Fix quantization and nightly tests (#4258)

2025-03-10 03:06:21 -07:00
parent 1a5023e05d
commit 00d25a7f5e
7 changed files with 142 additions and 70 deletions
--- a/test/srt/test_nightly_gsm8k_eval.py
+++ b/test/srt/test_nightly_gsm8k_eval.py
@@ -38,6 +38,7 @@ MODEL_SCORE_THRESHOLDS = {
    "neuralmagic/Qwen2-57B-A14B-Instruct-FP8": 0.82,
    "hugging-quants/Meta-Llama-3.1-8B-Instruct-AWQ-INT4": 0.84,
    "hugging-quants/Meta-Llama-3.1-8B-Instruct-GPTQ-INT4": 0.83,
+    "hugging-quants/Mixtral-8x7B-Instruct-v0.1-AWQ-INT4": 0.60,
 }