初始化项目，由ModelHub XC社区提供模型

Model: RedHatAI/Llama-3.2-1B-Instruct-quantized.w8a8 Source: Original Platform
2026-05-06 12:14:02 +08:00
commit fe294906a5
10 changed files with 421 additions and 0 deletions
--- a/recipe.yaml
+++ b/recipe.yaml
@@ -0,0 +1,18 @@
+quant_stage:
+  quant_modifiers:
+    SmoothQuantModifier:
+      smoothing_strength: 0.7
+      mappings:
+      - - ['re:.*q_proj', 're:.*k_proj', 're:.*v_proj']
+        - re:.*input_layernorm
+      - - ['re:.*gate_proj', 're:.*up_proj']
+        - re:.*post_attention_layernorm
+      - - ['re:.*down_proj']
+        - re:.*up_proj
+    GPTQModifier:
+      sequential_update: true
+      dampening_frac: 0.01
+      ignore: [lm_head]
+      scheme: W8A8
+      targets: [Linear]
+      observer: mse