Upload 74 files

user-baichuan2-13b-v2-3.6/README.md

@@ -0,0 +1,23 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+### Framework versions
+
+
+- PEFT 0.4.0

user-baichuan2-13b-v2-3.6/adapter_config.json

@@ -0,0 +1,24 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/jiakangxiang/.cache/modelscope/hub/baichuan-inc/Baichuan2-13B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "W_pack",
+    "down_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

user-baichuan2-13b-v2-3.6/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5fa2928717257a823e5ece47fa40497bfd62df2de1dad1d22b7189be1eaae1fc
+size 223203704

user-baichuan2-13b-v2-3.6/all_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 1.0,
+    "train_loss": 0.5017403132133569,
+    "train_runtime": 75900.5046,
+    "train_samples_per_second": 0.102,
+    "train_steps_per_second": 0.006
+}

user-baichuan2-13b-v2-3.6/checkpoint-200/README.md

@@ -0,0 +1,53 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+### Framework versions
+
+- PEFT 0.4.0
+- PEFT 0.4.0
+
+- PEFT 0.4.0

user-baichuan2-13b-v2-3.6/checkpoint-200/adapter_config.json

@@ -0,0 +1,24 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/jiakangxiang/.cache/modelscope/hub/baichuan-inc/Baichuan2-13B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "W_pack",
+    "down_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

user-baichuan2-13b-v2-3.6/checkpoint-200/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:38a30f766c18c946f733b21a20bbed87b6c9fc7fcd352632d9f10275b9bcafec
+size 223203704

user-baichuan2-13b-v2-3.6/checkpoint-200/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:244f09da90f8637279a30da6c1ab06000f824a132e5721b107eae147dab5ec76
+size 446541509

user-baichuan2-13b-v2-3.6/checkpoint-200/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:82057afe356e5d07dcf306e6c4328a22809f76704092cbfcfc589f5f6ca4ecfb
+size 14575

user-baichuan2-13b-v2-3.6/checkpoint-200/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7d91431adcc602a99f3b7dfb114d8f98fcd9283452a4e00f4d0e20d836e409d
+size 627

user-baichuan2-13b-v2-3.6/checkpoint-200/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
+  }
+}

user-baichuan2-13b-v2-3.6/checkpoint-200/tokenization_baichuan.py

@@ -0,0 +1,258 @@
+# 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
+        )
+        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)
+        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,
+        )
+
+    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

user-baichuan2-13b-v2-3.6/checkpoint-200/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79452955be6b419a65984273a9f08af86042e1c2a75ee3ba989cbf620a133cc2
+size 2001107

user-baichuan2-13b-v2-3.6/checkpoint-200/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    }
+  },
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_baichuan.BaichuanTokenizer",
+      null
+    ]
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "model_max_length": 4096,
+  "pad_token": "<unk>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "BaichuanTokenizer",
+  "unk_token": "<unk>"
+}

user-baichuan2-13b-v2-3.6/checkpoint-200/trainer_state.json

@@ -0,0 +1,161 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.41450777202072536,
+  "eval_steps": 500,
+  "global_step": 200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.99941873550415,
+      "learning_rate": 2e-05,
+      "loss": 9.9329,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.741065502166748,
+      "learning_rate": 4e-05,
+      "loss": 11.0746,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4727320671081543,
+      "learning_rate": 6e-05,
+      "loss": 2.7159,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1335960477590561,
+      "learning_rate": 8e-05,
+      "loss": 0.3969,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0014472692273557186,
+      "learning_rate": 0.0001,
+      "loss": 0.0032,
+      "step": 50
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0010780546581372619,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.03132963180542,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 70
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.008827299810945988,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 80
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0002956670359708369,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 90
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0003419867134653032,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0003681881644297391,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0002884200366679579,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.00011985149467363954,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0003195986500941217,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.00010149635636480525,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.00010508792183827609,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.00011793687008321285,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.076676749624312e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0007808339432813227,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11711683869361877,
+      "learning_rate": 0.0001,
+      "loss": 0.003,
+      "step": 200
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 482,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 3.335841878562816e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

user-baichuan2-13b-v2-3.6/checkpoint-200/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1abab34fc571ab2be46c8abdf765b96b9a09ab4144528e95270f1af465c0f19c
+size 4475

user-baichuan2-13b-v2-3.6/checkpoint-300/README.md

@@ -0,0 +1,23 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+### Framework versions
+
+
+- PEFT 0.4.0

user-baichuan2-13b-v2-3.6/checkpoint-300/adapter_config.json

@@ -0,0 +1,24 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/jiakangxiang/.cache/modelscope/hub/baichuan-inc/Baichuan2-13B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "W_pack",
+    "down_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

user-baichuan2-13b-v2-3.6/checkpoint-300/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3abb43148c7b50ee5cbd1d6ca51b35ce5a55e74d837bf441cb247d3ea6a7c56a
+size 223203704

user-baichuan2-13b-v2-3.6/checkpoint-300/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6626b117cfcf7dfb7c1d3cf79bd87cac30bfdbb5b6d6138c5282bda6d59f86cb
+size 446541893

user-baichuan2-13b-v2-3.6/checkpoint-300/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:939982b34814a29d20630db8d8bb9ca8ffaaca659852333f9bf40ca5f715ffd0
+size 14575

user-baichuan2-13b-v2-3.6/checkpoint-300/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47e1286a69d6cf9d5865b0808d9a438a85cffe270c5273c3518b2a5557084aa5
+size 627

user-baichuan2-13b-v2-3.6/checkpoint-300/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
+  }
+}

user-baichuan2-13b-v2-3.6/checkpoint-300/tokenization_baichuan.py

@@ -0,0 +1,258 @@
+# 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
+        )
+        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)
+        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,
+        )
+
+    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

user-baichuan2-13b-v2-3.6/checkpoint-300/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79452955be6b419a65984273a9f08af86042e1c2a75ee3ba989cbf620a133cc2
+size 2001107

user-baichuan2-13b-v2-3.6/checkpoint-300/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    }
+  },
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_baichuan.BaichuanTokenizer",
+      null
+    ]
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "model_max_length": 4096,
+  "pad_token": "<unk>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "BaichuanTokenizer",
+  "unk_token": "<unk>"
+}

user-baichuan2-13b-v2-3.6/checkpoint-300/trainer_state.json

@@ -0,0 +1,231 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6217616580310881,
+  "eval_steps": 500,
+  "global_step": 300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.99941873550415,
+      "learning_rate": 2e-05,
+      "loss": 9.9329,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.741065502166748,
+      "learning_rate": 4e-05,
+      "loss": 11.0746,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4727320671081543,
+      "learning_rate": 6e-05,
+      "loss": 2.7159,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1335960477590561,
+      "learning_rate": 8e-05,
+      "loss": 0.3969,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0014472692273557186,
+      "learning_rate": 0.0001,
+      "loss": 0.0032,
+      "step": 50
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0010780546581372619,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.03132963180542,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 70
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.008827299810945988,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 80
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0002956670359708369,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 90
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0003419867134653032,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0003681881644297391,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0002884200366679579,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.00011985149467363954,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0003195986500941217,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.00010149635636480525,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.00010508792183827609,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.00011793687008321285,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.076676749624312e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0007808339432813227,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11711683869361877,
+      "learning_rate": 0.0001,
+      "loss": 0.003,
+      "step": 200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0002039404644165188,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.00873592495918274,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0506539344787598,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05903371796011925,
+      "learning_rate": 0.0001,
+      "loss": 0.0026,
+      "step": 240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0002484666183590889,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0003493047261144966,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0008058947860263288,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0004198936221655458,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0002983050071634352,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0002279053587699309,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 300
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 482,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 5.004587702980301e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

user-baichuan2-13b-v2-3.6/checkpoint-300/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1abab34fc571ab2be46c8abdf765b96b9a09ab4144528e95270f1af465c0f19c
+size 4475

user-baichuan2-13b-v2-3.6/checkpoint-400/README.md

@@ -0,0 +1,23 @@
+---
+library_name: peft
+---
+## Training procedure
+
+
+The following `bitsandbytes` quantization config was used during training:
+- quant_method: bitsandbytes
+- _load_in_8bit: False
+- _load_in_4bit: True
+- llm_int8_threshold: 6.0
+- llm_int8_skip_modules: None
+- llm_int8_enable_fp32_cpu_offload: False
+- llm_int8_has_fp16_weight: False
+- bnb_4bit_quant_type: nf4
+- bnb_4bit_use_double_quant: True
+- bnb_4bit_compute_dtype: float16
+- load_in_4bit: True
+- load_in_8bit: False
+### Framework versions
+
+
+- PEFT 0.4.0

user-baichuan2-13b-v2-3.6/checkpoint-400/adapter_config.json

@@ -0,0 +1,24 @@
+{
+  "auto_mapping": null,
+  "base_model_name_or_path": "/home/jiakangxiang/.cache/modelscope/hub/baichuan-inc/Baichuan2-13B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "revision": null,
+  "target_modules": [
+    "o_proj",
+    "W_pack",
+    "down_proj",
+    "up_proj",
+    "gate_proj"
+  ],
+  "task_type": "CAUSAL_LM"
+}

user-baichuan2-13b-v2-3.6/checkpoint-400/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8a761a3ff3848cad96ebcb4d93d82af85ae86c2703ec031d10338d35d93aff15
+size 223203704

user-baichuan2-13b-v2-3.6/checkpoint-400/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8b55a4a50d2c66907dc3b95e3f979aac9b320a415e3a978859a220e686a76d9a
+size 446541893

user-baichuan2-13b-v2-3.6/checkpoint-400/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54a11743bad40439e1bbf592fb8f66b1a3c8dbde2539b8897aec5e85c29fcc1c
+size 14575

user-baichuan2-13b-v2-3.6/checkpoint-400/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6536a67a17be52e1d2b7b314f3abff272bd0f976aca319628b666d64bd161a64
+size 627

user-baichuan2-13b-v2-3.6/checkpoint-400/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
+  }
+}

user-baichuan2-13b-v2-3.6/checkpoint-400/tokenization_baichuan.py

@@ -0,0 +1,258 @@
+# 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
+        )
+        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)
+        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,
+        )
+
+    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

user-baichuan2-13b-v2-3.6/checkpoint-400/tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79452955be6b419a65984273a9f08af86042e1c2a75ee3ba989cbf620a133cc2
+size 2001107

user-baichuan2-13b-v2-3.6/checkpoint-400/tokenizer_config.json

@@ -0,0 +1,44 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": true,
+      "special": true
+    }
+  },
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_baichuan.BaichuanTokenizer",
+      null
+    ]
+  },
+  "bos_token": "<s>",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "model_max_length": 4096,
+  "pad_token": "<unk>",
+  "sp_model_kwargs": {},
+  "tokenizer_class": "BaichuanTokenizer",
+  "unk_token": "<unk>"
+}

user-baichuan2-13b-v2-3.6/checkpoint-400/trainer_state.json

@@ -0,0 +1,301 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8290155440414507,
+  "eval_steps": 500,
+  "global_step": 400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.99941873550415,
+      "learning_rate": 2e-05,
+      "loss": 9.9329,
+      "step": 10
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.741065502166748,
+      "learning_rate": 4e-05,
+      "loss": 11.0746,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4727320671081543,
+      "learning_rate": 6e-05,
+      "loss": 2.7159,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1335960477590561,
+      "learning_rate": 8e-05,
+      "loss": 0.3969,
+      "step": 40
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0014472692273557186,
+      "learning_rate": 0.0001,
+      "loss": 0.0032,
+      "step": 50
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0010780546581372619,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 60
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.03132963180542,
+      "learning_rate": 0.0001,
+      "loss": 0.0002,
+      "step": 70
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.008827299810945988,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 80
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0002956670359708369,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 90
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0003419867134653032,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 100
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0003681881644297391,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 110
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0002884200366679579,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.00011985149467363954,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0003195986500941217,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 140
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.00010149635636480525,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.00010508792183827609,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 160
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.00011793687008321285,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 170
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.076676749624312e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 180
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0007808339432813227,
+      "learning_rate": 0.0001,
+      "loss": 0.006,
+      "step": 190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.11711683869361877,
+      "learning_rate": 0.0001,
+      "loss": 0.003,
+      "step": 200
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0002039404644165188,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 210
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.00873592495918274,
+      "learning_rate": 0.0001,
+      "loss": 0.0209,
+      "step": 220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0506539344787598,
+      "learning_rate": 0.0001,
+      "loss": 0.0201,
+      "step": 230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05903371796011925,
+      "learning_rate": 0.0001,
+      "loss": 0.0026,
+      "step": 240
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0002484666183590889,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0003493047261144966,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0008058947860263288,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0004198936221655458,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 280
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.0002983050071634352,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 290
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.0002279053587699309,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.00015332824841607362,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 310
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.00011723622446879745,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.0001235378731507808,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.00010625163122313097,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.50239341869019e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 350
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.00010148331784876063,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 360
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.368126534856856e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.00012744461128022522,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 380
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.87971335335169e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0,
+      "step": 390
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.405858584912494e-05,
+      "learning_rate": 0.0001,
+      "loss": 0.0001,
+      "step": 400
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 482,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 6.625291354656461e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

user-baichuan2-13b-v2-3.6/checkpoint-400/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1abab34fc571ab2be46c8abdf765b96b9a09ab4144528e95270f1af465c0f19c
+size 4475

user-baichuan2-13b-v2-3.6/runs/Mar06_16-02-53_u/events.out.tfevents.1709741241.u.349083.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3014282355673d8fdcaa570d4c5296d240c5dd75bf87a842af9ec4b76a4e9116
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_16-15-19_u/events.out.tfevents.1709741991.u.349593.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9058c4ffc24a76a37dcc37c3d6b8fc0c8aa81301fb6b58fe0305a01b7081d8f2
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_16-27-57_u/events.out.tfevents.1709742755.u.350734.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a35a2b472b2864493ef9958235b77f8c4c7af27b4b88b478d703f99b4f10f718
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_16-37-25_u/events.out.tfevents.1709743386.u.351776.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c388e6f60bbcf55c3e512b8af87941e03d9a7648390ab77bfe714ffdc48c7bac
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_16-46-23_u/events.out.tfevents.1709743925.u.352180.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a087c12a11786c7f7b12c642ee0a6c4098d3859d1e2df9abf3e218f2e3365ec
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_16-55-14_u/events.out.tfevents.1709744402.u.352650.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0aca9a5ecc84224df331de22088bfa3fe20a25c6903fd402a0be863a9566699d
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_17-03-22_u/events.out.tfevents.1709744890.u.353116.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dfea604250d4aac14bc3ed7c865337977033589c98558e199805f39e062fa8d8
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_17-13-29_u/events.out.tfevents.1709745516.u.353684.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b3e44490b8dd88a5a161c7a6b9a835026ef7a8218a9fb171acaa8737033db7
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_17-30-51_u/events.out.tfevents.1709746552.u.354572.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b2a7c82c63e3cb38c34975b02f2f0e377684b0dfa72aa1dfcfaad954da7c93e2
+size 5162

user-baichuan2-13b-v2-3.6/runs/Mar06_17-42-56_u/events.out.tfevents.1709747302.u.355650.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cf7786e2daf0473ba392d547dc007aaa018e30f7c97ce3bca26d196f68798142
+size 5164