FIX: Fix multiple generations for new HF cache format (#444)

awq/modules/fused/attn.py

@@ -189,16 +189,19 @@ def forward(
             self.start_pos = 0
 
         hf_is_generating = False
+        hf_is_first_forward = "past_key_value" in kwargs and kwargs["past_key_value"] is None
+        hf_is_new_cache_first_forward = "past_key_value" in kwargs and isinstance(kwargs["past_key_value"], DynamicCache) and kwargs["past_key_value"].get_seq_length() == 0
 
         if self.is_hf_transformers and "use_cache" in kwargs:
             hf_is_generating = kwargs["use_cache"]
 
+        # print(kwargs["past_key_value"].get_seq_length())
 
         # In case we re-generate, we need to refresh the starting position
         # to 0. We detect it by checking if `past_key_values` is set to None,
         # which indicates that we are on the first step of `generate()`.
         # This is only applicable for `transformers` integration
-        if (self.is_hf_transformers and "past_key_value" in kwargs and kwargs["past_key_value"] is None) or (self.is_hf_transformers and not hf_is_generating):
+        if (self.is_hf_transformers and (hf_is_first_forward or hf_is_new_cache_first_forward)) or (self.is_hf_transformers and not hf_is_generating):
             self.start_pos = 0