[CI][AMD] Move 17 test groups from MI300 to MI355 and test large DeepSeek/Nemotron evals on AITER (#61955)

Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
Andreas Karatzas
2026-10-16 10:58:59 -05:00
committed by GitHub
co-authored by Claude
parent 76081b9ee4
commit 706791d516
13 changed files with 228 additions and 350 deletions
@@ -1,12 +0,0 @@
model_name: "deepseek-ai/DeepSeek-R1"
accuracy_threshold: 0.95
num_questions: 1319
num_fewshot: 5
startup_max_wait_seconds: 1200
server_args: >-
--enforce-eager
--max-model-len 4096
--data-parallel-size 8
--enable-expert-parallel
--attention-backend=TRITON_ATTN
--speculative-config '{"method":"mtp","num_speculative_tokens":3}'
@@ -1,12 +0,0 @@
model_name: "deepseek-ai/DeepSeek-R1"
accuracy_threshold: 0.95
num_questions: 1319
num_fewshot: 5
startup_max_wait_seconds: 1200
server_args: >-
--enforce-eager
--max-model-len 4096
--tensor-parallel-size 8
--enable-expert-parallel
--attention-backend=TRITON_ATTN
--speculative-config '{"method":"mtp","num_speculative_tokens":3}'
@@ -1,12 +0,0 @@
model_name: "deepseek-ai/DeepSeek-V3.2"
accuracy_threshold: 0.95
num_questions: 1319
num_fewshot: 5
startup_max_wait_seconds: 1200
server_args: >-
--enforce-eager
--max-model-len 4096
--data-parallel-size 8
--enable-expert-parallel
--attention-backend=TRITON_ATTN
--speculative-config '{"method":"mtp","num_speculative_tokens":3}'
@@ -1,12 +0,0 @@
model_name: "deepseek-ai/DeepSeek-V3.2"
accuracy_threshold: 0.95
num_questions: 1319
num_fewshot: 5
startup_max_wait_seconds: 1200
server_args: >-
--enforce-eager
--max-model-len 4096
--tensor-parallel-size 8
--enable-expert-parallel
--attention-backend=TRITON_ATTN
--speculative-config '{"method":"mtp","num_speculative_tokens":3}'
@@ -12,4 +12,5 @@ server_args: >-
--enable-expert-parallel
--mamba-backend triton
--moe-backend triton
--attention-backend ROCM_AITER_FA
--speculative-config '{"method":"mtp","num_speculative_tokens":5}'
@@ -12,4 +12,5 @@ server_args: >-
--enable-expert-parallel
--mamba-backend triton
--moe-backend triton
--attention-backend ROCM_AITER_FA
--speculative-config '{"method":"mtp","num_speculative_tokens":5}'
@@ -1,8 +0,0 @@
DeepSeek-R1-TP_MI325.yaml
DeepSeek-R1-DP_MI325.yaml
DeepSeek-V3.2-TP_MI325.yaml
DeepSeek-V3.2-DP_MI325.yaml
Qwen3-30B-A3B-NVFP4.yaml
Qwen3-30B-A3B-NVFP4-quark.yaml
Qwen3.5-35B-A3B-MXFP4-AITER-TP2.yaml
Qwen3.5-35B-A3B-MXFP4-EMU-TP2.yaml