Merge pull request #1401 from alibaba-damo-academy/dev_gzf

init param

funasr/auto/auto_model.py

@@ -95,7 +95,7 @@ def prepare_data_iterator(data_in, input_len=None, data_type=None, key=None):
 class AutoModel:
     
     def __init__(self, **kwargs):
-        if not kwargs.get("disable_log", False):
+        if not kwargs.get("disable_log", True):
             tables.print()
         
         model, kwargs = self.build_model(**kwargs)
@@ -193,7 +193,7 @@ class AutoModel:
                     path=init_param,
                     ignore_init_mismatch=kwargs.get("ignore_init_mismatch", False),
                     oss_bucket=kwargs.get("oss_bucket", None),
-                    scope_map=kwargs.get("scope_map", None),
+                    scope_map=kwargs.get("scope_map", "module.,None"),
                     excludes=kwargs.get("excludes", None),
                 )
             else:

funasr/bin/train.py

@@ -105,7 +105,7 @@ def main(**kwargs):
                     path=p,
                     ignore_init_mismatch=kwargs.get("ignore_init_mismatch", True),
                     oss_bucket=kwargs.get("oss_bucket", None),
-                    scope_map=kwargs.get("scope_map", None),
+                    scope_map=kwargs.get("scope_map", "module.,none"),
                     excludes=kwargs.get("excludes", None),
                 )
             else:

funasr/models/llm_asr/model.py → funasr/models/llm_asr_nar/model.py

@@ -294,24 +294,29 @@ class LLMASRNAR(nn.Module):
         inputs_embeds = torch.cat((inputs_embeds[None, :, :], encoder_out), dim=1)  # [prompt, audio]
         attention_mask = torch.ones(inputs_embeds.size()[:-1], dtype=torch.long).to(kwargs["device"])
         
-        model_outputs = self.llm.generate(
-            inputs_embeds=inputs_embeds,
-            max_length=kwargs.get("max_length", 200),
-            max_new_tokens=kwargs.get("max_new_tokens", 200),
-            num_beams=kwargs.get("num_beams", 4),
-            do_sample=kwargs.get("do_sample", False),
-            min_length=kwargs.get("min_length", 1),
-            top_p=kwargs.get("top_p", 1.0),
-            repetition_penalty=kwargs.get("repetition_penalty", 1.0),
-            length_penalty=kwargs.get("length_penalty", 1.0),
-            temperature=kwargs.get("temperature", 1.0),
-            attention_mask=attention_mask,
-            bos_token_id=tokenizer.bos_token_id,
-            eos_token_id=tokenizer.eos_token_id,
-            pad_token_id=tokenizer.pad_token_id
-        )
+        # model_outputs = self.llm.generate(
+        #     inputs_embeds=inputs_embeds,
+        #     max_length=kwargs.get("max_length", 200),
+        #     max_new_tokens=kwargs.get("max_new_tokens", 200),
+        #     num_beams=kwargs.get("num_beams", 4),
+        #     do_sample=kwargs.get("do_sample", False),
+        #     min_length=kwargs.get("min_length", 1),
+        #     top_p=kwargs.get("top_p", 1.0),
+        #     repetition_penalty=kwargs.get("repetition_penalty", 1.0),
+        #     length_penalty=kwargs.get("length_penalty", 1.0),
+        #     temperature=kwargs.get("temperature", 1.0),
+        #     attention_mask=attention_mask,
+        #     bos_token_id=tokenizer.bos_token_id,
+        #     eos_token_id=tokenizer.eos_token_id,
+        #     pad_token_id=tokenizer.pad_token_id
+        # )
+
 
-        text = tokenizer.batch_decode(model_outputs, add_special_tokens=False, skip_special_tokens=True)
+        model_outputs = self.llm(inputs_embeds=inputs_embeds, attention_mask=attention_mask, labels=None)
+        preds = torch.argmax(model_outputs.logits, -1)
+        text = tokenizer.batch_decode(preds, add_special_tokens=False, skip_special_tokens=True)
+        text = text.split(': "\n')[-1]
+        # preds = torch.argmax(model_outputs.logits, -1)
         
         ibest_writer = None
         if kwargs.get("output_dir") is not None:

funasr/train_utils/load_pretrained_model.py

@@ -82,7 +82,7 @@ def load_pretrained_model(
 	ignore_init_mismatch: bool,
 	map_location: str = "cpu",
 	oss_bucket=None,
-	scope_map=None,
+	scope_map="module.:none",
 	excludes=None,
 ):
 	"""Load a model state and set it to the model.
@@ -108,15 +108,40 @@ def load_pretrained_model(
 	
 	src_state = src_state["model"] if "model" in src_state else src_state
 	
+	if isinstance(scope_map, str):
+		scope_map = scope_map.split(",")
+	
 	for k in dst_state.keys():
-		if not k.startswith("module.") and "module." + k in src_state.keys():
-			k_ddp = "module." + k
-		else:
-			k_ddp = k
-		if k_ddp in src_state:
-			dst_state[k] = src_state[k_ddp]
+		# if not k.startswith("module.") and "module." + k in src_state.keys():
+		# 	k_ddp = "module." + k
+		# else:
+		# 	k_ddp = k
+		k_src = k
+
+		if scope_map is not None:
+			src_prefix = ""
+			dst_prefix = ""
+			for i in range(0, len(scope_map), 2):
+				src_prefix = scope_map[i] if scope_map[i].lower() != "none" else ""
+				dst_prefix = scope_map[i+1] if scope_map[i+1].lower() != "none" else ""
+
+				if k.startswith(dst_prefix) and k.replace(dst_prefix, src_prefix) in src_state.keys():
+					k_src = k.replace(dst_prefix, src_prefix)
+					print(f"init param, map: {k} from {k_src} in ckpt")
+					
+		if k_src in src_state.keys():
+			dst_state[k] = src_state[k_src]
+				
+		# if k_ddp.startswith("audio_encoder"):
+		# 	if k_ddp.replace("audio_encoder", "encoder.model") in src_state.keys():
+		# 		k_ddp = k_ddp.replace("audio_encoder", "encoder.model")
+		# if k_ddp.startswith("adaptor"):
+		# 	if k_ddp.replace("adaptor", "encoder_projector") in src_state.keys():
+		# 		k_ddp = k_ddp.replace("adaptor", "encoder_projector")
+		# if k_ddp in src_state:
+		# 	dst_state[k] = src_state[k_ddp]
 		else:
-			print(f"Warning, miss key in ckpt: {k}, mapped: {k_ddp}")
+			print(f"Warning, miss key in ckpt: {k}, mapped: {k_src}")
 			
 	flag = obj.load_state_dict(dst_state, strict=False)
 	# print(flag)