From 9afcf0ea7d2877ddbbafec5b1a77f5cf025dab17 Mon Sep 17 00:00:00 2001
From: 游雁 <zhifu.gzf@alibaba-inc.com>
Date: 星期三, 12 六月 2024 17:17:03 +0800
Subject: [PATCH] decoding

---
 funasr/models/llm_asr/model.py |  109 ++++++++++++++++++++++++++++++++++++++++++++----------
 1 files changed, 88 insertions(+), 21 deletions(-)

diff --git a/funasr/models/llm_asr/model.py b/funasr/models/llm_asr/model.py
index 0955e84..2a55cd6 100644
--- a/funasr/models/llm_asr/model.py
+++ b/funasr/models/llm_asr/model.py
@@ -19,6 +19,7 @@
 from funasr.utils.datadir_writer import DatadirWriter
 from funasr.register import tables
 from funasr.train_utils.device_funcs import to_device
+import traceback
 
 
 @tables.register("model_classes", "LLMASR")
@@ -406,9 +407,21 @@
             audio_encoder = encoder_class(input_size=input_size, **audio_encoder_conf)
             audio_encoder_output_size = audio_encoder.output_size()
         freeze = audio_encoder_conf.get("freeze", True)
+        freeze_layer_num = int(audio_encoder_conf.get("freeze_layer_num", -1))
+        if freeze_layer_num > 0:
+            freeze_layer_num = range(freeze_layer_num)
+
         if freeze:
             for name, param in audio_encoder.named_parameters():
-                param.requires_grad = False
+                idx = re.search(r"\.\d+\.", name)
+                if idx is not None:
+                    beg, end = idx.regs[0]
+                    layer_id = int(name[beg + 1 : end - 1])
+                    if isinstance(freeze_layer_num, (list, tuple)):
+                        if layer_id in freeze_layer_num:
+                            param.requires_grad = False
+                    else:
+                        param.requires_grad = False
             audio_encoder.eval()
 
         self.audio_encoder = audio_encoder
@@ -489,6 +502,7 @@
             fbank_fake_len = fbank_fake_lens[batch_idx].item()
             fbank_beg_idx = fbank_beg[batch_idx, 0].item()
             min_len = min(fbank_fake_len, inputs_embeds.shape[1] - fbank_beg_idx)
+
             try:
                 inputs_embeds[batch_idx, fbank_beg_idx : fbank_beg_idx + min_len, :] = encoder_out[
                     batch_idx, :min_len, :
@@ -496,10 +510,10 @@
             except Exception as e:
                 logging.error(f"{str(e)}, {traceback.format_exc()}")
                 logging.info(
-                    f"batch_idx: {batch_idx}, inputs_embeds: {inputs_embeds.shape}, fbank_beg_idx: {fbank_beg_idx}, min_len: {min_len}, fbank_fake_len: {fbank_fake_len}"
+                    f"batch_idx: {batch_idx}, inputs_embeds: {inputs_embeds.shape}, fbank_beg_idx: {fbank_beg_idx}, min_len: {min_len}, fbank_fake_len: {fbank_fake_len}, encoder_out: {encoder_out.shape}, encoder_out_lens: {encoder_out_lens[batch_idx].item()}"
                 )
                 fbank_fake_len = encoder_out_lens[batch_idx].item()
-                min_len = min(fbank_fake_len, inputs_embeds.shape[1] - fbank_beg_idx)
+                min_len = min(fbank_fake_len, min_len)
                 inputs_embeds[batch_idx, fbank_beg_idx : fbank_beg_idx + min_len, :] = encoder_out[
                     batch_idx, :min_len, :
                 ]
@@ -554,13 +568,14 @@
 
         return contents
 
-    def data_load_speech(self, contents: dict, tokenizer, frontend, **kwargs):
+    def data_load_speech(self, contents: dict, tokenizer, frontend, meta_data={}, **kwargs):
 
         system = contents["system"]
         user = contents["user"]
         assistant = contents["assistant"]
         pattern = re.compile(r"(<\|startofspeech\|>.*?<\|endofspeech\|>)")
         input_ids, labels, source_ids, target_ids, fbank, fbank_lens, fbank_mask, fbank_beg = (
+            [],
             [],
             [],
             [],
@@ -591,7 +606,10 @@
                     )
                     if sub_str.startswith("!"):
                         try:
+                            time1 = time.perf_counter()
                             data_src = load_audio_text_image_video(sub_str[1:], fs=frontend.fs)
+                            time2 = time.perf_counter()
+                            meta_data["load_data"] = f"{time2 - time1:0.3f}"
                         except Exception as e:
                             logging.error(f"Loading wav failed! {str(e)}, {traceback.format_exc()}")
 
@@ -601,6 +619,15 @@
                             frontend=frontend,
                             is_final=True,
                         )  # speech: [b, T, d]
+
+                        time3 = time.perf_counter()
+                        meta_data["extract_feat"] = f"{time3 - time2:0.3f}"
+                        meta_data["batch_data_time"] = (
+                            speech_lengths.sum().item()
+                            * frontend.frame_shift
+                            * frontend.lfr_n
+                            / 1000
+                        )
 
                         if kwargs.get("permute", True):
                             speech = speech.permute(0, 2, 1)
@@ -624,7 +651,7 @@
         input_ids = torch.tensor(input_ids, dtype=torch.int64)  # [: self.max_token_length]
         attention_mask = torch.tensor([1] * len(input_ids), dtype=torch.int32)
         labels = torch.tensor(labels, dtype=torch.int64)  # [: self.max_token_length]
-        source_ids = torch.tensor(source_ids, dtype=torch.int64)
+        source_ids = torch.tensor(source_ids_i, dtype=torch.int64)
         target_ids = torch.tensor(target_ids, dtype=torch.int64)
 
         fbank = speech[0, :, :]
@@ -662,13 +689,18 @@
         if kwargs.get("batch_size", 1) > 1:
             raise NotImplementedError("batch decoding is not implemented")
 
-        contents = self.data_template(data_in)
-        output = self.data_load_speech(contents, tokenizer, frontend, **kwargs)
+        contents = self.data_template(data_in[0])
+        output = self.data_load_speech(contents, tokenizer, frontend, meta_data=meta_data, **kwargs)
         batch = to_device(output, kwargs["device"])
 
         # audio encoder
         speech = batch["speech"]
         speech_lengths = batch["speech_lengths"][:, 0]
+        # fp16
+        if kwargs.get("fp16", False):
+            speech = speech.to(torch.float16)
+        elif kwargs.get("bf16", False):
+            speech = speech.to(torch.bfloat16)
         encoder_out, encoder_out_lens = self.audio_encoder(speech.permute(0, 2, 1), speech_lengths)
 
         # audio_adaptor
@@ -676,7 +708,7 @@
 
         input_ids = batch["input_ids"]
         source_ids = batch["source_ids"]
-        if kwargs.get("tearchforing", False):
+        if not kwargs.get("tearchforing", False):
             input_ids = source_ids
         input_ids[input_ids < 0] = 0
         inputs_embeds = self.llm.model.get_input_embeddings()(input_ids)
@@ -691,19 +723,50 @@
                 batch_idx, :min_len, :
             ]
 
-        if not kwargs.get("tearchforing", False):
+        llm_dtype = kwargs.get("llm_dtype", "fp32")
+        if llm_dtype == "fp32":
+            llm_dtype = "fp16" if kwargs.get("fp16", False) else llm_dtype
+            llm_dtype = "bf16" if kwargs.get("bf16", False) else llm_dtype
 
-            generated_ids = self.llm.generate(
-                inputs_embeds=inputs_embeds, max_new_tokens=kwargs.get("max_length", 512)
-            )
-            generated_ids = [
-                output_ids[len(input_id) :]
-                for input_id, output_ids in zip(input_ids, generated_ids)
-            ]
-            response = tokenizer.batch_decode(
-                generated_ids, skip_special_tokens=kwargs.get("skip_special_tokens", True)
-            )[0]
+        dtype_map = {"bf16": torch.bfloat16, "fp16": torch.float16, "fp32": torch.float32}
+        with torch.cuda.amp.autocast(
+            enabled=True if llm_dtype != "fp32" else False, dtype=dtype_map[llm_dtype]
+        ):
             label = contents["assistant"][0]
+            self.llm = self.llm.to(dtype_map[llm_dtype])
+            inputs_embeds = inputs_embeds.to(dtype_map[llm_dtype])
+
+            if not kwargs.get("tearchforing", False):
+
+                generated_ids = self.llm.generate(
+                    inputs_embeds=inputs_embeds, max_new_tokens=kwargs.get("max_length", 512)
+                )
+                # generated_ids = [
+                #     output_ids[len(input_id) :]
+                #     for input_id, output_ids in zip(input_ids, generated_ids)
+                # ]
+                response = tokenizer.batch_decode(
+                    generated_ids, skip_special_tokens=kwargs.get("skip_special_tokens", True)
+                )[0]
+
+                loss = None
+            else:
+
+                labels_ids = batch["labels_ids"]
+                labels_ids[labels_ids == -1] = -100
+                attention_mask = batch.get("attention_mask", None)
+                # attention_mask = attention_mask.to(dtype_map[llm_dtype])
+                model_outputs = self.llm(
+                    inputs_embeds=inputs_embeds, attention_mask=attention_mask, labels=labels_ids
+                )
+
+                preds = torch.argmax(model_outputs.logits, -1)[:, source_ids.shape[1] :]
+                response = tokenizer.batch_decode(
+                    preds,
+                    add_special_tokens=False,
+                    skip_special_tokens=kwargs.get("skip_special_tokens", True),
+                )[0]
+                loss = model_outputs.loss.item()
 
         ibest_writer = None
         if kwargs.get("output_dir") is not None:
@@ -712,11 +775,15 @@
             ibest_writer = self.writer[f"{0 + 1}best_recog"]
 
         results = []
-        result_i = {"key": key[0], "text": response, "label": label}
+        response_clean = re.sub("[^\w\s\u3000\u4e00-\u9fff]+", "", response)
+        result_i = {"key": key[0], "text": response, "text_tn": response_clean, "label": label}
+        if loss is not None:
+            result_i["loss"] = loss
         results.append(result_i)
 
         if ibest_writer is not None:
-            ibest_writer["text"][key[0]] = text
+            ibest_writer["text"][key[0]] = response
             ibest_writer["label"][key[0]] = label
+            ibest_writer["text_tn"][key[0]] = response_clean
 
         return results, meta_data

--
Gitblit v1.9.1