add EN

LLM-Detector-44w-EN/README.md

@@ -0,0 +1,56 @@
+---
+license: other
+base_model: ./Qwen-1_8B-Chat
+tags:
+- llama-factory
+- lora
+- generated_from_trainer
+model-index:
+- name: qwen-1.8b-deepfake
+  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. -->
+
+# qwen-1.8b-deepfake
+
+This model is a fine-tuned version of [./Qwen-1_8B-Chat](https://huggingface.co/./Qwen-1_8B-Chat) on the df dataset.
+
+## 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: 5e-05
+- train_batch_size: 8
+- eval_batch_size: 8
+- seed: 42
+- gradient_accumulation_steps: 4
+- total_train_batch_size: 32
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: cosine
+- num_epochs: 3.0
+
+### Training results
+
+
+
+### Framework versions
+
+- Transformers 4.33.0
+- Pytorch 2.1.1+cu121
+- Datasets 2.14.7
+- Tokenizers 0.13.3

LLM-Detector-44w-EN/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "./Qwen-1_8B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "task_type": "CAUSAL_LM"
+}

LLM-Detector-44w-EN/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7313389f2f7fe453280142487569a71a98cce4c20f6ba8b25f2f9a10d6f0a10a
+size 6308670

LLM-Detector-44w-EN/all_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 3.0,
+    "train_loss": 0.05192308982466081,
+    "train_runtime": 264704.0502,
+    "train_samples_per_second": 4.904,
+    "train_steps_per_second": 0.153
+}

LLM-Detector-44w-EN/checkpoint-10000/README.md

@@ -0,0 +1,207 @@
+---
+library_name: peft
+base_model: ./Qwen-1_8B-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Data Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+
+
+## Training procedure
+
+
+### Framework versions
+
+
+- PEFT 0.6.2

LLM-Detector-44w-EN/checkpoint-10000/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "./Qwen-1_8B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "task_type": "CAUSAL_LM"
+}

LLM-Detector-44w-EN/checkpoint-10000/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:78608cfbb6ea330c0e64fbd48525138e3ac4323e39f7eab19edfdd566ba2962e
+size 6308670

LLM-Detector-44w-EN/checkpoint-10000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:16f2adcf695963a604c175611a7f7a5c592a789764ce264e138f125d610ad9a4
+size 12623610

LLM-Detector-44w-EN/checkpoint-10000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ce73e53cece6c81c88f4ba1fbf614a65143e271e5eb8ef956133078ca0c21a69
+size 14244

LLM-Detector-44w-EN/checkpoint-10000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:837d56630b1e5c783a00e38f82bd5ffd484240d093f73e7220e094b83452858b
+size 1064

LLM-Detector-44w-EN/checkpoint-10000/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "additional_special_tokens": [
+    "<|im_end|>"
+  ],
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

LLM-Detector-44w-EN/checkpoint-10000/tokenization_qwen.py

@@ -0,0 +1,276 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Tokenization classes for QWen."""
+
+import base64
+import logging
+import os
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union
+
+import tiktoken
+from transformers import PreTrainedTokenizer, AddedToken
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken"}
+
+PAT_STR = r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+"""
+ENDOFTEXT = "<|endoftext|>"
+IMSTART = "<|im_start|>"
+IMEND = "<|im_end|>"
+# as the default behavior is changed to allow special tokens in
+# regular texts, the surface forms of special tokens need to be
+# as different as possible to minimize the impact
+EXTRAS = tuple((f"<|extra_{i}|>" for i in range(205)))
+# changed to use actual index to avoid misconfiguration with vocabulary expansion
+SPECIAL_START_ID = 151643
+SPECIAL_TOKENS = tuple(
+    enumerate(
+        (
+            (
+                ENDOFTEXT,
+                IMSTART,
+                IMEND,
+            )
+            + EXTRAS
+        ),
+        start=SPECIAL_START_ID,
+    )
+)
+SPECIAL_TOKENS_SET = set(t for i, t in SPECIAL_TOKENS)
+
+
+def _load_tiktoken_bpe(tiktoken_bpe_file: str) -> Dict[bytes, int]:
+    with open(tiktoken_bpe_file, "rb") as f:
+        contents = f.read()
+    return {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in contents.splitlines() if line)
+    }
+
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        extra_vocab_file=None,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        # how to handle errors in decoding UTF-8 byte sequences
+        # use ignore if you are in streaming inference
+        self.errors = errors  
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: Dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in SPECIAL_TOKENS
+        }
+
+        # try load extra vocab from file
+        if extra_vocab_file is not None:
+            used_ids = set(self.mergeable_ranks.values()) | set(self.special_tokens.values())
+            extra_mergeable_ranks = _load_tiktoken_bpe(extra_vocab_file)
+            for token, index in extra_mergeable_ranks.items():
+                if token in self.mergeable_ranks:
+                    logger.info(f"extra token {token} exists, skipping")
+                    continue
+                if index in used_ids:
+                    logger.info(f'the index {index} for extra token {token} exists, skipping')
+                    continue
+                self.mergeable_ranks[token] = index
+            # the index may be sparse after this, but don't worry tiktoken.Encoding will handle this
+
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        assert (
+            len(self.mergeable_ranks) + len(self.special_tokens) == enc.n_vocab
+        ), f"{len(self.mergeable_ranks) + len(self.special_tokens)} != {enc.n_vocab} in encoding"
+
+        self.decoder = {
+            v: k for k, v in self.mergeable_ranks.items()
+        }  # type: dict[int, bytes|str]
+        self.decoder.update({v: k for k, v in self.special_tokens.items()})
+
+        self.tokenizer = enc  # type: tiktoken.Encoding
+
+        self.eod_id = self.tokenizer.eot_token
+        self.im_start_id = self.special_tokens[IMSTART]
+        self.im_end_id = self.special_tokens[IMEND]
+
+    def __getstate__(self):
+        # for pickle lovers
+        state = self.__dict__.copy()
+        del state["tokenizer"]
+        return state
+
+    def __setstate__(self, state):
+        # tokenizer is not python native; don't pass it; rebuild it
+        self.__dict__.update(state)
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        self.tokenizer = enc
+
+    def __len__(self) -> int:
+        return self.tokenizer.n_vocab
+
+    def get_vocab(self) -> Dict[bytes, int]:
+        return self.mergeable_ranks
+
+    def convert_tokens_to_ids(
+        self, tokens: Union[bytes, str, List[Union[bytes, str]]]
+    ) -> List[int]:
+        ids = []
+        if isinstance(tokens, (str, bytes)):
+            if tokens in self.special_tokens:
+                return self.special_tokens[tokens]
+            else:
+                return self.mergeable_ranks.get(tokens)
+        for token in tokens:
+            if token in self.special_tokens:
+                ids.append(self.special_tokens[token])
+            else:
+                ids.append(self.mergeable_ranks.get(token))
+        return ids
+
+    def _add_tokens(
+        self,
+        new_tokens: Union[List[str], List[AddedToken]],
+        special_tokens: bool = False,
+    ) -> int:
+        if not special_tokens and new_tokens:
+            raise ValueError("Adding regular tokens is not supported")
+        for token in new_tokens:
+            surface_form = token.content if isinstance(token, AddedToken) else token
+            if surface_form not in SPECIAL_TOKENS_SET:
+                raise ValueError("Adding unknown special tokens is not supported")
+        return 0
+
+    def save_vocabulary(self, save_directory: str, **kwargs) -> Tuple[str]:
+        """
+        Save only the vocabulary of the tokenizer (vocabulary).
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        file_path = os.path.join(save_directory, "qwen.tiktoken")
+        with open(file_path, "w", encoding="utf8") as w:
+            for k, v in self.mergeable_ranks.items():
+                line = base64.b64encode(k).decode("utf8") + " " + str(v) + "\n"
+                w.write(line)
+        return (file_path,)
+
+    def tokenize(
+        self,
+        text: str,
+        allowed_special: Union[Set, str] = "all",
+        disallowed_special: Union[Collection, str] = (),
+        **kwargs,
+    ) -> List[Union[bytes, str]]:
+        """
+        Converts a string in a sequence of tokens.
+
+        Args:
+            text (`str`):
+                The sequence to be encoded.
+            allowed_special (`Literal["all"]` or `set`):
+                The surface forms of the tokens to be encoded as special tokens in regular texts.
+                Default to "all".
+            disallowed_special (`Literal["all"]` or `Collection`):
+                The surface forms of the tokens that should not be in regular texts and trigger errors.
+                Default to an empty tuple.
+
+            kwargs (additional keyword arguments, *optional*):
+                Will be passed to the underlying model specific encode method.
+
+        Returns:
+            `List[bytes|str]`: The list of tokens.
+        """
+        tokens = []
+        text = unicodedata.normalize("NFC", text)
+
+        # this implementation takes a detour: text -> token id -> token surface forms
+        for t in self.tokenizer.encode(
+            text, allowed_special=allowed_special, disallowed_special=disallowed_special
+        ):
+            tokens.append(self.decoder[t])
+        return tokens
+
+    def convert_tokens_to_string(self, tokens: List[Union[bytes, str]]) -> str:
+        """
+        Converts a sequence of tokens in a single string.
+        """
+        text = ""
+        temp = b""
+        for t in tokens:
+            if isinstance(t, str):
+                if temp:
+                    text += temp.decode("utf-8", errors=self.errors)
+                    temp = b""
+                text += t
+            elif isinstance(t, bytes):
+                temp += t
+            else:
+                raise TypeError("token should only be of type types or str")
+        if temp:
+            text += temp.decode("utf-8", errors=self.errors)
+        return text
+
+    @property
+    def vocab_size(self):
+        return self.tokenizer.n_vocab
+
+    def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
+        """Converts an id to a token, special tokens included"""
+        if index in self.decoder:
+            return self.decoder[index]
+        raise ValueError("unknown ids")
+
+    def _convert_token_to_id(self, token: Union[bytes, str]) -> int:
+        """Converts a token to an id using the vocab, special tokens included"""
+        if token in self.special_tokens:
+            return self.special_tokens[token]
+        if token in self.mergeable_ranks:
+            return self.mergeable_ranks[token]
+        raise ValueError("unknown token")
+
+    def _tokenize(self, text: str, **kwargs):
+        """
+        Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based
+        vocabulary or sub-words for sub-word-based vocabularies (BPE/SentencePieces/WordPieces).
+
+        Do NOT take care of added tokens.
+        """
+        raise NotImplementedError
+
+    def _decode(
+        self,
+        token_ids: Union[int, List[int]],
+        skip_special_tokens: bool = False,
+        errors: str = None,
+        **kwargs,
+    ) -> str:
+        if isinstance(token_ids, int):
+            token_ids = [token_ids]
+        if skip_special_tokens:
+            token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)

LLM-Detector-44w-EN/checkpoint-10000/tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 8192,
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "QWenTokenizer"
+}

LLM-Detector-44w-EN/checkpoint-10000/trainer_state.json

@@ -0,0 +1,619 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7395629183152757,
+  "eval_steps": 500,
+  "global_step": 10000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999925019651876e-05,
+      "loss": 3.2124,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9997000831051485e-05,
+      "loss": 0.2834,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999325203852471e-05,
+      "loss": 0.2178,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998800404380707e-05,
+      "loss": 0.1782,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9981257161695735e-05,
+      "loss": 0.1501,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997301179689756e-05,
+      "loss": 0.1376,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.996326844400481e-05,
+      "loss": 0.1161,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9952027687465465e-05,
+      "loss": 0.1163,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9939290201548214e-05,
+      "loss": 0.1157,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.992505675030195e-05,
+      "loss": 0.1048,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9909328187509964e-05,
+      "loss": 0.1073,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.989210545663877e-05,
+      "loss": 0.1047,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987338959078144e-05,
+      "loss": 0.1066,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9853181712595686e-05,
+      "loss": 0.1011,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.98314830342365e-05,
+      "loss": 0.0947,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9808294857283454e-05,
+      "loss": 0.0916,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.97836185726626e-05,
+      "loss": 0.0933,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9757455660563085e-05,
+      "loss": 0.089,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9729807690348297e-05,
+      "loss": 0.0872,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9700676320461794e-05,
+      "loss": 0.0898,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9670376756771875e-05,
+      "loss": 0.0894,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.963829870746861e-05,
+      "loss": 0.0859,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.960474274759395e-05,
+      "loss": 0.0826,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.956971088997792e-05,
+      "loss": 0.0816,
+      "step": 2400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.953320523598123e-05,
+      "loss": 0.074,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9495227975369186e-05,
+      "loss": 0.0896,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.945578138618039e-05,
+      "loss": 0.0768,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.941486783459001e-05,
+      "loss": 0.0779,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.937248977476793e-05,
+      "loss": 0.0814,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9328649748731495e-05,
+      "loss": 0.0727,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.928335038619302e-05,
+      "loss": 0.0806,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9236594404402104e-05,
+      "loss": 0.0709,
+      "step": 3200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9188384607982565e-05,
+      "loss": 0.0746,
+      "step": 3300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.913872388876427e-05,
+      "loss": 0.0805,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.908761522560962e-05,
+      "loss": 0.0726,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.903506168423491e-05,
+      "loss": 0.0715,
+      "step": 3600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.898106641702639e-05,
+      "loss": 0.0751,
+      "step": 3700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.892563266285121e-05,
+      "loss": 0.0818,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.886876374686313e-05,
+      "loss": 0.0651,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.881046308030303e-05,
+      "loss": 0.0806,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.87513385076558e-05,
+      "loss": 0.0803,
+      "step": 4100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.869019914568559e-05,
+      "loss": 0.0717,
+      "step": 4200
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8627638744212125e-05,
+      "loss": 0.0678,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8563661055875955e-05,
+      "loss": 0.0681,
+      "step": 4400
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.849826991833256e-05,
+      "loss": 0.0668,
+      "step": 4500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8431469254022146e-05,
+      "loss": 0.0717,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8363263069934364e-05,
+      "loss": 0.0681,
+      "step": 4700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829365545736794e-05,
+      "loss": 0.0703,
+      "step": 4800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8222650591685316e-05,
+      "loss": 0.0625,
+      "step": 4900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.815025273206212e-05,
+      "loss": 0.066,
+      "step": 5000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.807646622123172e-05,
+      "loss": 0.0707,
+      "step": 5100
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.800129548522474e-05,
+      "loss": 0.061,
+      "step": 5200
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7924745033103533e-05,
+      "loss": 0.0645,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.784681945669176e-05,
+      "loss": 0.0669,
+      "step": 5400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.776752343029888e-05,
+      "loss": 0.059,
+      "step": 5500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.768686171043982e-05,
+      "loss": 0.0641,
+      "step": 5600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.760483913554966e-05,
+      "loss": 0.0619,
+      "step": 5700
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7522301106315205e-05,
+      "loss": 0.0616,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7437585147204955e-05,
+      "loss": 0.0625,
+      "step": 5900
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7351523285738384e-05,
+      "loss": 0.0594,
+      "step": 6000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.726412068427418e-05,
+      "loss": 0.0616,
+      "step": 6100
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.717538258559431e-05,
+      "loss": 0.0657,
+      "step": 6200
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7085314312589614e-05,
+      "loss": 0.0631,
+      "step": 6300
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.699392126794045e-05,
+      "loss": 0.0503,
+      "step": 6400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6901208933792675e-05,
+      "loss": 0.0652,
+      "step": 6500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.680718287142875e-05,
+      "loss": 0.0571,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.671184872093419e-05,
+      "loss": 0.054,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.661521220085923e-05,
+      "loss": 0.0628,
+      "step": 6800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.65172791078758e-05,
+      "loss": 0.0554,
+      "step": 6900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.641805531642983e-05,
+      "loss": 0.059,
+      "step": 7000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.631754677838885e-05,
+      "loss": 0.0557,
+      "step": 7100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.621575952268501e-05,
+      "loss": 0.0579,
+      "step": 7200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6112699654953394e-05,
+      "loss": 0.0594,
+      "step": 7300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.600837335716581e-05,
+      "loss": 0.0652,
+      "step": 7400
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.590278688725998e-05,
+      "loss": 0.0509,
+      "step": 7500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.579594657876408e-05,
+      "loss": 0.0658,
+      "step": 7600
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.568785884041696e-05,
+      "loss": 0.054,
+      "step": 7700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.557853015578361e-05,
+      "loss": 0.0543,
+      "step": 7800
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5467967082866306e-05,
+      "loss": 0.0552,
+      "step": 7900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.535730021743883e-05,
+      "loss": 0.0615,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.52443005148096e-05,
+      "loss": 0.0567,
+      "step": 8100
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5130086472420265e-05,
+      "loss": 0.0519,
+      "step": 8200
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5014664941317766e-05,
+      "loss": 0.0573,
+      "step": 8300
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.489804284497936e-05,
+      "loss": 0.053,
+      "step": 8400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.4780227178897366e-05,
+      "loss": 0.0658,
+      "step": 8500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.46612250101595e-05,
+      "loss": 0.0574,
+      "step": 8600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.4541043477025e-05,
+      "loss": 0.0608,
+      "step": 8700
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.441968978849641e-05,
+      "loss": 0.0471,
+      "step": 8800
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.4297171223887194e-05,
+      "loss": 0.0598,
+      "step": 8900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4173495132385035e-05,
+      "loss": 0.0549,
+      "step": 9000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.404866893261106e-05,
+      "loss": 0.0529,
+      "step": 9100
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.39227001121748e-05,
+      "loss": 0.0529,
+      "step": 9200
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.3795596227225066e-05,
+      "loss": 0.0568,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.366736490199668e-05,
+      "loss": 0.0602,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.353801382835318e-05,
+      "loss": 0.053,
+      "step": 9500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.340755076532538e-05,
+      "loss": 0.0498,
+      "step": 9600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3275983538645995e-05,
+      "loss": 0.0509,
+      "step": 9700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.314332004028019e-05,
+      "loss": 0.0552,
+      "step": 9800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.3009568227952194e-05,
+      "loss": 0.0565,
+      "step": 9900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.287473612466796e-05,
+      "loss": 0.0573,
+      "step": 10000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 40563,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 2.2312434217439724e+18,
+  "trial_name": null,
+  "trial_params": null
+}

LLM-Detector-44w-EN/checkpoint-10000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207ad05ea260ffab22b19df12f661a83e61f91ab639a8553b2202d5916732318
+size 4664

LLM-Detector-44w-EN/checkpoint-15000/README.md

@@ -0,0 +1,207 @@
+---
+library_name: peft
+base_model: ./Qwen-1_8B-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Data Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+
+
+## Training procedure
+
+
+### Framework versions
+
+
+- PEFT 0.6.2

LLM-Detector-44w-EN/checkpoint-15000/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "./Qwen-1_8B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "task_type": "CAUSAL_LM"
+}

LLM-Detector-44w-EN/checkpoint-15000/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8e9328bcaf2d8ae8b0b8433b28702a985fa756bfd86d7d7adc25956a25db851c
+size 6308670

LLM-Detector-44w-EN/checkpoint-15000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:664e91079cdddce5ea1a018f569a2ae51b1a35ebed161f89ed193d9c51a21bd7
+size 12623610

LLM-Detector-44w-EN/checkpoint-15000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:81e137f4596972ee86e52aca8e40aa1b0c41f3ecd6da8878e985c8cd4dd6c344
+size 14244

LLM-Detector-44w-EN/checkpoint-15000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b6d27af7c1cb6ccfba45767401ba2d56a3189ead973f5bb014df36bc378d6002
+size 1064

LLM-Detector-44w-EN/checkpoint-15000/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "additional_special_tokens": [
+    "<|im_end|>"
+  ],
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

LLM-Detector-44w-EN/checkpoint-15000/tokenization_qwen.py

@@ -0,0 +1,276 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Tokenization classes for QWen."""
+
+import base64
+import logging
+import os
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union
+
+import tiktoken
+from transformers import PreTrainedTokenizer, AddedToken
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken"}
+
+PAT_STR = r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+"""
+ENDOFTEXT = "<|endoftext|>"
+IMSTART = "<|im_start|>"
+IMEND = "<|im_end|>"
+# as the default behavior is changed to allow special tokens in
+# regular texts, the surface forms of special tokens need to be
+# as different as possible to minimize the impact
+EXTRAS = tuple((f"<|extra_{i}|>" for i in range(205)))
+# changed to use actual index to avoid misconfiguration with vocabulary expansion
+SPECIAL_START_ID = 151643
+SPECIAL_TOKENS = tuple(
+    enumerate(
+        (
+            (
+                ENDOFTEXT,
+                IMSTART,
+                IMEND,
+            )
+            + EXTRAS
+        ),
+        start=SPECIAL_START_ID,
+    )
+)
+SPECIAL_TOKENS_SET = set(t for i, t in SPECIAL_TOKENS)
+
+
+def _load_tiktoken_bpe(tiktoken_bpe_file: str) -> Dict[bytes, int]:
+    with open(tiktoken_bpe_file, "rb") as f:
+        contents = f.read()
+    return {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in contents.splitlines() if line)
+    }
+
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        extra_vocab_file=None,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        # how to handle errors in decoding UTF-8 byte sequences
+        # use ignore if you are in streaming inference
+        self.errors = errors  
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: Dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in SPECIAL_TOKENS
+        }
+
+        # try load extra vocab from file
+        if extra_vocab_file is not None:
+            used_ids = set(self.mergeable_ranks.values()) | set(self.special_tokens.values())
+            extra_mergeable_ranks = _load_tiktoken_bpe(extra_vocab_file)
+            for token, index in extra_mergeable_ranks.items():
+                if token in self.mergeable_ranks:
+                    logger.info(f"extra token {token} exists, skipping")
+                    continue
+                if index in used_ids:
+                    logger.info(f'the index {index} for extra token {token} exists, skipping')
+                    continue
+                self.mergeable_ranks[token] = index
+            # the index may be sparse after this, but don't worry tiktoken.Encoding will handle this
+
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        assert (
+            len(self.mergeable_ranks) + len(self.special_tokens) == enc.n_vocab
+        ), f"{len(self.mergeable_ranks) + len(self.special_tokens)} != {enc.n_vocab} in encoding"
+
+        self.decoder = {
+            v: k for k, v in self.mergeable_ranks.items()
+        }  # type: dict[int, bytes|str]
+        self.decoder.update({v: k for k, v in self.special_tokens.items()})
+
+        self.tokenizer = enc  # type: tiktoken.Encoding
+
+        self.eod_id = self.tokenizer.eot_token
+        self.im_start_id = self.special_tokens[IMSTART]
+        self.im_end_id = self.special_tokens[IMEND]
+
+    def __getstate__(self):
+        # for pickle lovers
+        state = self.__dict__.copy()
+        del state["tokenizer"]
+        return state
+
+    def __setstate__(self, state):
+        # tokenizer is not python native; don't pass it; rebuild it
+        self.__dict__.update(state)
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        self.tokenizer = enc
+
+    def __len__(self) -> int:
+        return self.tokenizer.n_vocab
+
+    def get_vocab(self) -> Dict[bytes, int]:
+        return self.mergeable_ranks
+
+    def convert_tokens_to_ids(
+        self, tokens: Union[bytes, str, List[Union[bytes, str]]]
+    ) -> List[int]:
+        ids = []
+        if isinstance(tokens, (str, bytes)):
+            if tokens in self.special_tokens:
+                return self.special_tokens[tokens]
+            else:
+                return self.mergeable_ranks.get(tokens)
+        for token in tokens:
+            if token in self.special_tokens:
+                ids.append(self.special_tokens[token])
+            else:
+                ids.append(self.mergeable_ranks.get(token))
+        return ids
+
+    def _add_tokens(
+        self,
+        new_tokens: Union[List[str], List[AddedToken]],
+        special_tokens: bool = False,
+    ) -> int:
+        if not special_tokens and new_tokens:
+            raise ValueError("Adding regular tokens is not supported")
+        for token in new_tokens:
+            surface_form = token.content if isinstance(token, AddedToken) else token
+            if surface_form not in SPECIAL_TOKENS_SET:
+                raise ValueError("Adding unknown special tokens is not supported")
+        return 0
+
+    def save_vocabulary(self, save_directory: str, **kwargs) -> Tuple[str]:
+        """
+        Save only the vocabulary of the tokenizer (vocabulary).
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        file_path = os.path.join(save_directory, "qwen.tiktoken")
+        with open(file_path, "w", encoding="utf8") as w:
+            for k, v in self.mergeable_ranks.items():
+                line = base64.b64encode(k).decode("utf8") + " " + str(v) + "\n"
+                w.write(line)
+        return (file_path,)
+
+    def tokenize(
+        self,
+        text: str,
+        allowed_special: Union[Set, str] = "all",
+        disallowed_special: Union[Collection, str] = (),
+        **kwargs,
+    ) -> List[Union[bytes, str]]:
+        """
+        Converts a string in a sequence of tokens.
+
+        Args:
+            text (`str`):
+                The sequence to be encoded.
+            allowed_special (`Literal["all"]` or `set`):
+                The surface forms of the tokens to be encoded as special tokens in regular texts.
+                Default to "all".
+            disallowed_special (`Literal["all"]` or `Collection`):
+                The surface forms of the tokens that should not be in regular texts and trigger errors.
+                Default to an empty tuple.
+
+            kwargs (additional keyword arguments, *optional*):
+                Will be passed to the underlying model specific encode method.
+
+        Returns:
+            `List[bytes|str]`: The list of tokens.
+        """
+        tokens = []
+        text = unicodedata.normalize("NFC", text)
+
+        # this implementation takes a detour: text -> token id -> token surface forms
+        for t in self.tokenizer.encode(
+            text, allowed_special=allowed_special, disallowed_special=disallowed_special
+        ):
+            tokens.append(self.decoder[t])
+        return tokens
+
+    def convert_tokens_to_string(self, tokens: List[Union[bytes, str]]) -> str:
+        """
+        Converts a sequence of tokens in a single string.
+        """
+        text = ""
+        temp = b""
+        for t in tokens:
+            if isinstance(t, str):
+                if temp:
+                    text += temp.decode("utf-8", errors=self.errors)
+                    temp = b""
+                text += t
+            elif isinstance(t, bytes):
+                temp += t
+            else:
+                raise TypeError("token should only be of type types or str")
+        if temp:
+            text += temp.decode("utf-8", errors=self.errors)
+        return text
+
+    @property
+    def vocab_size(self):
+        return self.tokenizer.n_vocab
+
+    def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
+        """Converts an id to a token, special tokens included"""
+        if index in self.decoder:
+            return self.decoder[index]
+        raise ValueError("unknown ids")
+
+    def _convert_token_to_id(self, token: Union[bytes, str]) -> int:
+        """Converts a token to an id using the vocab, special tokens included"""
+        if token in self.special_tokens:
+            return self.special_tokens[token]
+        if token in self.mergeable_ranks:
+            return self.mergeable_ranks[token]
+        raise ValueError("unknown token")
+
+    def _tokenize(self, text: str, **kwargs):
+        """
+        Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based
+        vocabulary or sub-words for sub-word-based vocabularies (BPE/SentencePieces/WordPieces).
+
+        Do NOT take care of added tokens.
+        """
+        raise NotImplementedError
+
+    def _decode(
+        self,
+        token_ids: Union[int, List[int]],
+        skip_special_tokens: bool = False,
+        errors: str = None,
+        **kwargs,
+    ) -> str:
+        if isinstance(token_ids, int):
+            token_ids = [token_ids]
+        if skip_special_tokens:
+            token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)

LLM-Detector-44w-EN/checkpoint-15000/tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 8192,
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "QWenTokenizer"
+}

LLM-Detector-44w-EN/checkpoint-15000/trainer_state.json

@@ -0,0 +1,919 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.1093443774729135,
+  "eval_steps": 500,
+  "global_step": 15000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999925019651876e-05,
+      "loss": 3.2124,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9997000831051485e-05,
+      "loss": 0.2834,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999325203852471e-05,
+      "loss": 0.2178,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998800404380707e-05,
+      "loss": 0.1782,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9981257161695735e-05,
+      "loss": 0.1501,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997301179689756e-05,
+      "loss": 0.1376,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.996326844400481e-05,
+      "loss": 0.1161,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9952027687465465e-05,
+      "loss": 0.1163,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9939290201548214e-05,
+      "loss": 0.1157,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.992505675030195e-05,
+      "loss": 0.1048,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9909328187509964e-05,
+      "loss": 0.1073,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.989210545663877e-05,
+      "loss": 0.1047,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987338959078144e-05,
+      "loss": 0.1066,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9853181712595686e-05,
+      "loss": 0.1011,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.98314830342365e-05,
+      "loss": 0.0947,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9808294857283454e-05,
+      "loss": 0.0916,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.97836185726626e-05,
+      "loss": 0.0933,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9757455660563085e-05,
+      "loss": 0.089,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9729807690348297e-05,
+      "loss": 0.0872,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9700676320461794e-05,
+      "loss": 0.0898,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9670376756771875e-05,
+      "loss": 0.0894,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.963829870746861e-05,
+      "loss": 0.0859,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.960474274759395e-05,
+      "loss": 0.0826,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.956971088997792e-05,
+      "loss": 0.0816,
+      "step": 2400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.953320523598123e-05,
+      "loss": 0.074,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9495227975369186e-05,
+      "loss": 0.0896,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.945578138618039e-05,
+      "loss": 0.0768,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.941486783459001e-05,
+      "loss": 0.0779,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.937248977476793e-05,
+      "loss": 0.0814,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9328649748731495e-05,
+      "loss": 0.0727,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.928335038619302e-05,
+      "loss": 0.0806,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9236594404402104e-05,
+      "loss": 0.0709,
+      "step": 3200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9188384607982565e-05,
+      "loss": 0.0746,
+      "step": 3300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.913872388876427e-05,
+      "loss": 0.0805,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.908761522560962e-05,
+      "loss": 0.0726,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.903506168423491e-05,
+      "loss": 0.0715,
+      "step": 3600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.898106641702639e-05,
+      "loss": 0.0751,
+      "step": 3700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.892563266285121e-05,
+      "loss": 0.0818,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.886876374686313e-05,
+      "loss": 0.0651,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.881046308030303e-05,
+      "loss": 0.0806,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.87513385076558e-05,
+      "loss": 0.0803,
+      "step": 4100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.869019914568559e-05,
+      "loss": 0.0717,
+      "step": 4200
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8627638744212125e-05,
+      "loss": 0.0678,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8563661055875955e-05,
+      "loss": 0.0681,
+      "step": 4400
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.849826991833256e-05,
+      "loss": 0.0668,
+      "step": 4500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8431469254022146e-05,
+      "loss": 0.0717,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8363263069934364e-05,
+      "loss": 0.0681,
+      "step": 4700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829365545736794e-05,
+      "loss": 0.0703,
+      "step": 4800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8222650591685316e-05,
+      "loss": 0.0625,
+      "step": 4900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.815025273206212e-05,
+      "loss": 0.066,
+      "step": 5000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.807646622123172e-05,
+      "loss": 0.0707,
+      "step": 5100
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.800129548522474e-05,
+      "loss": 0.061,
+      "step": 5200
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7924745033103533e-05,
+      "loss": 0.0645,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.784681945669176e-05,
+      "loss": 0.0669,
+      "step": 5400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.776752343029888e-05,
+      "loss": 0.059,
+      "step": 5500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.768686171043982e-05,
+      "loss": 0.0641,
+      "step": 5600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.760483913554966e-05,
+      "loss": 0.0619,
+      "step": 5700
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7522301106315205e-05,
+      "loss": 0.0616,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7437585147204955e-05,
+      "loss": 0.0625,
+      "step": 5900
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7351523285738384e-05,
+      "loss": 0.0594,
+      "step": 6000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.726412068427418e-05,
+      "loss": 0.0616,
+      "step": 6100
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.717538258559431e-05,
+      "loss": 0.0657,
+      "step": 6200
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7085314312589614e-05,
+      "loss": 0.0631,
+      "step": 6300
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.699392126794045e-05,
+      "loss": 0.0503,
+      "step": 6400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6901208933792675e-05,
+      "loss": 0.0652,
+      "step": 6500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.680718287142875e-05,
+      "loss": 0.0571,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.671184872093419e-05,
+      "loss": 0.054,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.661521220085923e-05,
+      "loss": 0.0628,
+      "step": 6800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.65172791078758e-05,
+      "loss": 0.0554,
+      "step": 6900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.641805531642983e-05,
+      "loss": 0.059,
+      "step": 7000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.631754677838885e-05,
+      "loss": 0.0557,
+      "step": 7100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.621575952268501e-05,
+      "loss": 0.0579,
+      "step": 7200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6112699654953394e-05,
+      "loss": 0.0594,
+      "step": 7300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.600837335716581e-05,
+      "loss": 0.0652,
+      "step": 7400
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.590278688725998e-05,
+      "loss": 0.0509,
+      "step": 7500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.579594657876408e-05,
+      "loss": 0.0658,
+      "step": 7600
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.568785884041696e-05,
+      "loss": 0.054,
+      "step": 7700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.557853015578361e-05,
+      "loss": 0.0543,
+      "step": 7800
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5467967082866306e-05,
+      "loss": 0.0552,
+      "step": 7900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.535730021743883e-05,
+      "loss": 0.0615,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.52443005148096e-05,
+      "loss": 0.0567,
+      "step": 8100
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5130086472420265e-05,
+      "loss": 0.0519,
+      "step": 8200
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5014664941317766e-05,
+      "loss": 0.0573,
+      "step": 8300
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.489804284497936e-05,
+      "loss": 0.053,
+      "step": 8400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.4780227178897366e-05,
+      "loss": 0.0658,
+      "step": 8500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.46612250101595e-05,
+      "loss": 0.0574,
+      "step": 8600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.4541043477025e-05,
+      "loss": 0.0608,
+      "step": 8700
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.441968978849641e-05,
+      "loss": 0.0471,
+      "step": 8800
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.4297171223887194e-05,
+      "loss": 0.0598,
+      "step": 8900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4173495132385035e-05,
+      "loss": 0.0549,
+      "step": 9000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.404866893261106e-05,
+      "loss": 0.0529,
+      "step": 9100
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.39227001121748e-05,
+      "loss": 0.0529,
+      "step": 9200
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.3795596227225066e-05,
+      "loss": 0.0568,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.366736490199668e-05,
+      "loss": 0.0602,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.353801382835318e-05,
+      "loss": 0.053,
+      "step": 9500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.340755076532538e-05,
+      "loss": 0.0498,
+      "step": 9600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3275983538645995e-05,
+      "loss": 0.0509,
+      "step": 9700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.314332004028019e-05,
+      "loss": 0.0552,
+      "step": 9800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.3009568227952194e-05,
+      "loss": 0.0565,
+      "step": 9900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.287473612466796e-05,
+      "loss": 0.0573,
+      "step": 10000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.273883181823394e-05,
+      "loss": 0.0477,
+      "step": 10100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2603238384474695e-05,
+      "loss": 0.0522,
+      "step": 10200
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.24652247094062e-05,
+      "loss": 0.0515,
+      "step": 10300
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.23261633954347e-05,
+      "loss": 0.0478,
+      "step": 10400
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.218606278405277e-05,
+      "loss": 0.0535,
+      "step": 10500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.20449312790945e-05,
+      "loss": 0.0481,
+      "step": 10600
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.1902777346231383e-05,
+      "loss": 0.0487,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.175960951246454e-05,
+      "loss": 0.0429,
+      "step": 10800
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.161543636561316e-05,
+      "loss": 0.0485,
+      "step": 10900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.147026655379945e-05,
+      "loss": 0.0486,
+      "step": 11000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.132410878492983e-05,
+      "loss": 0.0484,
+      "step": 11100
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.1176971826172596e-05,
+      "loss": 0.0485,
+      "step": 11200
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.1028864503432085e-05,
+      "loss": 0.0545,
+      "step": 11300
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.087979570081917e-05,
+      "loss": 0.0436,
+      "step": 11400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.072977436011844e-05,
+      "loss": 0.0492,
+      "step": 11500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.057880948025177e-05,
+      "loss": 0.0534,
+      "step": 11600
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.042691011673855e-05,
+      "loss": 0.0551,
+      "step": 11700
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.027408538115252e-05,
+      "loss": 0.0488,
+      "step": 11800
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.0120344440575165e-05,
+      "loss": 0.0525,
+      "step": 11900
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.996569651704589e-05,
+      "loss": 0.0458,
+      "step": 12000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.9810150887008806e-05,
+      "loss": 0.0492,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.965528558751401e-05,
+      "loss": 0.0489,
+      "step": 12200
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.949798133193112e-05,
+      "loss": 0.0465,
+      "step": 12300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.933980742539834e-05,
+      "loss": 0.0502,
+      "step": 12400
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.9180773355863344e-05,
+      "loss": 0.0506,
+      "step": 12500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.902088866287004e-05,
+      "loss": 0.0525,
+      "step": 12600
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.886016293698637e-05,
+      "loss": 0.0472,
+      "step": 12700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.869860581922905e-05,
+      "loss": 0.0443,
+      "step": 12800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.8536227000485234e-05,
+      "loss": 0.042,
+      "step": 12900
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.837303622093119e-05,
+      "loss": 0.0447,
+      "step": 13000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.8209043269448096e-05,
+      "loss": 0.046,
+      "step": 13100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.804425798303483e-05,
+      "loss": 0.0434,
+      "step": 13200
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.787869024621789e-05,
+      "loss": 0.0479,
+      "step": 13300
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.7712349990458524e-05,
+      "loss": 0.0484,
+      "step": 13400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.754524719355695e-05,
+      "loss": 0.0506,
+      "step": 13500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.737739187905389e-05,
+      "loss": 0.0449,
+      "step": 13600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208794115629266e-05,
+      "loss": 0.0419,
+      "step": 13700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.703946401649827e-05,
+      "loss": 0.0356,
+      "step": 13800
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.6869411738804735e-05,
+      "loss": 0.0433,
+      "step": 13900
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.669864748301185e-05,
+      "loss": 0.0408,
+      "step": 14000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.6527181492290277e-05,
+      "loss": 0.0435,
+      "step": 14100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.635502405190375e-05,
+      "loss": 0.0394,
+      "step": 14200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.618391721177532e-05,
+      "loss": 0.0442,
+      "step": 14300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.601041454923619e-05,
+      "loss": 0.0448,
+      "step": 14400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.583625143492436e-05,
+      "loss": 0.047,
+      "step": 14500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.5661438315888565e-05,
+      "loss": 0.0374,
+      "step": 14600
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.5485985678167643e-05,
+      "loss": 0.0397,
+      "step": 14700
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.530990404616147e-05,
+      "loss": 0.0398,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.5133203981999684e-05,
+      "loss": 0.0461,
+      "step": 14900
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.4955896084908166e-05,
+      "loss": 0.0374,
+      "step": 15000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 40563,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 3.3477356574825185e+18,
+  "trial_name": null,
+  "trial_params": null
+}

LLM-Detector-44w-EN/checkpoint-15000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207ad05ea260ffab22b19df12f661a83e61f91ab639a8553b2202d5916732318
+size 4664

LLM-Detector-44w-EN/checkpoint-20000/README.md

@@ -0,0 +1,207 @@
+---
+library_name: peft
+base_model: ./Qwen-1_8B-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Data Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+
+
+## Training procedure
+
+
+### Framework versions
+
+
+- PEFT 0.6.2

LLM-Detector-44w-EN/checkpoint-20000/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "./Qwen-1_8B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "task_type": "CAUSAL_LM"
+}

LLM-Detector-44w-EN/checkpoint-20000/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:54aa4209357e0e092a924b8931d84311ab742706f8d12247fbb01ddd6ad1e191
+size 6308670

LLM-Detector-44w-EN/checkpoint-20000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8376c90a899e06dbba730be00352928f4ab55dcfc80eebcbf30d591d556ee51c
+size 12623610

LLM-Detector-44w-EN/checkpoint-20000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:669139e3d9745b87e9f2b66b1e696b22d8259c50c520380073d1d2e82fe53b8f
+size 14244

LLM-Detector-44w-EN/checkpoint-20000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:592afe39d005c6878933be2e25e00543d1542ce13e6db95ff894694639adfd9b
+size 1064

LLM-Detector-44w-EN/checkpoint-20000/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "additional_special_tokens": [
+    "<|im_end|>"
+  ],
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

LLM-Detector-44w-EN/checkpoint-20000/tokenization_qwen.py

@@ -0,0 +1,276 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Tokenization classes for QWen."""
+
+import base64
+import logging
+import os
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union
+
+import tiktoken
+from transformers import PreTrainedTokenizer, AddedToken
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken"}
+
+PAT_STR = r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+"""
+ENDOFTEXT = "<|endoftext|>"
+IMSTART = "<|im_start|>"
+IMEND = "<|im_end|>"
+# as the default behavior is changed to allow special tokens in
+# regular texts, the surface forms of special tokens need to be
+# as different as possible to minimize the impact
+EXTRAS = tuple((f"<|extra_{i}|>" for i in range(205)))
+# changed to use actual index to avoid misconfiguration with vocabulary expansion
+SPECIAL_START_ID = 151643
+SPECIAL_TOKENS = tuple(
+    enumerate(
+        (
+            (
+                ENDOFTEXT,
+                IMSTART,
+                IMEND,
+            )
+            + EXTRAS
+        ),
+        start=SPECIAL_START_ID,
+    )
+)
+SPECIAL_TOKENS_SET = set(t for i, t in SPECIAL_TOKENS)
+
+
+def _load_tiktoken_bpe(tiktoken_bpe_file: str) -> Dict[bytes, int]:
+    with open(tiktoken_bpe_file, "rb") as f:
+        contents = f.read()
+    return {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in contents.splitlines() if line)
+    }
+
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        extra_vocab_file=None,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        # how to handle errors in decoding UTF-8 byte sequences
+        # use ignore if you are in streaming inference
+        self.errors = errors  
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: Dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in SPECIAL_TOKENS
+        }
+
+        # try load extra vocab from file
+        if extra_vocab_file is not None:
+            used_ids = set(self.mergeable_ranks.values()) | set(self.special_tokens.values())
+            extra_mergeable_ranks = _load_tiktoken_bpe(extra_vocab_file)
+            for token, index in extra_mergeable_ranks.items():
+                if token in self.mergeable_ranks:
+                    logger.info(f"extra token {token} exists, skipping")
+                    continue
+                if index in used_ids:
+                    logger.info(f'the index {index} for extra token {token} exists, skipping')
+                    continue
+                self.mergeable_ranks[token] = index
+            # the index may be sparse after this, but don't worry tiktoken.Encoding will handle this
+
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        assert (
+            len(self.mergeable_ranks) + len(self.special_tokens) == enc.n_vocab
+        ), f"{len(self.mergeable_ranks) + len(self.special_tokens)} != {enc.n_vocab} in encoding"
+
+        self.decoder = {
+            v: k for k, v in self.mergeable_ranks.items()
+        }  # type: dict[int, bytes|str]
+        self.decoder.update({v: k for k, v in self.special_tokens.items()})
+
+        self.tokenizer = enc  # type: tiktoken.Encoding
+
+        self.eod_id = self.tokenizer.eot_token
+        self.im_start_id = self.special_tokens[IMSTART]
+        self.im_end_id = self.special_tokens[IMEND]
+
+    def __getstate__(self):
+        # for pickle lovers
+        state = self.__dict__.copy()
+        del state["tokenizer"]
+        return state
+
+    def __setstate__(self, state):
+        # tokenizer is not python native; don't pass it; rebuild it
+        self.__dict__.update(state)
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        self.tokenizer = enc
+
+    def __len__(self) -> int:
+        return self.tokenizer.n_vocab
+
+    def get_vocab(self) -> Dict[bytes, int]:
+        return self.mergeable_ranks
+
+    def convert_tokens_to_ids(
+        self, tokens: Union[bytes, str, List[Union[bytes, str]]]
+    ) -> List[int]:
+        ids = []
+        if isinstance(tokens, (str, bytes)):
+            if tokens in self.special_tokens:
+                return self.special_tokens[tokens]
+            else:
+                return self.mergeable_ranks.get(tokens)
+        for token in tokens:
+            if token in self.special_tokens:
+                ids.append(self.special_tokens[token])
+            else:
+                ids.append(self.mergeable_ranks.get(token))
+        return ids
+
+    def _add_tokens(
+        self,
+        new_tokens: Union[List[str], List[AddedToken]],
+        special_tokens: bool = False,
+    ) -> int:
+        if not special_tokens and new_tokens:
+            raise ValueError("Adding regular tokens is not supported")
+        for token in new_tokens:
+            surface_form = token.content if isinstance(token, AddedToken) else token
+            if surface_form not in SPECIAL_TOKENS_SET:
+                raise ValueError("Adding unknown special tokens is not supported")
+        return 0
+
+    def save_vocabulary(self, save_directory: str, **kwargs) -> Tuple[str]:
+        """
+        Save only the vocabulary of the tokenizer (vocabulary).
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        file_path = os.path.join(save_directory, "qwen.tiktoken")
+        with open(file_path, "w", encoding="utf8") as w:
+            for k, v in self.mergeable_ranks.items():
+                line = base64.b64encode(k).decode("utf8") + " " + str(v) + "\n"
+                w.write(line)
+        return (file_path,)
+
+    def tokenize(
+        self,
+        text: str,
+        allowed_special: Union[Set, str] = "all",
+        disallowed_special: Union[Collection, str] = (),
+        **kwargs,
+    ) -> List[Union[bytes, str]]:
+        """
+        Converts a string in a sequence of tokens.
+
+        Args:
+            text (`str`):
+                The sequence to be encoded.
+            allowed_special (`Literal["all"]` or `set`):
+                The surface forms of the tokens to be encoded as special tokens in regular texts.
+                Default to "all".
+            disallowed_special (`Literal["all"]` or `Collection`):
+                The surface forms of the tokens that should not be in regular texts and trigger errors.
+                Default to an empty tuple.
+
+            kwargs (additional keyword arguments, *optional*):
+                Will be passed to the underlying model specific encode method.
+
+        Returns:
+            `List[bytes|str]`: The list of tokens.
+        """
+        tokens = []
+        text = unicodedata.normalize("NFC", text)
+
+        # this implementation takes a detour: text -> token id -> token surface forms
+        for t in self.tokenizer.encode(
+            text, allowed_special=allowed_special, disallowed_special=disallowed_special
+        ):
+            tokens.append(self.decoder[t])
+        return tokens
+
+    def convert_tokens_to_string(self, tokens: List[Union[bytes, str]]) -> str:
+        """
+        Converts a sequence of tokens in a single string.
+        """
+        text = ""
+        temp = b""
+        for t in tokens:
+            if isinstance(t, str):
+                if temp:
+                    text += temp.decode("utf-8", errors=self.errors)
+                    temp = b""
+                text += t
+            elif isinstance(t, bytes):
+                temp += t
+            else:
+                raise TypeError("token should only be of type types or str")
+        if temp:
+            text += temp.decode("utf-8", errors=self.errors)
+        return text
+
+    @property
+    def vocab_size(self):
+        return self.tokenizer.n_vocab
+
+    def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
+        """Converts an id to a token, special tokens included"""
+        if index in self.decoder:
+            return self.decoder[index]
+        raise ValueError("unknown ids")
+
+    def _convert_token_to_id(self, token: Union[bytes, str]) -> int:
+        """Converts a token to an id using the vocab, special tokens included"""
+        if token in self.special_tokens:
+            return self.special_tokens[token]
+        if token in self.mergeable_ranks:
+            return self.mergeable_ranks[token]
+        raise ValueError("unknown token")
+
+    def _tokenize(self, text: str, **kwargs):
+        """
+        Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based
+        vocabulary or sub-words for sub-word-based vocabularies (BPE/SentencePieces/WordPieces).
+
+        Do NOT take care of added tokens.
+        """
+        raise NotImplementedError
+
+    def _decode(
+        self,
+        token_ids: Union[int, List[int]],
+        skip_special_tokens: bool = False,
+        errors: str = None,
+        **kwargs,
+    ) -> str:
+        if isinstance(token_ids, int):
+            token_ids = [token_ids]
+        if skip_special_tokens:
+            token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)

LLM-Detector-44w-EN/checkpoint-20000/tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 8192,
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "QWenTokenizer"
+}

LLM-Detector-44w-EN/checkpoint-20000/trainer_state.json

@@ -0,0 +1,1219 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.4791258366305513,
+  "eval_steps": 500,
+  "global_step": 20000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999925019651876e-05,
+      "loss": 3.2124,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9997000831051485e-05,
+      "loss": 0.2834,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999325203852471e-05,
+      "loss": 0.2178,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998800404380707e-05,
+      "loss": 0.1782,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9981257161695735e-05,
+      "loss": 0.1501,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997301179689756e-05,
+      "loss": 0.1376,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.996326844400481e-05,
+      "loss": 0.1161,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9952027687465465e-05,
+      "loss": 0.1163,
+      "step": 800
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9939290201548214e-05,
+      "loss": 0.1157,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.992505675030195e-05,
+      "loss": 0.1048,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9909328187509964e-05,
+      "loss": 0.1073,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.989210545663877e-05,
+      "loss": 0.1047,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.987338959078144e-05,
+      "loss": 0.1066,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9853181712595686e-05,
+      "loss": 0.1011,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.98314830342365e-05,
+      "loss": 0.0947,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9808294857283454e-05,
+      "loss": 0.0916,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.97836185726626e-05,
+      "loss": 0.0933,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9757455660563085e-05,
+      "loss": 0.089,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9729807690348297e-05,
+      "loss": 0.0872,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9700676320461794e-05,
+      "loss": 0.0898,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9670376756771875e-05,
+      "loss": 0.0894,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.963829870746861e-05,
+      "loss": 0.0859,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.960474274759395e-05,
+      "loss": 0.0826,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.956971088997792e-05,
+      "loss": 0.0816,
+      "step": 2400
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.953320523598123e-05,
+      "loss": 0.074,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9495227975369186e-05,
+      "loss": 0.0896,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.945578138618039e-05,
+      "loss": 0.0768,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.941486783459001e-05,
+      "loss": 0.0779,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.937248977476793e-05,
+      "loss": 0.0814,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.9328649748731495e-05,
+      "loss": 0.0727,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.928335038619302e-05,
+      "loss": 0.0806,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9236594404402104e-05,
+      "loss": 0.0709,
+      "step": 3200
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.9188384607982565e-05,
+      "loss": 0.0746,
+      "step": 3300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.913872388876427e-05,
+      "loss": 0.0805,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.908761522560962e-05,
+      "loss": 0.0726,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.903506168423491e-05,
+      "loss": 0.0715,
+      "step": 3600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.898106641702639e-05,
+      "loss": 0.0751,
+      "step": 3700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.892563266285121e-05,
+      "loss": 0.0818,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.886876374686313e-05,
+      "loss": 0.0651,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.881046308030303e-05,
+      "loss": 0.0806,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.87513385076558e-05,
+      "loss": 0.0803,
+      "step": 4100
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.869019914568559e-05,
+      "loss": 0.0717,
+      "step": 4200
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.8627638744212125e-05,
+      "loss": 0.0678,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.8563661055875955e-05,
+      "loss": 0.0681,
+      "step": 4400
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.849826991833256e-05,
+      "loss": 0.0668,
+      "step": 4500
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.8431469254022146e-05,
+      "loss": 0.0717,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.8363263069934364e-05,
+      "loss": 0.0681,
+      "step": 4700
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.829365545736794e-05,
+      "loss": 0.0703,
+      "step": 4800
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.8222650591685316e-05,
+      "loss": 0.0625,
+      "step": 4900
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.815025273206212e-05,
+      "loss": 0.066,
+      "step": 5000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.807646622123172e-05,
+      "loss": 0.0707,
+      "step": 5100
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.800129548522474e-05,
+      "loss": 0.061,
+      "step": 5200
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.7924745033103533e-05,
+      "loss": 0.0645,
+      "step": 5300
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.784681945669176e-05,
+      "loss": 0.0669,
+      "step": 5400
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.776752343029888e-05,
+      "loss": 0.059,
+      "step": 5500
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.768686171043982e-05,
+      "loss": 0.0641,
+      "step": 5600
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.760483913554966e-05,
+      "loss": 0.0619,
+      "step": 5700
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.7522301106315205e-05,
+      "loss": 0.0616,
+      "step": 5800
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7437585147204955e-05,
+      "loss": 0.0625,
+      "step": 5900
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.7351523285738384e-05,
+      "loss": 0.0594,
+      "step": 6000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.726412068427418e-05,
+      "loss": 0.0616,
+      "step": 6100
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.717538258559431e-05,
+      "loss": 0.0657,
+      "step": 6200
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.7085314312589614e-05,
+      "loss": 0.0631,
+      "step": 6300
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.699392126794045e-05,
+      "loss": 0.0503,
+      "step": 6400
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.6901208933792675e-05,
+      "loss": 0.0652,
+      "step": 6500
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.680718287142875e-05,
+      "loss": 0.0571,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.671184872093419e-05,
+      "loss": 0.054,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.661521220085923e-05,
+      "loss": 0.0628,
+      "step": 6800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 4.65172791078758e-05,
+      "loss": 0.0554,
+      "step": 6900
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 4.641805531642983e-05,
+      "loss": 0.059,
+      "step": 7000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.631754677838885e-05,
+      "loss": 0.0557,
+      "step": 7100
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 4.621575952268501e-05,
+      "loss": 0.0579,
+      "step": 7200
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 4.6112699654953394e-05,
+      "loss": 0.0594,
+      "step": 7300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.600837335716581e-05,
+      "loss": 0.0652,
+      "step": 7400
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 4.590278688725998e-05,
+      "loss": 0.0509,
+      "step": 7500
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 4.579594657876408e-05,
+      "loss": 0.0658,
+      "step": 7600
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 4.568785884041696e-05,
+      "loss": 0.054,
+      "step": 7700
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.557853015578361e-05,
+      "loss": 0.0543,
+      "step": 7800
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 4.5467967082866306e-05,
+      "loss": 0.0552,
+      "step": 7900
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 4.535730021743883e-05,
+      "loss": 0.0615,
+      "step": 8000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 4.52443005148096e-05,
+      "loss": 0.0567,
+      "step": 8100
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5130086472420265e-05,
+      "loss": 0.0519,
+      "step": 8200
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 4.5014664941317766e-05,
+      "loss": 0.0573,
+      "step": 8300
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 4.489804284497936e-05,
+      "loss": 0.053,
+      "step": 8400
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 4.4780227178897366e-05,
+      "loss": 0.0658,
+      "step": 8500
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.46612250101595e-05,
+      "loss": 0.0574,
+      "step": 8600
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 4.4541043477025e-05,
+      "loss": 0.0608,
+      "step": 8700
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 4.441968978849641e-05,
+      "loss": 0.0471,
+      "step": 8800
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 4.4297171223887194e-05,
+      "loss": 0.0598,
+      "step": 8900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.4173495132385035e-05,
+      "loss": 0.0549,
+      "step": 9000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 4.404866893261106e-05,
+      "loss": 0.0529,
+      "step": 9100
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.39227001121748e-05,
+      "loss": 0.0529,
+      "step": 9200
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.3795596227225066e-05,
+      "loss": 0.0568,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.366736490199668e-05,
+      "loss": 0.0602,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.353801382835318e-05,
+      "loss": 0.053,
+      "step": 9500
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.340755076532538e-05,
+      "loss": 0.0498,
+      "step": 9600
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.3275983538645995e-05,
+      "loss": 0.0509,
+      "step": 9700
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 4.314332004028019e-05,
+      "loss": 0.0552,
+      "step": 9800
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 4.3009568227952194e-05,
+      "loss": 0.0565,
+      "step": 9900
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 4.287473612466796e-05,
+      "loss": 0.0573,
+      "step": 10000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.273883181823394e-05,
+      "loss": 0.0477,
+      "step": 10100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 4.2603238384474695e-05,
+      "loss": 0.0522,
+      "step": 10200
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 4.24652247094062e-05,
+      "loss": 0.0515,
+      "step": 10300
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 4.23261633954347e-05,
+      "loss": 0.0478,
+      "step": 10400
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.218606278405277e-05,
+      "loss": 0.0535,
+      "step": 10500
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.20449312790945e-05,
+      "loss": 0.0481,
+      "step": 10600
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.1902777346231383e-05,
+      "loss": 0.0487,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.175960951246454e-05,
+      "loss": 0.0429,
+      "step": 10800
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.161543636561316e-05,
+      "loss": 0.0485,
+      "step": 10900
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 4.147026655379945e-05,
+      "loss": 0.0486,
+      "step": 11000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 4.132410878492983e-05,
+      "loss": 0.0484,
+      "step": 11100
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 4.1176971826172596e-05,
+      "loss": 0.0485,
+      "step": 11200
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.1028864503432085e-05,
+      "loss": 0.0545,
+      "step": 11300
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 4.087979570081917e-05,
+      "loss": 0.0436,
+      "step": 11400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 4.072977436011844e-05,
+      "loss": 0.0492,
+      "step": 11500
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 4.057880948025177e-05,
+      "loss": 0.0534,
+      "step": 11600
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.042691011673855e-05,
+      "loss": 0.0551,
+      "step": 11700
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 4.027408538115252e-05,
+      "loss": 0.0488,
+      "step": 11800
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 4.0120344440575165e-05,
+      "loss": 0.0525,
+      "step": 11900
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.996569651704589e-05,
+      "loss": 0.0458,
+      "step": 12000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 3.9810150887008806e-05,
+      "loss": 0.0492,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 3.965528558751401e-05,
+      "loss": 0.0489,
+      "step": 12200
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.949798133193112e-05,
+      "loss": 0.0465,
+      "step": 12300
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.933980742539834e-05,
+      "loss": 0.0502,
+      "step": 12400
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.9180773355863344e-05,
+      "loss": 0.0506,
+      "step": 12500
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 3.902088866287004e-05,
+      "loss": 0.0525,
+      "step": 12600
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.886016293698637e-05,
+      "loss": 0.0472,
+      "step": 12700
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.869860581922905e-05,
+      "loss": 0.0443,
+      "step": 12800
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.8536227000485234e-05,
+      "loss": 0.042,
+      "step": 12900
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 3.837303622093119e-05,
+      "loss": 0.0447,
+      "step": 13000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.8209043269448096e-05,
+      "loss": 0.046,
+      "step": 13100
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.804425798303483e-05,
+      "loss": 0.0434,
+      "step": 13200
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.787869024621789e-05,
+      "loss": 0.0479,
+      "step": 13300
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.7712349990458524e-05,
+      "loss": 0.0484,
+      "step": 13400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.754524719355695e-05,
+      "loss": 0.0506,
+      "step": 13500
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.737739187905389e-05,
+      "loss": 0.0449,
+      "step": 13600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7208794115629266e-05,
+      "loss": 0.0419,
+      "step": 13700
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.703946401649827e-05,
+      "loss": 0.0356,
+      "step": 13800
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 3.6869411738804735e-05,
+      "loss": 0.0433,
+      "step": 13900
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.669864748301185e-05,
+      "loss": 0.0408,
+      "step": 14000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 3.6527181492290277e-05,
+      "loss": 0.0435,
+      "step": 14100
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 3.635502405190375e-05,
+      "loss": 0.0394,
+      "step": 14200
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.618391721177532e-05,
+      "loss": 0.0442,
+      "step": 14300
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 3.601041454923619e-05,
+      "loss": 0.0448,
+      "step": 14400
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 3.583625143492436e-05,
+      "loss": 0.047,
+      "step": 14500
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 3.5661438315888565e-05,
+      "loss": 0.0374,
+      "step": 14600
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.5485985678167643e-05,
+      "loss": 0.0397,
+      "step": 14700
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 3.530990404616147e-05,
+      "loss": 0.0398,
+      "step": 14800
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 3.5133203981999684e-05,
+      "loss": 0.0461,
+      "step": 14900
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 3.4955896084908166e-05,
+      "loss": 0.0374,
+      "step": 15000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.4777990990573174e-05,
+      "loss": 0.0415,
+      "step": 15100
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 3.459949937050345e-05,
+      "loss": 0.0378,
+      "step": 15200
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 3.442043193139005e-05,
+      "loss": 0.0434,
+      "step": 15300
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 3.424079941446407e-05,
+      "loss": 0.0409,
+      "step": 15400
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.4060612594852486e-05,
+      "loss": 0.0384,
+      "step": 15500
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 3.387988228093163e-05,
+      "loss": 0.0394,
+      "step": 15600
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 3.3698619313679e-05,
+      "loss": 0.0343,
+      "step": 15700
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 3.3516834566022906e-05,
+      "loss": 0.0371,
+      "step": 15800
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.333453894219027e-05,
+      "loss": 0.0364,
+      "step": 15900
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 3.315174337705257e-05,
+      "loss": 0.04,
+      "step": 16000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 3.296845883546988e-05,
+      "loss": 0.0422,
+      "step": 16100
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 3.2784696311633213e-05,
+      "loss": 0.0343,
+      "step": 16200
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.260046682840495e-05,
+      "loss": 0.0381,
+      "step": 16300
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 3.241578143665773e-05,
+      "loss": 0.041,
+      "step": 16400
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 3.223250468230082e-05,
+      "loss": 0.0336,
+      "step": 16500
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 3.204694501705715e-05,
+      "loss": 0.0364,
+      "step": 16600
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.186096264590101e-05,
+      "loss": 0.0328,
+      "step": 16700
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 3.1674568724850744e-05,
+      "loss": 0.0403,
+      "step": 16800
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 3.148777443461123e-05,
+      "loss": 0.0425,
+      "step": 16900
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.13005909799032e-05,
+      "loss": 0.0432,
+      "step": 17000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 3.111302958879111e-05,
+      "loss": 0.0423,
+      "step": 17100
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 3.0925101512009715e-05,
+      "loss": 0.0328,
+      "step": 17200
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 3.073681802228907e-05,
+      "loss": 0.0385,
+      "step": 17300
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.054819041367849e-05,
+      "loss": 0.0378,
+      "step": 17400
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 3.035923000086896e-05,
+      "loss": 0.0461,
+      "step": 17500
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 3.016994811851453e-05,
+      "loss": 0.0332,
+      "step": 17600
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.9980356120552333e-05,
+      "loss": 0.0343,
+      "step": 17700
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.9790465379521572e-05,
+      "loss": 0.0413,
+      "step": 17800
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.960028728588135e-05,
+      "loss": 0.0363,
+      "step": 17900
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.94098332473274e-05,
+      "loss": 0.0486,
+      "step": 18000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.9219114688107802e-05,
+      "loss": 0.0388,
+      "step": 18100
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.9028143048337736e-05,
+      "loss": 0.0389,
+      "step": 18200
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8836929783313228e-05,
+      "loss": 0.0424,
+      "step": 18300
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8645486362824016e-05,
+      "loss": 0.04,
+      "step": 18400
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.845574193607597e-05,
+      "loss": 0.0414,
+      "step": 18500
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.826387468336204e-05,
+      "loss": 0.0421,
+      "step": 18600
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.807181164948013e-05,
+      "loss": 0.0387,
+      "step": 18700
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.7879564355192734e-05,
+      "loss": 0.0342,
+      "step": 18800
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.7687144332315106e-05,
+      "loss": 0.0377,
+      "step": 18900
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.749456312302348e-05,
+      "loss": 0.0394,
+      "step": 19000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.7301832279162748e-05,
+      "loss": 0.0343,
+      "step": 19100
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.7108963361553523e-05,
+      "loss": 0.0383,
+      "step": 19200
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.6915967939298682e-05,
+      "loss": 0.0399,
+      "step": 19300
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.672285758908937e-05,
+      "loss": 0.0394,
+      "step": 19400
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.6529643894510626e-05,
+      "loss": 0.0379,
+      "step": 19500
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.63363384453465e-05,
+      "loss": 0.0353,
+      "step": 19600
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.6142952836884905e-05,
+      "loss": 0.0335,
+      "step": 19700
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.5949498669222026e-05,
+      "loss": 0.0324,
+      "step": 19800
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.5755987546566546e-05,
+      "loss": 0.0392,
+      "step": 19900
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.556243107654353e-05,
+      "loss": 0.0391,
+      "step": 20000
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 40563,
+  "num_train_epochs": 3,
+  "save_steps": 5000,
+  "total_flos": 4.463551046392185e+18,
+  "trial_name": null,
+  "trial_params": null
+}

LLM-Detector-44w-EN/checkpoint-20000/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:207ad05ea260ffab22b19df12f661a83e61f91ab639a8553b2202d5916732318
+size 4664

LLM-Detector-44w-EN/checkpoint-25000/README.md

@@ -0,0 +1,207 @@
+---
+library_name: peft
+base_model: ./Qwen-1_8B-Chat
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure 
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Data Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+
+
+## Training procedure
+
+
+### Framework versions
+
+
+- PEFT 0.6.2

LLM-Detector-44w-EN/checkpoint-25000/adapter_config.json

@@ -0,0 +1,22 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "./Qwen-1_8B-Chat",
+  "bias": "none",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "lora_alpha": 16.0,
+  "lora_dropout": 0.1,
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 8,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "c_attn"
+  ],
+  "task_type": "CAUSAL_LM"
+}

LLM-Detector-44w-EN/checkpoint-25000/adapter_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:68862f8ff515896d80d1af8824d3960f7ce9c83c8883bcb4d5bd1fdce3e4ff11
+size 6308670

LLM-Detector-44w-EN/checkpoint-25000/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14f35a03839f958014b4eedf5aaf79a90a672e1dbc7bf89f393c9eaee84e0e32
+size 12623610

LLM-Detector-44w-EN/checkpoint-25000/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ade8524bbf39cb986fd50ff46e11c127e7e3e0a36ec22207bec345dbe6d13d1d
+size 14244

LLM-Detector-44w-EN/checkpoint-25000/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a239119ccce7ad2c75d7d2ebb5cdd895972f6f31486f6c4df47965339676a56c
+size 1064

LLM-Detector-44w-EN/checkpoint-25000/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "additional_special_tokens": [
+    "<|im_end|>"
+  ],
+  "eos_token": "<|endoftext|>",
+  "pad_token": "<|endoftext|>"
+}

LLM-Detector-44w-EN/checkpoint-25000/tokenization_qwen.py

@@ -0,0 +1,276 @@
+# Copyright (c) Alibaba Cloud.
+#
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+"""Tokenization classes for QWen."""
+
+import base64
+import logging
+import os
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union
+
+import tiktoken
+from transformers import PreTrainedTokenizer, AddedToken
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken"}
+
+PAT_STR = r"""(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\r\n\p{L}\p{N}]?\p{L}+|\p{N}| ?[^\s\p{L}\p{N}]+[\r\n]*|\s*[\r\n]+|\s+(?!\S)|\s+"""
+ENDOFTEXT = "<|endoftext|>"
+IMSTART = "<|im_start|>"
+IMEND = "<|im_end|>"
+# as the default behavior is changed to allow special tokens in
+# regular texts, the surface forms of special tokens need to be
+# as different as possible to minimize the impact
+EXTRAS = tuple((f"<|extra_{i}|>" for i in range(205)))
+# changed to use actual index to avoid misconfiguration with vocabulary expansion
+SPECIAL_START_ID = 151643
+SPECIAL_TOKENS = tuple(
+    enumerate(
+        (
+            (
+                ENDOFTEXT,
+                IMSTART,
+                IMEND,
+            )
+            + EXTRAS
+        ),
+        start=SPECIAL_START_ID,
+    )
+)
+SPECIAL_TOKENS_SET = set(t for i, t in SPECIAL_TOKENS)
+
+
+def _load_tiktoken_bpe(tiktoken_bpe_file: str) -> Dict[bytes, int]:
+    with open(tiktoken_bpe_file, "rb") as f:
+        contents = f.read()
+    return {
+        base64.b64decode(token): int(rank)
+        for token, rank in (line.split() for line in contents.splitlines() if line)
+    }
+
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        extra_vocab_file=None,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+
+        # how to handle errors in decoding UTF-8 byte sequences
+        # use ignore if you are in streaming inference
+        self.errors = errors  
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: Dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in SPECIAL_TOKENS
+        }
+
+        # try load extra vocab from file
+        if extra_vocab_file is not None:
+            used_ids = set(self.mergeable_ranks.values()) | set(self.special_tokens.values())
+            extra_mergeable_ranks = _load_tiktoken_bpe(extra_vocab_file)
+            for token, index in extra_mergeable_ranks.items():
+                if token in self.mergeable_ranks:
+                    logger.info(f"extra token {token} exists, skipping")
+                    continue
+                if index in used_ids:
+                    logger.info(f'the index {index} for extra token {token} exists, skipping')
+                    continue
+                self.mergeable_ranks[token] = index
+            # the index may be sparse after this, but don't worry tiktoken.Encoding will handle this
+
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        assert (
+            len(self.mergeable_ranks) + len(self.special_tokens) == enc.n_vocab
+        ), f"{len(self.mergeable_ranks) + len(self.special_tokens)} != {enc.n_vocab} in encoding"
+
+        self.decoder = {
+            v: k for k, v in self.mergeable_ranks.items()
+        }  # type: dict[int, bytes|str]
+        self.decoder.update({v: k for k, v in self.special_tokens.items()})
+
+        self.tokenizer = enc  # type: tiktoken.Encoding
+
+        self.eod_id = self.tokenizer.eot_token
+        self.im_start_id = self.special_tokens[IMSTART]
+        self.im_end_id = self.special_tokens[IMEND]
+
+    def __getstate__(self):
+        # for pickle lovers
+        state = self.__dict__.copy()
+        del state["tokenizer"]
+        return state
+
+    def __setstate__(self, state):
+        # tokenizer is not python native; don't pass it; rebuild it
+        self.__dict__.update(state)
+        enc = tiktoken.Encoding(
+            "Qwen",
+            pat_str=PAT_STR,
+            mergeable_ranks=self.mergeable_ranks,
+            special_tokens=self.special_tokens,
+        )
+        self.tokenizer = enc
+
+    def __len__(self) -> int:
+        return self.tokenizer.n_vocab
+
+    def get_vocab(self) -> Dict[bytes, int]:
+        return self.mergeable_ranks
+
+    def convert_tokens_to_ids(
+        self, tokens: Union[bytes, str, List[Union[bytes, str]]]
+    ) -> List[int]:
+        ids = []
+        if isinstance(tokens, (str, bytes)):
+            if tokens in self.special_tokens:
+                return self.special_tokens[tokens]
+            else:
+                return self.mergeable_ranks.get(tokens)
+        for token in tokens:
+            if token in self.special_tokens:
+                ids.append(self.special_tokens[token])
+            else:
+                ids.append(self.mergeable_ranks.get(token))
+        return ids
+
+    def _add_tokens(
+        self,
+        new_tokens: Union[List[str], List[AddedToken]],
+        special_tokens: bool = False,
+    ) -> int:
+        if not special_tokens and new_tokens:
+            raise ValueError("Adding regular tokens is not supported")
+        for token in new_tokens:
+            surface_form = token.content if isinstance(token, AddedToken) else token
+            if surface_form not in SPECIAL_TOKENS_SET:
+                raise ValueError("Adding unknown special tokens is not supported")
+        return 0
+
+    def save_vocabulary(self, save_directory: str, **kwargs) -> Tuple[str]:
+        """
+        Save only the vocabulary of the tokenizer (vocabulary).
+
+        Returns:
+            `Tuple(str)`: Paths to the files saved.
+        """
+        file_path = os.path.join(save_directory, "qwen.tiktoken")
+        with open(file_path, "w", encoding="utf8") as w:
+            for k, v in self.mergeable_ranks.items():
+                line = base64.b64encode(k).decode("utf8") + " " + str(v) + "\n"
+                w.write(line)
+        return (file_path,)
+
+    def tokenize(
+        self,
+        text: str,
+        allowed_special: Union[Set, str] = "all",
+        disallowed_special: Union[Collection, str] = (),
+        **kwargs,
+    ) -> List[Union[bytes, str]]:
+        """
+        Converts a string in a sequence of tokens.
+
+        Args:
+            text (`str`):
+                The sequence to be encoded.
+            allowed_special (`Literal["all"]` or `set`):
+                The surface forms of the tokens to be encoded as special tokens in regular texts.
+                Default to "all".
+            disallowed_special (`Literal["all"]` or `Collection`):
+                The surface forms of the tokens that should not be in regular texts and trigger errors.
+                Default to an empty tuple.
+
+            kwargs (additional keyword arguments, *optional*):
+                Will be passed to the underlying model specific encode method.
+
+        Returns:
+            `List[bytes|str]`: The list of tokens.
+        """
+        tokens = []
+        text = unicodedata.normalize("NFC", text)
+
+        # this implementation takes a detour: text -> token id -> token surface forms
+        for t in self.tokenizer.encode(
+            text, allowed_special=allowed_special, disallowed_special=disallowed_special
+        ):
+            tokens.append(self.decoder[t])
+        return tokens
+
+    def convert_tokens_to_string(self, tokens: List[Union[bytes, str]]) -> str:
+        """
+        Converts a sequence of tokens in a single string.
+        """
+        text = ""
+        temp = b""
+        for t in tokens:
+            if isinstance(t, str):
+                if temp:
+                    text += temp.decode("utf-8", errors=self.errors)
+                    temp = b""
+                text += t
+            elif isinstance(t, bytes):
+                temp += t
+            else:
+                raise TypeError("token should only be of type types or str")
+        if temp:
+            text += temp.decode("utf-8", errors=self.errors)
+        return text
+
+    @property
+    def vocab_size(self):
+        return self.tokenizer.n_vocab
+
+    def _convert_id_to_token(self, index: int) -> Union[bytes, str]:
+        """Converts an id to a token, special tokens included"""
+        if index in self.decoder:
+            return self.decoder[index]
+        raise ValueError("unknown ids")
+
+    def _convert_token_to_id(self, token: Union[bytes, str]) -> int:
+        """Converts a token to an id using the vocab, special tokens included"""
+        if token in self.special_tokens:
+            return self.special_tokens[token]
+        if token in self.mergeable_ranks:
+            return self.mergeable_ranks[token]
+        raise ValueError("unknown token")
+
+    def _tokenize(self, text: str, **kwargs):
+        """
+        Converts a string in a sequence of tokens (string), using the tokenizer. Split in words for word-based
+        vocabulary or sub-words for sub-word-based vocabularies (BPE/SentencePieces/WordPieces).
+
+        Do NOT take care of added tokens.
+        """
+        raise NotImplementedError
+
+    def _decode(
+        self,
+        token_ids: Union[int, List[int]],
+        skip_special_tokens: bool = False,
+        errors: str = None,
+        **kwargs,
+    ) -> str:
+        if isinstance(token_ids, int):
+            token_ids = [token_ids]
+        if skip_special_tokens:
+            token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)

LLM-Detector-44w-EN/checkpoint-25000/tokenizer_config.json

@@ -0,0 +1,13 @@
+{
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 8192,
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "QWenTokenizer"
+}