commit from root

Baichuan-13B-Chat-lora-Retrieval/README.md

@@ -0,0 +1,9 @@
+---
+library_name: peft
+---
+## Training procedure
+
+### Framework versions
+
+
+- PEFT 0.5.0

Baichuan-13B-Chat-lora-Retrieval/adapter_config.json

@@ -0,0 +1,20 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "baichuan-inc/Baichuan-13B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 32.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "revision": null,
+  "target_modules": [
+    "W_pack"
+  ],
+  "task_type": "CAUSAL_LM"
+}

Baichuan-13B-Chat-lora-Retrieval/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6f743cb6f789074ff8abf9efd5bedc2451b00b041f8e357dea2d989b58d8dabd
+size 26243422

Baichuan-13B-Chat-lora-Retrieval/all_results.json

@@ -0,0 +1,11 @@
+{
+    "epoch": 1.99,
+    "eval_loss": 1.0294359922409058,
+    "eval_runtime": 7.9413,
+    "eval_samples_per_second": 25.436,
+    "eval_steps_per_second": 2.141,
+    "train_loss": 1.0595260426618052,
+    "train_runtime": 5379.5583,
+    "train_samples_per_second": 7.407,
+    "train_steps_per_second": 0.077
+}

Baichuan-13B-Chat-lora-Retrieval/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.99,
+    "eval_loss": 1.0294359922409058,
+    "eval_runtime": 7.9413,
+    "eval_samples_per_second": 25.436,
+    "eval_steps_per_second": 2.141
+}

Baichuan-13B-Chat-lora-Retrieval/special_tokens_map.json

@@ -0,0 +1,30 @@
+{
+  "bos_token": {
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "pad_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  }
+}

Baichuan-13B-Chat-lora-Retrieval/tokenization_baichuan.py

@@ -0,0 +1,232 @@
+# Copyright (c) 2023, Baichuan Intelligent Technology. All rights reserved.
+
+import os
+from shutil import copyfile
+from typing import Any, Dict, List, Optional, Tuple
+
+import sentencepiece as spm
+from transformers.tokenization_utils import AddedToken, PreTrainedTokenizer
+from transformers.utils import logging
+
+
+logger = logging.get_logger(__name__)
+
+VOCAB_FILES_NAMES = {"vocab_file": "tokenizer.model"}
+
+PRETRAINED_VOCAB_FILES_MAP = {
+    "vocab_file": {},
+    "tokenizer_file": {},
+}
+PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {}
+
+
+class BaichuanTokenizer(PreTrainedTokenizer):
+    """
+    Construct a Baichuan tokenizer. Based on byte-level Byte-Pair-Encoding.
+
+    Args:
+        vocab_file (`str`):
+            Path to the vocabulary file.
+    """
+
+    vocab_files_names = VOCAB_FILES_NAMES
+    pretrained_vocab_files_map = PRETRAINED_VOCAB_FILES_MAP
+    max_model_input_sizes = PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES
+    model_input_names = ["input_ids", "attention_mask"]
+
+    def __init__(
+        self,
+        vocab_file,
+        unk_token="<unk>",
+        bos_token="<s>",
+        eos_token="</s>",
+        pad_token=None,
+        sp_model_kwargs: Optional[Dict[str, Any]] = None,
+        add_bos_token=True,
+        add_eos_token=False,
+        clean_up_tokenization_spaces=False,
+        **kwargs,
+    ):
+        self.sp_model_kwargs = {} if sp_model_kwargs is None else sp_model_kwargs
+        bos_token = AddedToken(bos_token, lstrip=False, rstrip=False) if isinstance(bos_token, str) else bos_token
+        eos_token = AddedToken(eos_token, lstrip=False, rstrip=False) if isinstance(eos_token, str) else eos_token
+        unk_token = AddedToken(unk_token, lstrip=False, rstrip=False) if isinstance(unk_token, str) else unk_token
+        pad_token = AddedToken(pad_token, lstrip=False, rstrip=False) if isinstance(pad_token, str) else pad_token
+        super().__init__(
+            bos_token=bos_token,
+            eos_token=eos_token,
+            unk_token=unk_token,
+            pad_token=pad_token,
+            add_bos_token=add_bos_token,
+            add_eos_token=add_eos_token,
+            sp_model_kwargs=self.sp_model_kwargs,
+            clean_up_tokenization_spaces=clean_up_tokenization_spaces,
+            **kwargs,
+        )
+        self.vocab_file = vocab_file
+        self.add_bos_token = add_bos_token
+        self.add_eos_token = add_eos_token
+        self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
+        self.sp_model.Load(vocab_file)
+
+    def __getstate__(self):
+        state = self.__dict__.copy()
+        state["sp_model"] = None
+        return state
+
+    def __setstate__(self, d):
+        self.__dict__ = d
+        self.sp_model = spm.SentencePieceProcessor(**self.sp_model_kwargs)
+        self.sp_model.Load(self.vocab_file)
+
+    @property
+    def vocab_size(self):
+        """Returns vocab size"""
+        return self.sp_model.get_piece_size()
+
+    def get_vocab(self):
+        """Returns vocab as a dict"""
+        vocab = {self.convert_ids_to_tokens(i): i for i in range(self.vocab_size)}
+        vocab.update(self.added_tokens_encoder)
+        return vocab
+
+    def _tokenize(self, text):
+        """Returns a tokenized string."""
+        return self.sp_model.encode(text, out_type=str)
+
+    def _convert_token_to_id(self, token):
+        """Converts a token (str) in an id using the vocab."""
+        return self.sp_model.piece_to_id(token)
+
+    def _convert_id_to_token(self, index):
+        """Converts an index (integer) in a token (str) using the vocab."""
+        token = self.sp_model.IdToPiece(index)
+        return token
+
+    def convert_tokens_to_string(self, tokens):
+        """Converts a sequence of tokens (string) in a single string."""
+        current_sub_tokens = []
+        out_string = ""
+        prev_is_special = False
+        for i, token in enumerate(tokens):
+            # make sure that special tokens are not decoded using sentencepiece model
+            if token in self.all_special_tokens:
+                if not prev_is_special and i != 0:
+                    out_string += " "
+                out_string += self.sp_model.decode(current_sub_tokens) + token
+                prev_is_special = True
+                current_sub_tokens = []
+            else:
+                current_sub_tokens.append(token)
+                prev_is_special = False
+        out_string += self.sp_model.decode(current_sub_tokens)
+        return out_string
+
+    def save_vocabulary(self, save_directory, filename_prefix: Optional[str] = None) -> Tuple[str]:
+        """
+        Save the vocabulary and special tokens file to a directory.
+
+        Args:
+            save_directory (`str`):
+                The directory in which to save the vocabulary.
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        if not os.path.isdir(save_directory):
+            logger.error(f"Vocabulary path ({save_directory}) should be a directory")
+            return
+        out_vocab_file = os.path.join(
+            save_directory, (filename_prefix + "-" if filename_prefix else "") + VOCAB_FILES_NAMES["vocab_file"]
+        )
+
+        if os.path.abspath(self.vocab_file) != os.path.abspath(out_vocab_file) and os.path.isfile(self.vocab_file):
+            copyfile(self.vocab_file, out_vocab_file)
+        elif not os.path.isfile(self.vocab_file):
+            with open(out_vocab_file, "wb") as fi:
+                content_spiece_model = self.sp_model.serialized_model_proto()
+                fi.write(content_spiece_model)
+
+        return (out_vocab_file,)
+
+    def build_inputs_with_special_tokens(self, token_ids_0, token_ids_1=None):
+        bos_token_id = [self.bos_token_id] if self.add_bos_token else []
+        eos_token_id = [self.eos_token_id] if self.add_eos_token else []
+
+        output = bos_token_id + token_ids_0 + eos_token_id
+
+        if token_ids_1 is not None:
+            output = output + bos_token_id + token_ids_1 + eos_token_id
+
+        return output
+
+    def get_special_tokens_mask(
+        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, already_has_special_tokens: bool = False
+    ) -> List[int]:
+        """
+        Retrieve sequence ids from a token list that has no special tokens added. This method is called when adding
+        special tokens using the tokenizer `prepare_for_model` method.
+
+        Args:
+            token_ids_0 (`List[int]`):
+                List of IDs.
+            token_ids_1 (`List[int]`, *optional*):
+                Optional second list of IDs for sequence pairs.
+            already_has_special_tokens (`bool`, *optional*, defaults to `False`):
+                Whether or not the token list is already formatted with special tokens for the model.
+
+        Returns:
+            `List[int]`: A list of integers in the range [0, 1]: 1 for a special token, 0 for a sequence token.
+        """
+        if already_has_special_tokens:
+            return super().get_special_tokens_mask(
+                token_ids_0=token_ids_0, token_ids_1=token_ids_1, already_has_special_tokens=True
+            )
+
+        bos_token_id = [1] if self.add_bos_token else []
+        eos_token_id = [1] if self.add_eos_token else []
+
+        if token_ids_1 is None:
+            return bos_token_id + ([0] * len(token_ids_0)) + eos_token_id
+        return (
+            bos_token_id
+            + ([0] * len(token_ids_0))
+            + eos_token_id
+            + bos_token_id
+            + ([0] * len(token_ids_1))
+            + eos_token_id
+        )
+
+    def create_token_type_ids_from_sequences(
+        self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None
+    ) -> List[int]:
+        """
+        Creates a mask from the two sequences passed to be used in a sequence-pair classification task. An ALBERT
+        sequence pair mask has the following format:
+
+        ```
+        0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
+        | first sequence    | second sequence |
+        ```
+
+        if token_ids_1 is None, only returns the first portion of the mask (0s).
+
+        Args:
+            token_ids_0 (`List[int]`):
+                List of ids.
+            token_ids_1 (`List[int]`, *optional*):
+                Optional second list of IDs for sequence pairs.
+
+        Returns:
+            `List[int]`: List of [token type IDs](../glossary#token-type-ids) according to the given sequence(s).
+        """
+        bos_token_id = [self.bos_token_id] if self.add_bos_token else []
+        eos_token_id = [self.eos_token_id] if self.add_eos_token else []
+
+        output = [0] * len(bos_token_id + token_ids_0 + eos_token_id)
+
+        if token_ids_1 is not None:
+            output += [1] * len(bos_token_id + token_ids_1 + eos_token_id)
+
+        return output
+

Baichuan-13B-Chat-lora-Retrieval/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f7d1ab69d25c74644af5c5e4dcd1cc6e96d33783dbd257b6bdea55b643c72813
+size 1136765

Baichuan-13B-Chat-lora-Retrieval/tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_baichuan.BaichuanTokenizer",
+      null
+    ]
+  },
+  "bos_token": {
+    "__type": "AddedToken",
+    "content": "<s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "clean_up_tokenization_spaces": false,
+  "eos_token": {
+    "__type": "AddedToken",
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "model_max_length": 4096,
+  "pad_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  },
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "split_special_tokens": false,
+  "tokenizer_class": "BaichuanTokenizer",
+  "unk_token": {
+    "__type": "AddedToken",
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": true,
+    "rstrip": false,
+    "single_word": true
+  }
+}

Baichuan-13B-Chat-lora-Retrieval/train_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.99,
+    "train_loss": 1.0595260426618052,
+    "train_runtime": 5379.5583,
+    "train_samples_per_second": 7.407,
+    "train_steps_per_second": 0.077
+}

Baichuan-13B-Chat-lora-Retrieval/trainer_log.jsonl

@@ -0,0 +1,51 @@
+{"current_steps": 10, "total_steps": 414, "loss": 1.3488, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9928054992195985e-05, "epoch": 0.05, "percentage": 2.42, "elapsed_time": "0:02:06", "remaining_time": "1:24:56"}
+{"current_steps": 20, "total_steps": 414, "loss": 1.2097, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.971263405551576e-05, "epoch": 0.1, "percentage": 4.83, "elapsed_time": "0:04:15", "remaining_time": "1:23:44"}
+{"current_steps": 30, "total_steps": 414, "loss": 1.1719, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9354977066836986e-05, "epoch": 0.14, "percentage": 7.25, "elapsed_time": "0:06:23", "remaining_time": "1:21:51"}
+{"current_steps": 40, "total_steps": 414, "loss": 1.145, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.885714255694698e-05, "epoch": 0.19, "percentage": 9.66, "elapsed_time": "0:08:30", "remaining_time": "1:19:35"}
+{"current_steps": 50, "total_steps": 414, "loss": 1.1203, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.822199586246168e-05, "epoch": 0.24, "percentage": 12.08, "elapsed_time": "0:10:41", "remaining_time": "1:17:52"}
+{"current_steps": 50, "total_steps": 414, "loss": null, "eval_loss": 1.11279296875, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.24, "percentage": 12.08, "elapsed_time": "0:10:41", "remaining_time": "1:17:52"}
+{"current_steps": 60, "total_steps": 414, "loss": 1.1036, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.74531926340924e-05, "epoch": 0.29, "percentage": 14.49, "elapsed_time": "0:13:00", "remaining_time": "1:16:47"}
+{"current_steps": 70, "total_steps": 414, "loss": 1.0904, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.6555157796180335e-05, "epoch": 0.34, "percentage": 16.91, "elapsed_time": "0:15:08", "remaining_time": "1:14:24"}
+{"current_steps": 80, "total_steps": 414, "loss": 1.0818, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.5533060078599226e-05, "epoch": 0.39, "percentage": 19.32, "elapsed_time": "0:17:14", "remaining_time": "1:11:58"}
+{"current_steps": 90, "total_steps": 414, "loss": 1.0829, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.43927822676105e-05, "epoch": 0.43, "percentage": 21.74, "elapsed_time": "0:19:21", "remaining_time": "1:09:41"}
+{"current_steps": 100, "total_steps": 414, "loss": 1.0744, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.3140887346894974e-05, "epoch": 0.48, "percentage": 24.15, "elapsed_time": "0:21:34", "remaining_time": "1:07:43"}
+{"current_steps": 100, "total_steps": 414, "loss": null, "eval_loss": 1.065435767173767, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.48, "percentage": 24.15, "elapsed_time": "0:21:34", "remaining_time": "1:07:43"}
+{"current_steps": 110, "total_steps": 414, "loss": 1.0629, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.1784580723639923e-05, "epoch": 0.53, "percentage": 26.57, "elapsed_time": "0:23:52", "remaining_time": "1:05:58"}
+{"current_steps": 120, "total_steps": 414, "loss": 1.0659, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.033166875709291e-05, "epoch": 0.58, "percentage": 28.99, "elapsed_time": "0:25:56", "remaining_time": "1:03:33"}
+{"current_steps": 130, "total_steps": 414, "loss": 1.0523, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8790513828275683e-05, "epoch": 0.63, "percentage": 31.4, "elapsed_time": "0:28:05", "remaining_time": "1:01:21"}
+{"current_steps": 140, "total_steps": 414, "loss": 1.0397, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.716998620945871e-05, "epoch": 0.67, "percentage": 33.82, "elapsed_time": "0:30:13", "remaining_time": "0:59:09"}
+{"current_steps": 150, "total_steps": 414, "loss": 1.0435, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.547941301041661e-05, "epoch": 0.72, "percentage": 36.23, "elapsed_time": "0:32:22", "remaining_time": "0:56:58"}
+{"current_steps": 150, "total_steps": 414, "loss": null, "eval_loss": 1.048127293586731, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.72, "percentage": 36.23, "elapsed_time": "0:32:22", "remaining_time": "0:56:58"}
+{"current_steps": 160, "total_steps": 414, "loss": 1.0563, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.372852449530922e-05, "epoch": 0.77, "percentage": 38.65, "elapsed_time": "0:34:39", "remaining_time": "0:55:00"}
+{"current_steps": 170, "total_steps": 414, "loss": 1.0376, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.1927398079167226e-05, "epoch": 0.82, "percentage": 41.06, "elapsed_time": "0:36:44", "remaining_time": "0:52:44"}
+{"current_steps": 180, "total_steps": 414, "loss": 1.0436, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.008640032631585e-05, "epoch": 0.87, "percentage": 43.48, "elapsed_time": "0:38:52", "remaining_time": "0:50:32"}
+{"current_steps": 190, "total_steps": 414, "loss": 1.0301, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.821612728457078e-05, "epoch": 0.92, "percentage": 45.89, "elapsed_time": "0:41:04", "remaining_time": "0:48:25"}
+{"current_steps": 200, "total_steps": 414, "loss": 1.0338, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.632734349861874e-05, "epoch": 0.96, "percentage": 48.31, "elapsed_time": "0:43:13", "remaining_time": "0:46:14"}
+{"current_steps": 200, "total_steps": 414, "loss": null, "eval_loss": 1.0388035774230957, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.96, "percentage": 48.31, "elapsed_time": "0:43:13", "remaining_time": "0:46:14"}
+{"current_steps": 210, "total_steps": 414, "loss": 1.033, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.4430920053597356e-05, "epoch": 1.01, "percentage": 50.72, "elapsed_time": "0:45:29", "remaining_time": "0:44:11"}
+{"current_steps": 220, "total_steps": 414, "loss": 1.0399, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.2537772005470782e-05, "epoch": 1.06, "percentage": 53.14, "elapsed_time": "0:47:37", "remaining_time": "0:41:59"}
+{"current_steps": 230, "total_steps": 414, "loss": 1.0277, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.0658795558326743e-05, "epoch": 1.11, "percentage": 55.56, "elapsed_time": "0:49:44", "remaining_time": "0:39:47"}
+{"current_steps": 240, "total_steps": 414, "loss": 1.0325, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.8804805350177505e-05, "epoch": 1.16, "percentage": 57.97, "elapsed_time": "0:51:54", "remaining_time": "0:37:38"}
+{"current_steps": 250, "total_steps": 414, "loss": 1.0322, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.6986472208222576e-05, "epoch": 1.2, "percentage": 60.39, "elapsed_time": "0:54:02", "remaining_time": "0:35:26"}
+{"current_steps": 250, "total_steps": 414, "loss": null, "eval_loss": 1.0336002111434937, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.2, "percentage": 60.39, "elapsed_time": "0:54:02", "remaining_time": "0:35:26"}
+{"current_steps": 260, "total_steps": 414, "loss": 1.0248, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.5214261731829022e-05, "epoch": 1.25, "percentage": 62.8, "elapsed_time": "0:56:17", "remaining_time": "0:33:20"}
+{"current_steps": 270, "total_steps": 414, "loss": 1.0235, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.3498374056721197e-05, "epoch": 1.3, "percentage": 65.22, "elapsed_time": "0:58:28", "remaining_time": "0:31:11"}
+{"current_steps": 280, "total_steps": 414, "loss": 1.0446, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1848685147073222e-05, "epoch": 1.35, "percentage": 67.63, "elapsed_time": "1:00:38", "remaining_time": "0:29:01"}
+{"current_steps": 290, "total_steps": 414, "loss": 1.0271, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.0274689953403407e-05, "epoch": 1.4, "percentage": 70.05, "elapsed_time": "1:02:46", "remaining_time": "0:26:50"}
+{"current_steps": 300, "total_steps": 414, "loss": 1.0351, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.785447763431101e-06, "epoch": 1.44, "percentage": 72.46, "elapsed_time": "1:04:51", "remaining_time": "0:24:38"}
+{"current_steps": 300, "total_steps": 414, "loss": null, "eval_loss": 1.031198263168335, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.44, "percentage": 72.46, "elapsed_time": "1:04:51", "remaining_time": "0:24:38"}
+{"current_steps": 310, "total_steps": 414, "loss": 1.0341, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.389530060434696e-06, "epoch": 1.49, "percentage": 74.88, "elapsed_time": "1:07:07", "remaining_time": "0:22:31"}
+{"current_steps": 320, "total_steps": 414, "loss": 1.0267, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.094971189217042e-06, "epoch": 1.54, "percentage": 77.29, "elapsed_time": "1:09:15", "remaining_time": "0:20:20"}
+{"current_steps": 330, "total_steps": 414, "loss": 1.0122, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9092221136255444e-06, "epoch": 1.59, "percentage": 79.71, "elapsed_time": "1:11:24", "remaining_time": "0:18:10"}
+{"current_steps": 340, "total_steps": 414, "loss": 1.0197, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.839107531779978e-06, "epoch": 1.64, "percentage": 82.13, "elapsed_time": "1:13:31", "remaining_time": "0:16:00"}
+{"current_steps": 350, "total_steps": 414, "loss": 1.0193, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.890786595835693e-06, "epoch": 1.69, "percentage": 84.54, "elapsed_time": "1:15:41", "remaining_time": "0:13:50"}
+{"current_steps": 350, "total_steps": 414, "loss": null, "eval_loss": 1.0297337770462036, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.69, "percentage": 84.54, "elapsed_time": "1:15:41", "remaining_time": "0:13:50"}
+{"current_steps": 360, "total_steps": 414, "loss": 1.0231, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.0697174623636794e-06, "epoch": 1.73, "percentage": 86.96, "elapsed_time": "1:17:59", "remaining_time": "0:11:41"}
+{"current_steps": 370, "total_steps": 414, "loss": 1.0233, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.3806258773811476e-06, "epoch": 1.78, "percentage": 89.37, "elapsed_time": "1:20:09", "remaining_time": "0:09:31"}
+{"current_steps": 380, "total_steps": 414, "loss": 1.0224, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.274779768448482e-07, "epoch": 1.83, "percentage": 91.79, "elapsed_time": "1:22:18", "remaining_time": "0:07:21"}
+{"current_steps": 390, "total_steps": 414, "loss": 1.0181, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.134574591564494e-07, "epoch": 1.88, "percentage": 94.2, "elapsed_time": "1:24:25", "remaining_time": "0:05:11"}
+{"current_steps": 400, "total_steps": 414, "loss": 1.0189, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.4094726106603505e-07, "epoch": 1.93, "percentage": 96.62, "elapsed_time": "1:26:32", "remaining_time": "0:03:01"}
+{"current_steps": 400, "total_steps": 414, "loss": null, "eval_loss": 1.0294359922409058, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.93, "percentage": 96.62, "elapsed_time": "1:26:32", "remaining_time": "0:03:01"}
+{"current_steps": 410, "total_steps": 414, "loss": 1.0244, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 1.1515842439871472e-08, "epoch": 1.97, "percentage": 99.03, "elapsed_time": "1:28:47", "remaining_time": "0:00:51"}
+{"current_steps": 414, "total_steps": 414, "loss": null, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.99, "percentage": 100.0, "elapsed_time": "1:29:38", "remaining_time": "0:00:00"}
+{"current_steps": 17, "total_steps": 17, "loss": null, "eval_loss": 1.0294359922409058, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.99, "percentage": 100.0, "elapsed_time": "1:29:46", "remaining_time": "0:00:00"}

Baichuan-13B-Chat-lora-Retrieval/trainer_state.json

@@ -0,0 +1,335 @@
+{
+  "best_metric": 1.0294359922409058,
+  "best_model_checkpoint": "output/Baichuan-13B-Chat_lora_wqs_jiansuo/checkpoint-400",
+  "epoch": 1.9939795304033714,
+  "global_step": 414,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9928054992195985e-05,
+      "loss": 1.3488,
+      "step": 10
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.971263405551576e-05,
+      "loss": 1.2097,
+      "step": 20
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9354977066836986e-05,
+      "loss": 1.1719,
+      "step": 30
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.885714255694698e-05,
+      "loss": 1.145,
+      "step": 40
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.822199586246168e-05,
+      "loss": 1.1203,
+      "step": 50
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 1.11279296875,
+      "eval_runtime": 10.7965,
+      "eval_samples_per_second": 18.71,
+      "eval_steps_per_second": 1.575,
+      "step": 50
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.74531926340924e-05,
+      "loss": 1.1036,
+      "step": 60
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6555157796180335e-05,
+      "loss": 1.0904,
+      "step": 70
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5533060078599226e-05,
+      "loss": 1.0818,
+      "step": 80
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.43927822676105e-05,
+      "loss": 1.0829,
+      "step": 90
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3140887346894974e-05,
+      "loss": 1.0744,
+      "step": 100
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 1.065435767173767,
+      "eval_runtime": 7.9424,
+      "eval_samples_per_second": 25.433,
+      "eval_steps_per_second": 2.14,
+      "step": 100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.1784580723639923e-05,
+      "loss": 1.0629,
+      "step": 110
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.033166875709291e-05,
+      "loss": 1.0659,
+      "step": 120
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 3.8790513828275683e-05,
+      "loss": 1.0523,
+      "step": 130
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 3.716998620945871e-05,
+      "loss": 1.0397,
+      "step": 140
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.547941301041661e-05,
+      "loss": 1.0435,
+      "step": 150
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 1.048127293586731,
+      "eval_runtime": 8.0034,
+      "eval_samples_per_second": 25.239,
+      "eval_steps_per_second": 2.124,
+      "step": 150
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 3.372852449530922e-05,
+      "loss": 1.0563,
+      "step": 160
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.1927398079167226e-05,
+      "loss": 1.0376,
+      "step": 170
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 3.008640032631585e-05,
+      "loss": 1.0436,
+      "step": 180
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.821612728457078e-05,
+      "loss": 1.0301,
+      "step": 190
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.632734349861874e-05,
+      "loss": 1.0338,
+      "step": 200
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 1.0388035774230957,
+      "eval_runtime": 7.9339,
+      "eval_samples_per_second": 25.46,
+      "eval_steps_per_second": 2.143,
+      "step": 200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.4430920053597356e-05,
+      "loss": 1.033,
+      "step": 210
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.2537772005470782e-05,
+      "loss": 1.0399,
+      "step": 220
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 1.0277,
+      "step": 230
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.8804805350177505e-05,
+      "loss": 1.0325,
+      "step": 240
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.6986472208222576e-05,
+      "loss": 1.0322,
+      "step": 250
+    },
+    {
+      "epoch": 1.2,
+      "eval_loss": 1.0336002111434937,
+      "eval_runtime": 7.9742,
+      "eval_samples_per_second": 25.332,
+      "eval_steps_per_second": 2.132,
+      "step": 250
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.5214261731829022e-05,
+      "loss": 1.0248,
+      "step": 260
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3498374056721197e-05,
+      "loss": 1.0235,
+      "step": 270
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.1848685147073222e-05,
+      "loss": 1.0446,
+      "step": 280
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.0274689953403407e-05,
+      "loss": 1.0271,
+      "step": 290
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 8.785447763431101e-06,
+      "loss": 1.0351,
+      "step": 300
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 1.031198263168335,
+      "eval_runtime": 7.9378,
+      "eval_samples_per_second": 25.448,
+      "eval_steps_per_second": 2.142,
+      "step": 300
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 7.389530060434696e-06,
+      "loss": 1.0341,
+      "step": 310
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 6.094971189217042e-06,
+      "loss": 1.0267,
+      "step": 320
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 4.9092221136255444e-06,
+      "loss": 1.0122,
+      "step": 330
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 3.839107531779978e-06,
+      "loss": 1.0197,
+      "step": 340
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.890786595835693e-06,
+      "loss": 1.0193,
+      "step": 350
+    },
+    {
+      "epoch": 1.69,
+      "eval_loss": 1.0297337770462036,
+      "eval_runtime": 7.9869,
+      "eval_samples_per_second": 25.292,
+      "eval_steps_per_second": 2.128,
+      "step": 350
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.0697174623636794e-06,
+      "loss": 1.0231,
+      "step": 360
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.3806258773811476e-06,
+      "loss": 1.0233,
+      "step": 370
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 8.274779768448482e-07,
+      "loss": 1.0224,
+      "step": 380
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 4.134574591564494e-07,
+      "loss": 1.0181,
+      "step": 390
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.4094726106603505e-07,
+      "loss": 1.0189,
+      "step": 400
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.0294359922409058,
+      "eval_runtime": 7.9102,
+      "eval_samples_per_second": 25.537,
+      "eval_steps_per_second": 2.149,
+      "step": 400
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.1515842439871472e-08,
+      "loss": 1.0244,
+      "step": 410
+    },
+    {
+      "epoch": 1.99,
+      "step": 414,
+      "total_flos": 3.076881712207102e+18,
+      "train_loss": 1.0595260426618052,
+      "train_runtime": 5379.5583,
+      "train_samples_per_second": 7.407,
+      "train_steps_per_second": 0.077
+    }
+  ],
+  "max_steps": 414,
+  "num_train_epochs": 2,
+  "total_flos": 3.076881712207102e+18,
+  "trial_name": null,
+  "trial_params": null
+}

Baichuan-13B-Chat-lora-Retrieval/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0517bc6b507559268a02b20e558d9c2fc2bdf7c35709cc817272afdf90eb8b7d
+size 4600