游雁
2024-06-08 3d5e19792cd4bb510c2c0fc5749731d52b825c15
fix bug
2个文件已修改
32 ■■■■ 已修改文件
examples/industrial_data_pretraining/llm_asr/demo_speech2text.py 4 ●●● 补丁 | 查看 | 原始文档 | blame | 历史
funasr/models/llm_asr/model.py 28 ●●●● 补丁 | 查看 | 原始文档 | blame | 历史
examples/industrial_data_pretraining/llm_asr/demo_speech2text.py
@@ -16,12 +16,14 @@
with open(jsonl, "r") as f:
    lines = f.readlines()
tearchforing = True
for i, line in enumerate(lines):
    data_dict = json.loads(line.strip())
    data = data_dict["messages"]
    res = model.generate(
        input=data,
        input=[data],
        tearchforing=tearchforing,
        cache={},
    )
funasr/models/llm_asr/model.py
@@ -568,6 +568,7 @@
            [],
            [],
            [],
            [],
        )
        for i, (system_prompt, user_prompt, target_out) in enumerate(zip(system, user, assistant)):
@@ -624,7 +625,7 @@
        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, :, :]
@@ -662,7 +663,7 @@
        if kwargs.get("batch_size", 1) > 1:
            raise NotImplementedError("batch decoding is not implemented")
        contents = self.data_template(data_in)
        contents = self.data_template(data_in[0])
        output = self.data_load_speech(contents, tokenizer, frontend, **kwargs)
        batch = to_device(output, kwargs["device"])
@@ -676,7 +677,7 @@
        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)
@@ -704,6 +705,23 @@
                generated_ids, skip_special_tokens=kwargs.get("skip_special_tokens", True)
            )[0]
            label = contents["assistant"][0]
            loss = None
        else:
            labels_ids = batch["labels_ids"]
            labels_ids[labels_ids == -1] = -100
            attention_mask = batch.get("attention_mask", None)
            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
        ibest_writer = None
        if kwargs.get("output_dir") is not None:
@@ -713,10 +731,12 @@
        results = []
        result_i = {"key": key[0], "text": response, "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
        return results, meta_data