Upload 14 files

README.md

@@ -0,0 +1,63 @@
+---
+license: other
+library_name: peft
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+base_model: baichuan-inc/Baichuan2-7B-Chat
+model-index:
+- name: lora1
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# lora1
+
+This model is a fine-tuned version of [baichuan-inc/Baichuan2-7B-Chat](https://huggingface.co/baichuan-inc/Baichuan2-7B-Chat) on the identity and the alpaca_en_demo datasets.
+It achieves the following results on the evaluation set:
+- Loss: 1.1241
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0001
+- train_batch_size: 8
+- eval_batch_size: 1
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 2
+- total_train_batch_size: 16
+- total_eval_batch_size: 2
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 2
+
+### Training results
+
+
+
+### Framework versions
+
+- PEFT 0.11.1
+- Transformers 4.41.2
+- Pytorch 2.2.1+cu121
+- Datasets 2.18.0
+- Tokenizers 0.19.1

adapter_config.json

@@ -0,0 +1,32 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "baichuan-inc/Baichuan2-7B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "W_pack",
+    "down_proj",
+    "o_proj",
+    "gate_proj",
+    "up_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:193e9efc4360517708a7e800fd7703064a3ef9d8856fef033f5dbe4286321a5d
+size 71607456

all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 1.1241382360458374,
+    "eval_runtime": 2.8045,
+    "eval_samples_per_second": 39.222,
+    "eval_steps_per_second": 19.611,
+    "total_flos": 3.343333284695245e+16,
+    "train_loss": 1.1469936832304923,
+    "train_runtime": 212.0457,
+    "train_samples_per_second": 9.253,
+    "train_steps_per_second": 0.585
+}

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 2.0,
+    "eval_loss": 1.1241382360458374,
+    "eval_runtime": 2.8045,
+    "eval_samples_per_second": 39.222,
+    "eval_steps_per_second": 19.611
+}

special_tokens_map.json

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

tokenization_baichuan.py

@@ -0,0 +1,253 @@
+# Copyright 2023 Baichuan Inc. All Rights Reserved.
+
+# Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved.
+#
+# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
+# and OPT implementations in this library. It has been modified from its
+# original forms to accommodate minor architectural differences compared
+# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+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

tokenizer.model

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

tokenizer_config.json

@@ -0,0 +1,48 @@
+{
+  "add_bos_token": false,
+  "add_eos_token": false,
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "<s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": true,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    }
+  },
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_baichuan.BaichuanTokenizer",
+      null
+    ]
+  },
+  "bos_token": "<s>",
+  "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ '<reserved_106>' + content + '<reserved_107>' }}{% elif message['role'] == 'assistant' %}{{ content }}{% endif %}{% endfor %}",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "</s>",
+  "model_max_length": 4096,
+  "pad_token": "<unk>",
+  "padding_side": "right",
+  "sp_model_kwargs": {},
+  "split_special_tokens": false,
+  "tokenizer_class": "BaichuanTokenizer",
+  "unk_token": "<unk>",
+  "use_fast": false
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 2.0,
+    "total_flos": 3.343333284695245e+16,
+    "train_loss": 1.1469936832304923,
+    "train_runtime": 212.0457,
+    "train_samples_per_second": 9.253,
+    "train_steps_per_second": 0.585
+}

trainer_log.jsonl

@@ -0,0 +1,13 @@
+{"current_steps": 10, "total_steps": 124, "loss": 1.4799, "learning_rate": 7.692307692307693e-05, "epoch": 0.16129032258064516, "percentage": 8.06, "elapsed_time": "0:00:15", "remaining_time": "0:03:01", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 20, "total_steps": 124, "loss": 1.2357, "learning_rate": 9.902193239806635e-05, "epoch": 0.3225806451612903, "percentage": 16.13, "elapsed_time": "0:00:32", "remaining_time": "0:02:50", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 30, "total_steps": 124, "loss": 1.1836, "learning_rate": 9.432328436130493e-05, "epoch": 0.4838709677419355, "percentage": 24.19, "elapsed_time": "0:00:49", "remaining_time": "0:02:34", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 40, "total_steps": 124, "loss": 1.1089, "learning_rate": 8.609780469772623e-05, "epoch": 0.6451612903225806, "percentage": 32.26, "elapsed_time": "0:01:06", "remaining_time": "0:02:18", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 50, "total_steps": 124, "loss": 1.1531, "learning_rate": 7.500000000000001e-05, "epoch": 0.8064516129032258, "percentage": 40.32, "elapsed_time": "0:01:24", "remaining_time": "0:02:04", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 60, "total_steps": 124, "loss": 1.1684, "learning_rate": 6.191292957115825e-05, "epoch": 0.967741935483871, "percentage": 48.39, "elapsed_time": "0:01:40", "remaining_time": "0:01:46", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 70, "total_steps": 124, "loss": 1.1087, "learning_rate": 4.78779398401926e-05, "epoch": 1.129032258064516, "percentage": 56.45, "elapsed_time": "0:01:57", "remaining_time": "0:01:30", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 80, "total_steps": 124, "loss": 1.13, "learning_rate": 3.401180377143774e-05, "epoch": 1.2903225806451613, "percentage": 64.52, "elapsed_time": "0:02:14", "remaining_time": "0:01:13", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 90, "total_steps": 124, "loss": 1.0524, "learning_rate": 2.141785853707607e-05, "epoch": 1.4516129032258065, "percentage": 72.58, "elapsed_time": "0:02:32", "remaining_time": "0:00:57", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 100, "total_steps": 124, "loss": 1.0149, "learning_rate": 1.1098212284078036e-05, "epoch": 1.6129032258064515, "percentage": 80.65, "elapsed_time": "0:02:49", "remaining_time": "0:00:40", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 110, "total_steps": 124, "loss": 1.0677, "learning_rate": 3.87400575837657e-06, "epoch": 1.7741935483870968, "percentage": 88.71, "elapsed_time": "0:03:08", "remaining_time": "0:00:23", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 120, "total_steps": 124, "loss": 1.0709, "learning_rate": 3.2007361901485455e-07, "epoch": 1.935483870967742, "percentage": 96.77, "elapsed_time": "0:03:23", "remaining_time": "0:00:06", "throughput": "0.00", "total_tokens": 0}
+{"current_steps": 124, "total_steps": 124, "epoch": 2.0, "percentage": 100.0, "elapsed_time": "0:03:29", "remaining_time": "0:00:00", "throughput": "0.00", "total_tokens": 0}

trainer_state.json

@@ -0,0 +1,126 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 124,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.16129032258064516,
+      "grad_norm": 1.217032551765442,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 1.4799,
+      "step": 10
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 0.49437594413757324,
+      "learning_rate": 9.902193239806635e-05,
+      "loss": 1.2357,
+      "step": 20
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.38023412227630615,
+      "learning_rate": 9.432328436130493e-05,
+      "loss": 1.1836,
+      "step": 30
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 0.30302199721336365,
+      "learning_rate": 8.609780469772623e-05,
+      "loss": 1.1089,
+      "step": 40
+    },
+    {
+      "epoch": 0.8064516129032258,
+      "grad_norm": 0.3546346127986908,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.1531,
+      "step": 50
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.32698196172714233,
+      "learning_rate": 6.191292957115825e-05,
+      "loss": 1.1684,
+      "step": 60
+    },
+    {
+      "epoch": 1.129032258064516,
+      "grad_norm": 0.40984147787094116,
+      "learning_rate": 4.78779398401926e-05,
+      "loss": 1.1087,
+      "step": 70
+    },
+    {
+      "epoch": 1.2903225806451613,
+      "grad_norm": 0.6768341064453125,
+      "learning_rate": 3.401180377143774e-05,
+      "loss": 1.13,
+      "step": 80
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 0.38988226652145386,
+      "learning_rate": 2.141785853707607e-05,
+      "loss": 1.0524,
+      "step": 90
+    },
+    {
+      "epoch": 1.6129032258064515,
+      "grad_norm": 0.43906646966934204,
+      "learning_rate": 1.1098212284078036e-05,
+      "loss": 1.0149,
+      "step": 100
+    },
+    {
+      "epoch": 1.7741935483870968,
+      "grad_norm": 0.7597775459289551,
+      "learning_rate": 3.87400575837657e-06,
+      "loss": 1.0677,
+      "step": 110
+    },
+    {
+      "epoch": 1.935483870967742,
+      "grad_norm": 0.4260946810245514,
+      "learning_rate": 3.2007361901485455e-07,
+      "loss": 1.0709,
+      "step": 120
+    },
+    {
+      "epoch": 2.0,
+      "step": 124,
+      "total_flos": 3.343333284695245e+16,
+      "train_loss": 1.1469936832304923,
+      "train_runtime": 212.0457,
+      "train_samples_per_second": 9.253,
+      "train_steps_per_second": 0.585
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 124,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.343333284695245e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6b6a75705aea085b954f6ffa255790c2959a722f5e0351c026567198c802a6f7
+size 5240