wjh-change

ktransformers/optimize/optimize_rules/DeepSeek-V2-Lite-Chat.yaml

@@ -0,0 +1,56 @@
+- match:
+    class: ktransformers.models.modeling_deepseek.DeepseekV2YarnRotaryEmbedding
+  replace:
+    class: ktransformers.operators.RoPE.YarnRotaryEmbedding
+    kwargs:
+      generate_device: "cuda"
+      prefill_device: "cuda"
+- match:
+    name: "^model\\.layers\\.(?!.*self_attn\\.kv_b_proj).*$"  # regular expression 
+    class: torch.nn.Linear  # only match modules matching name and class simultaneously
+  replace:
+    class: ktransformers.operators.linear.KTransformersLinear  # optimized Kernel on quantized data types
+    kwargs:
+      generate_device: "cuda"
+      prefill_device: "cuda"
+      generate_op: "KLinearMarlin"
+      prefill_op: "KLinearTorch"
+- match:
+    name: "^model\\.layers\\..*\\.mlp$"
+    class: ktransformers.models.modeling_deepseek.DeepseekV2MoE
+  replace:
+    class: ktransformers.operators.experts.KDeepseekV2MoE     # mlp module with custom forward function
+    kwargs:
+      generate_device: "cuda"
+      prefill_device: "cuda"
+- match:
+    name: "^model\\.layers\\..*\\.mlp\\.experts$"
+  replace:
+    class: ktransformers.operators.experts.KTransformersExperts     # custom MoE Kernel with expert paralleism
+    kwargs:
+      prefill_device: "cuda"
+      prefill_op: "KExpertsTorch"
+      generate_device: "cpu"
+      generate_op: "KExpertsCPU"
+      out_device: "cuda"
+  recursive: False # don't recursively inject submodules of this module
+- match:
+    name: "^model\\.layers\\..*\\.self_attn$"
+  replace:
+    class: ktransformers.operators.attention.KDeepseekV2Attention # optimized MLA implementation
+    kwargs:
+      generate_device: "cuda"
+      prefill_device: "cuda"
+- match:
+    name: "^model$"
+  replace:
+    class: "ktransformers.operators.models.KDeepseekV2Model"
+    kwargs:
+      per_layer_prefill_intput_threshold: 0 # 0 is close layer wise prefill
+- match:
+    name: "^model.embed_tokens"
+  replace:
+    class: "default"
+    kwargs:
+      generate_device: "cpu"
+      prefill_device: "cpu"

ktransformers/server/args.py

@@ -9,19 +9,19 @@ class ArgumentParser:
     def parse_args(self):
         parser = argparse.ArgumentParser(prog="kvcache.ai", description="Ktransformers")
         parser.add_argument("--host", type=str, default=self.cfg.server_ip)
-        parser.add_argument("--port", type=int, default=self.cfg.server_port)
+        parser.add_argument("--port", type=int, default=8082)
         parser.add_argument("--ssl_keyfile", type=str)
         parser.add_argument("--ssl_certfile", type=str)
-        parser.add_argument("--web", type=bool, default=self.cfg.mount_web)
-        parser.add_argument("--model_name", type=str, default=self.cfg.model_name)
-        parser.add_argument("--model_dir", type=str, default=self.cfg.model_dir)
+        parser.add_argument("--web", type=bool, default=True)
+        parser.add_argument("--model_name", type=str, default='DeepSeek-V2-Lite-Chat')
+        parser.add_argument("--model_dir", type=str, default='/mnt/data/model/DeepSeek-V2-Lite-Chat')
         parser.add_argument(
             "--device", type=str, default=self.cfg.model_device, help="Warning: Abandoning this parameter"
         )
-        parser.add_argument("--gguf_path", type=str, default=self.cfg.gguf_path)
-        parser.add_argument("--optimize_config_path", default=self.cfg.optimize_config_path, type=str, required=False)
+        parser.add_argument("--gguf_path", type=str, default='/mnt/data/model/DeepSeek-V2-Lite-Chat-GGUF')
+        parser.add_argument("--optimize_config_path", default='/mnt/data/benchmark/ktransformers-dev/ktransformers/optimize/optimize_rules/DeepSeek-V2-Lite-Chat.yaml', type=str, required=False)
         parser.add_argument("--cpu_infer", type=int, default=self.cfg.cpu_infer)
-        parser.add_argument("--type", type=str, default=self.cfg.backend_type)
+        parser.add_argument("--type", type=str, default='ktransformers')
 
         # model configs
         # parser.add_argument("--model_cache_lens", type=int, default=self.cfg.cache_lens)  # int?
@@ -69,7 +69,7 @@ class ArgumentParser:
         parser.add_argument("--print_timings", type=bool, default=self.cfg.print_timings)
         parser.add_argument("--amnesia", type=bool, default=self.cfg.amnesia)
         parser.add_argument("--batch_size", type=int, default=self.cfg.batch_size)
-        parser.add_argument("--cache_lens", type=int, default=self.cfg.cache_lens)
+        parser.add_argument("--cache_lens", type=int, default='32768')
 
         # log configs
         # log level: debug, info, warn, error, crit

ktransformers/server/backend/interfaces/transformers.py

@@ -164,6 +164,7 @@ class TransformersInterface(BackendInterfaceBase):
             if m["role"] == "system":
                 logger.warning(f'change {m["role"]} to user')
                 m["role"] = "user"
+
         new_messages = [messages[0]]
         for m in messages[1:]:
             if m["role"] == "user" and new_messages[-1]["role"] == "user":
@@ -172,25 +173,12 @@ class TransformersInterface(BackendInterfaceBase):
             else:
                 new_messages.append(m)
 
-        # if (self.last_request_id is not None) and self.last_request_id == thread_id:
-        #     logger.debug(f"last message: {new_messages[-1]}")
-        #     input_ids = self.tokenizer.apply_chat_template([new_messages[-1]], return_tensors="pt",add_generation_prompt=False).to(self.args.device)
-        # else:
-        #     input_ids = self.tokenizer.apply_chat_template(
-        #         new_messages, return_tensors="pt", add_generation_prompt=True
-        #     ).to(self.args.device)
-
-        input_ids = self.tokenizer.apply_chat_template(new_messages,return_tensors='pt',add_generation_prompt=True).to(self.args.device)
         if (self.last_request_id is not None) and self.last_request_id == thread_id:
-            x = self.generated_ids[:,:self.seq_length]
-            y = input_ids[:,:self.seq_length]
-            # We can only hope that the input_ids are the same
-            unequal_mask = torch.ne(x,y)
-            unequal_positions = torch.nonzero(unequal_mask)
-            num_unequal_elements = unequal_mask.sum().item()
-            logger.warning(f'num_unequal_elements: {num_unequal_elements}') 
-
-            input_ids = input_ids[:,self.seq_length:]
+            input_ids = self.tokenizer.encode(self.tokenizer.eos_token+self.tokenizer.apply_chat_template([new_messages[-1]], return_tensors="pt",tokenize=False, add_generation_prompt=True), add_special_tokens = False, return_tensors="pt").to(self.args.device)
+        else:
+            input_ids = self.tokenizer.apply_chat_template(
+                new_messages, return_tensors="pt", add_generation_prompt=True
+            ).to(self.args.device)
         logger.debug(f"get input ids of shape {input_ids.shape}")
         return input_ids
 

ktransformers/server/main.py

@@ -4,6 +4,10 @@ from fastapi import FastAPI
 from fastapi.staticfiles import StaticFiles
 import uvicorn.logging
 import uvicorn
+import sys
+
+project_dir = os.path.dirname(os.path.dirname(os.path.dirname(__file__)))
+sys.path.insert(0, project_dir)
 from fastapi.middleware.cors import CORSMiddleware
 from ktransformers.server.args import ArgumentParser
 from ktransformers.server.config.config import Config