32
+ "TitanML/tiny-mixtral",
33
]
34
35
@@ -83,9 +83,8 @@ AITER_MODEL_LIST = [
83
pytest.param("stabilityai/stablelm-3b-4e1t"), # stablelm
84
pytest.param("bigcode/starcoder2-3b"), # starcoder2
85
pytest.param(
86
- "ehristoforu/Falcon3-MoE-2x7B-Insruct", # mixtral
87
- marks=[pytest.mark.cpu_model,
88
- large_gpu_mark(min_gb=48)],
86
+ "TitanML/tiny-mixtral", # mixtral
87
+ marks=[pytest.mark.cpu_model],
88
)
89
])
90
@pytest.mark.parametrize("max_tokens", [32])
tests/models/registry.py
+1
-1
@@ -191,7 +191,7 @@ _TEXT_GENERATION_EXAMPLE_MODELS = {
191
trust_remote_code=True),
192
"MistralForCausalLM": _HfExamplesInfo("mistralai/Mistral-7B-Instruct-v0.1"),
193
"MixtralForCausalLM": _HfExamplesInfo("mistralai/Mixtral-8x7B-Instruct-v0.1", # noqa: E501
194
- {"falcon3": "ehristoforu/Falcon3-MoE-2x7B-Insruct"}), # noqa: E501
194
+ {"tiny": "TitanML/tiny-mixtral"}), # noqa: E501
195
"QuantMixtralForCausalLM": _HfExamplesInfo("mistral-community/Mixtral-8x22B-v0.1-AWQ"), # noqa: E501
196
"MptForCausalLM": _HfExamplesInfo("mpt", is_available_online=False),
197
"MPTForCausalLM": _HfExamplesInfo("mosaicml/mpt-7b"),