commit from

README.md

@@ -0,0 +1,202 @@
+---
+library_name: peft
+base_model: /home/wentao/models--Qwen--Qwen-VL-Chat/snapshots/f57cfbd358cb56b710d963669ad1bcfb44cdcdd8
+---
+
+# 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]
+- **Funded by [optional]:** [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 Dataset 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 Dataset 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]
+### Framework versions
+
+- PEFT 0.10.0

adapter_config.json

@@ -0,0 +1,30 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path":  "F:/kelperliu/docker_data/models--Qwen--Qwen-VL-Chat/snapshots/f57cfbd358cb56b710d963669ad1bcfb44cdcdd8",
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 64,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "w1",
+    "c_attn",
+    "attn.c_proj",
+    "w2"
+  ],
+  "task_type": "CAUSAL_LM",
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6639723fd37096e64df5cb6528157004ccad0b3704e5573bf74e4d9528d459bc
+size 224428728

special_tokens_map.json

@@ -0,0 +1,3 @@
+{
+  "pad_token": "<|endoftext|>"
+}

tokenization_qwen.py

@@ -0,0 +1,598 @@
+# 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 requests
+import unicodedata
+from typing import Collection, Dict, List, Set, Tuple, Union, Any, Callable, Optional
+
+import tiktoken
+import numpy as np
+from PIL import Image
+from PIL import ImageFont
+from PIL import ImageDraw
+from transformers import PreTrainedTokenizer, AddedToken
+from transformers.utils import try_to_load_from_cache
+
+import matplotlib.colors as mcolors
+from matplotlib.font_manager import FontProperties
+
+logger = logging.getLogger(__name__)
+
+
+VOCAB_FILES_NAMES = {"vocab_file": "qwen.tiktoken", "ttf": "SimSun.ttf"}
+FONT_PATH = try_to_load_from_cache("Qwen/Qwen-VL-Chat", "SimSun.ttf")
+if FONT_PATH is None:
+    if not os.path.exists("SimSun.ttf"):
+        ttf = requests.get("https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/SimSun.ttf")
+        open("SimSun.ttf", "wb").write(ttf.content)
+    FONT_PATH = "SimSun.ttf"
+
+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)))
+SPECIAL_TOKENS = (
+    ENDOFTEXT,
+    IMSTART,
+    IMEND,
+) + EXTRAS
+IMG_TOKEN_SPAN = 256
+
+
+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)
+    }
+
+def _list_find(
+    input_list: List[Any],
+    candidates: Tuple[Any],
+    start: int = 0,
+):
+    for i in range(start, len(input_list)):
+        if input_list[i] in candidates:
+            return i
+    return -1
+
+def _replace_closed_tag(
+    input_tokens: List[Any],
+    start_tags: Union[Any, Tuple[Any]],
+    end_tags: Union[Any, Tuple[Any]],
+    inclusive_replace_func: Callable,
+    exclusive_replace_func: Callable = lambda x: x,
+):
+    if isinstance(start_tags, (str, int)):
+        start_tags = (start_tags,)
+    if isinstance(end_tags, (str, int)):
+        end_tags = (end_tags,)
+    assert len(start_tags) == len(end_tags)
+
+    output_tokens = []
+    end = 0
+    while True:
+        start = _list_find(input_tokens, start_tags, end)
+        if start == -1:
+            break
+        output_tokens.extend(exclusive_replace_func(input_tokens[end : start]))
+        tag_idx = start_tags.index(input_tokens[start])
+        end = _list_find(input_tokens, (end_tags[tag_idx],), start)
+        if end == -1:
+            raise ValueError("Unclosed image token")
+        output_tokens.extend(inclusive_replace_func(input_tokens[start : end + 1]))
+        end += 1
+    output_tokens.extend(exclusive_replace_func(input_tokens[end : ]))
+    return output_tokens
+
+class QWenTokenizer(PreTrainedTokenizer):
+    """QWen tokenizer."""
+
+    vocab_files_names = VOCAB_FILES_NAMES
+
+    def __init__(
+        self,
+        vocab_file,
+        errors="replace",
+        image_start_tag='<img>',
+        image_end_tag='</img>',
+        image_pad_tag='<imgpad>',
+        ref_start_tag='<ref>',
+        ref_end_tag='</ref>',
+        box_start_tag='<box>',
+        box_end_tag='</box>',
+        quad_start_tag='<quad>',
+        quad_end_tag='</quad>',
+        **kwargs,
+        ): 
+        # super().__init__(**kwargs)
+        self.image_start_tag = image_start_tag
+        self.image_end_tag = image_end_tag
+        self.image_pad_tag = image_pad_tag
+        self.ref_start_tag = ref_start_tag
+        self.ref_end_tag = ref_end_tag
+        self.box_start_tag = box_start_tag
+        self.box_end_tag = box_end_tag
+        self.quad_start_tag = quad_start_tag
+        self.quad_end_tag = quad_end_tag
+        self.IMAGE_ST = (
+            ref_start_tag, ref_end_tag,
+            box_start_tag, box_end_tag,
+            quad_start_tag, quad_end_tag,
+            image_start_tag, image_end_tag,
+            image_pad_tag
+        )
+        super().__init__(**kwargs)
+        self.errors = errors  # how to handle errors in decoding
+
+        self.mergeable_ranks = _load_tiktoken_bpe(vocab_file)  # type: dict[bytes, int]
+        self.special_tokens = {
+            token: index
+            for index, token in enumerate(
+                SPECIAL_TOKENS + self.IMAGE_ST, start=len(self.mergeable_ranks)
+            )
+        }
+        self.img_start_id = self.special_tokens[self.image_start_tag]
+        self.img_end_id = self.special_tokens[self.image_end_tag]
+        self.img_pad_id = self.special_tokens[self.image_pad_tag]
+        self.ref_start_id = self.special_tokens[self.ref_start_tag]
+        self.ref_end_id = self.special_tokens[self.ref_end_tag]
+        self.box_start_id = self.special_tokens[self.box_start_tag]
+        self.box_end_id = self.special_tokens[self.box_end_tag]
+        self.quad_start_id = self.special_tokens[self.quad_start_tag]
+        self.quad_end_id = self.special_tokens[self.quad_end_tag]
+        self.image_special_tokens = set([
+            self.ref_start_id, self.ref_end_id, self.box_start_id, self.box_end_id,
+            self.quad_start_id, self.quad_end_id,
+        ])
+
+        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 + self.IMAGE_ST:
+                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])
+
+        def _encode_imgurl(img_tokens):
+            assert img_tokens[0] == self.image_start_tag and img_tokens[-1] == self.image_end_tag
+            img_tokens = img_tokens[1:-1]
+            img_url = b''.join(img_tokens)
+            out_img_tokens = list(map(self.decoder.get, img_url))
+            if len(out_img_tokens) > IMG_TOKEN_SPAN:
+                raise ValueError("The content in {}..{} is too long".format(
+                    self.image_start_tag, self.image_end_tag))
+            out_img_tokens.extend([self.image_pad_tag] * (IMG_TOKEN_SPAN - len(out_img_tokens)))
+            out_img_tokens = [self.image_start_tag] + out_img_tokens + [self.image_end_tag]
+            return out_img_tokens
+
+        return _replace_closed_tag(tokens, self.image_start_tag, self.image_end_tag, _encode_imgurl)
+
+    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]
+
+        def _decode_imgurl(img_token_ids):
+            assert img_token_ids[0] == self.img_start_id and img_token_ids[-1] == self.img_end_id
+            img_token_ids = img_token_ids[1:-1]
+            img_token_ids = img_token_ids[ : img_token_ids.index(self.img_pad_id)]
+            img_url = bytes(img_token_ids).decode('utf-8')
+            return [self.img_start_id] + self.tokenizer.encode(img_url) + [self.img_end_id]
+
+        token_ids = _replace_closed_tag(token_ids, self.img_start_id, self.img_end_id, _decode_imgurl)
+
+        if skip_special_tokens:
+            if kwargs.get('keep_image_special', False):
+                token_ids = [i for i in token_ids if i < self.eod_id 
+                    or i in self.image_special_tokens]
+            else:
+                token_ids = [i for i in token_ids if i < self.eod_id]
+        return self.tokenizer.decode(token_ids, errors=errors or self.errors)
+
+    def to_list_format(self, text: str):
+        text = unicodedata.normalize("NFC", text)
+        token_ids = self.tokenizer.encode(
+            text, allowed_special=set(self.IMAGE_ST + (ENDOFTEXT,)))
+
+        def _encode_vl_info(tokens):
+            if len(tokens) == 0:
+                return []
+            if tokens[0] == self.img_start_id and tokens[-1] == self.img_end_id:
+                key = 'image'
+            elif tokens[0] == self.ref_start_id and tokens[-1] == self.ref_end_id:
+                key = 'ref'
+            elif tokens[0] == self.box_start_id and tokens[-1] == self.box_end_id:
+                key = 'box'
+            elif tokens[0] == self.quad_start_id and tokens[-1] == self.quad_end_id:
+                key = 'quad'
+            else:
+                _tobytes = lambda x: x.encode('utf-8') if isinstance(x, str) else x
+                return [{'text': b''.join(map(_tobytes, map(self.decoder.get, tokens))).decode('utf-8')}]
+            _tobytes = lambda x: x.encode('utf-8') if isinstance(x, str) else x
+            val = b''.join(map(_tobytes, map(self.decoder.get, tokens[1:-1]))).decode('utf-8')
+            return [{key: val}]
+
+        return _replace_closed_tag(
+            token_ids,
+            (self.img_start_id, self.ref_start_id, self.box_start_id, self.quad_start_id),
+            (self.img_end_id, self.ref_end_id, self.box_end_id, self.quad_end_id),
+            _encode_vl_info,
+            _encode_vl_info,
+        )
+
+    def from_list_format(self, list_format: List[Dict]):
+        text = ''
+        num_images = 0
+        for ele in list_format:
+            if 'image' in ele:
+                num_images += 1
+                text += f'Picture {num_images}: '
+                text += self.image_start_tag + ele['image'] + self.image_end_tag
+                text += '\n'
+            elif 'text' in ele:
+                text += ele['text']
+            elif 'box' in ele:
+                if 'ref' in ele:
+                    text += self.ref_start_tag + ele['ref'] + self.ref_end_tag
+                for box in ele['box']:
+                    text += self.box_start_tag + '(%d,%d),(%d,%d)' % (box[0], box[1], box[2], box[3]) + self.box_end_tag
+            else:
+                raise ValueError("Unsupport element: " + str(ele))
+        return text
+
+    def _fetch_latest_picture(self, response, history):
+        if history is None:
+            history = []
+        _history = history + [(response, None)]
+        for q, r in _history[::-1]:
+            for ele in self.to_list_format(q)[::-1]:
+                if 'image' in ele:
+                    return ele['image']
+        return None
+
+    def _fetch_all_box_with_ref(self, text):
+        list_format = self.to_list_format(text)
+        output = []
+        for i, ele in enumerate(list_format):
+            if 'box' in ele:
+                bbox = tuple(map(int, ele['box'].replace('(', '').replace(')', '').split(',')))
+                assert len(bbox) == 4
+                output.append({'box': bbox})
+                if i > 0 and 'ref' in list_format[i-1]:
+                    output[-1]['ref'] = list_format[i-1]['ref'].strip()
+        return output
+
+    def draw_bbox_on_latest_picture(
+        self,
+        response,
+        history=None,
+    ) -> Optional[Image.Image]:
+        image = self._fetch_latest_picture(response, history)
+        if image is None:
+            return None
+        if image.startswith("http://") or image.startswith("https://"):
+            image = Image.open(requests.get(image, stream=True).raw).convert("RGB")
+            h, w = image.height, image.width
+        else:
+            image = np.asarray(Image.open(image).convert("RGB"))
+            h, w = image.shape[0], image.shape[1]
+        visualizer = Visualizer(image)
+
+        boxes = self._fetch_all_box_with_ref(response)
+        if not boxes:
+            return None
+        color = random.choice([_ for _ in mcolors.TABLEAU_COLORS.keys()]) # init color
+        for box in boxes:
+            if 'ref' in box: # random new color for new refexps
+                color = random.choice([_ for _ in mcolors.TABLEAU_COLORS.keys()])
+            x1, y1, x2, y2 = box['box']
+            x1, y1, x2, y2 = (int(x1 / 1000 * w), int(y1 / 1000 * h), int(x2 / 1000 * w), int(y2 / 1000 * h))
+            visualizer.draw_box((x1, y1, x2, y2), alpha=1, edge_color=color)
+            if 'ref' in box:
+                visualizer.draw_text(box['ref'], (x1, y1), color=color, horizontal_alignment="left")
+        return visualizer.output
+
+
+import colorsys
+import logging
+import math
+import numpy as np
+import matplotlib as mpl
+import matplotlib.colors as mplc
+import matplotlib.figure as mplfigure
+import torch
+from matplotlib.backends.backend_agg import FigureCanvasAgg
+from PIL import Image
+import random
+
+logger = logging.getLogger(__name__)
+
+
+class VisImage:
+    def __init__(self, img, scale=1.0):
+        self.img = img
+        self.scale = scale
+        self.width, self.height = img.shape[1], img.shape[0]
+        self._setup_figure(img)
+
+    def _setup_figure(self, img):
+        fig = mplfigure.Figure(frameon=False)
+        self.dpi = fig.get_dpi()
+        # add a small 1e-2 to avoid precision lost due to matplotlib's truncation
+        # (https://github.com/matplotlib/matplotlib/issues/15363)
+        fig.set_size_inches(
+            (self.width * self.scale + 1e-2) / self.dpi,
+            (self.height * self.scale + 1e-2) / self.dpi,
+        )
+        self.canvas = FigureCanvasAgg(fig)
+        # self.canvas = mpl.backends.backend_cairo.FigureCanvasCairo(fig)
+        ax = fig.add_axes([0.0, 0.0, 1.0, 1.0])
+        ax.axis("off")
+        self.fig = fig
+        self.ax = ax
+        self.reset_image(img)
+
+    def reset_image(self, img):
+        img = img.astype("uint8")
+        self.ax.imshow(img, extent=(0, self.width, self.height, 0), interpolation="nearest")
+
+    def save(self, filepath):
+        self.fig.savefig(filepath)
+
+    def get_image(self):
+        canvas = self.canvas
+        s, (width, height) = canvas.print_to_buffer()
+
+        buffer = np.frombuffer(s, dtype="uint8")
+
+        img_rgba = buffer.reshape(height, width, 4)
+        rgb, alpha = np.split(img_rgba, [3], axis=2)
+        return rgb.astype("uint8")
+
+
+class Visualizer:
+    def __init__(self, img_rgb, metadata=None, scale=1.0):
+        self.img = np.asarray(img_rgb).clip(0, 255).astype(np.uint8)
+        self.font_path = FONT_PATH
+        self.output = VisImage(self.img, scale=scale)
+        self.cpu_device = torch.device("cpu")
+
+        # too small texts are useless, therefore clamp to 14
+        self._default_font_size = max(
+            np.sqrt(self.output.height * self.output.width) // 30, 15 // scale
+        )
+
+    def draw_text(
+        self,
+        text,
+        position,
+        *,
+        font_size=None,
+        color="g",
+        horizontal_alignment="center",
+        rotation=0,
+    ):
+        if not font_size:
+            font_size = self._default_font_size
+
+        # since the text background is dark, we don't want the text to be dark
+        color = np.maximum(list(mplc.to_rgb(color)), 0.2)
+        color[np.argmax(color)] = max(0.8, np.max(color))
+
+        x, y = position
+        self.output.ax.text(
+            x,
+            y,
+            text,
+            size=font_size * self.output.scale,
+            fontproperties=FontProperties(fname=self.font_path),
+            bbox={"facecolor": "black", "alpha": 0.8, "pad": 0.7, "edgecolor": "none"},
+            verticalalignment="top",
+            horizontalalignment=horizontal_alignment,
+            color=color,
+            zorder=10,
+            rotation=rotation,
+        )
+        return self.output
+
+    def draw_box(self, box_coord, alpha=0.5, edge_color="g", line_style="-"):
+        
+        x0, y0, x1, y1 = box_coord
+        width = x1 - x0
+        height = y1 - y0
+
+        linewidth = max(self._default_font_size / 4, 1)
+
+        self.output.ax.add_patch(
+            mpl.patches.Rectangle(
+                (x0, y0),
+                width,
+                height,
+                fill=False,
+                edgecolor=edge_color,
+                linewidth=linewidth * self.output.scale,
+                alpha=alpha,
+                linestyle=line_style,
+            )
+        )
+        return self.output
+
+    def get_output(self):
+        
+        return self.output

tokenizer_config.json

@@ -0,0 +1,14 @@
+{
+  "added_tokens_decoder": {},
+  "auto_map": {
+    "AutoTokenizer": [
+      "tokenization_qwen.QWenTokenizer",
+      null
+    ]
+  },
+  "clean_up_tokenization_spaces": true,
+  "model_max_length": 512,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "tokenizer_class": "QWenTokenizer"
+}

trainer_state.json

@@ -0,0 +1,3630 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.2396,
+      "step": 1
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.2073,
+      "step": 2
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-06,
+      "loss": 0.1975,
+      "step": 3
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.2081,
+      "step": 4
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.1832,
+      "step": 5
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1e-05,
+      "loss": 0.2585,
+      "step": 6
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 9.999930069625869e-06,
+      "loss": 0.2094,
+      "step": 7
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 9.999720280459576e-06,
+      "loss": 0.163,
+      "step": 8
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 9.999370638369377e-06,
+      "loss": 0.2687,
+      "step": 9
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 9.99888115313551e-06,
+      "loss": 0.2199,
+      "step": 10
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 9.998251838449932e-06,
+      "loss": 0.254,
+      "step": 11
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 9.997482711915926e-06,
+      "loss": 0.185,
+      "step": 12
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 9.996573795047616e-06,
+      "loss": 0.1471,
+      "step": 13
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 9.99552511326936e-06,
+      "loss": 0.2113,
+      "step": 14
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 9.994336695915041e-06,
+      "loss": 0.1411,
+      "step": 15
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.993008576227248e-06,
+      "loss": 0.215,
+      "step": 16
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.991540791356342e-06,
+      "loss": 0.2258,
+      "step": 17
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.989933382359423e-06,
+      "loss": 0.1529,
+      "step": 18
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.988186394199175e-06,
+      "loss": 0.1812,
+      "step": 19
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.986299875742612e-06,
+      "loss": 0.2054,
+      "step": 20
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.984273879759713e-06,
+      "loss": 0.207,
+      "step": 21
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.982108462921938e-06,
+      "loss": 0.1292,
+      "step": 22
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.979803685800651e-06,
+      "loss": 0.1894,
+      "step": 23
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 0.1512,
+      "step": 24
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.97477631248223e-06,
+      "loss": 0.2282,
+      "step": 25
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.972053856911534e-06,
+      "loss": 0.1739,
+      "step": 26
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.969192322306271e-06,
+      "loss": 0.1747,
+      "step": 27
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.1059,
+      "step": 28
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.963052340053242e-06,
+      "loss": 0.2257,
+      "step": 29
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.959774064153977e-06,
+      "loss": 0.1449,
+      "step": 30
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.956357052712347e-06,
+      "loss": 0.1357,
+      "step": 31
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.952801401309504e-06,
+      "loss": 0.1424,
+      "step": 32
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.1842,
+      "step": 33
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.945274580332316e-06,
+      "loss": 0.1781,
+      "step": 34
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.941303621299332e-06,
+      "loss": 0.1825,
+      "step": 35
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.937194443381972e-06,
+      "loss": 0.1603,
+      "step": 36
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.932947161522779e-06,
+      "loss": 0.1809,
+      "step": 37
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.928561894527354e-06,
+      "loss": 0.188,
+      "step": 38
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.112,
+      "step": 39
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.919377899645497e-06,
+      "loss": 0.1943,
+      "step": 40
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.914579428655144e-06,
+      "loss": 0.13,
+      "step": 41
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 0.14,
+      "step": 42
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.904570210689584e-06,
+      "loss": 0.1408,
+      "step": 43
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.899359743693715e-06,
+      "loss": 0.1618,
+      "step": 44
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.894012231073895e-06,
+      "loss": 0.1458,
+      "step": 45
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.888527822411543e-06,
+      "loss": 0.1471,
+      "step": 46
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.882906671117363e-06,
+      "loss": 0.1468,
+      "step": 47
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.877148934427037e-06,
+      "loss": 0.1638,
+      "step": 48
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.871254773396837e-06,
+      "loss": 0.1769,
+      "step": 49
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.1476,
+      "step": 50
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.859057841617709e-06,
+      "loss": 0.1301,
+      "step": 51
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.85275541204318e-06,
+      "loss": 0.1658,
+      "step": 52
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 9.84631724046804e-06,
+      "loss": 0.1639,
+      "step": 53
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.839743506981783e-06,
+      "loss": 0.1746,
+      "step": 54
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 9.833034395465866e-06,
+      "loss": 0.1701,
+      "step": 55
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 9.826190093588564e-06,
+      "loss": 0.1561,
+      "step": 56
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 9.819210792799711e-06,
+      "loss": 0.1779,
+      "step": 57
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.812096688325354e-06,
+      "loss": 0.1498,
+      "step": 58
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 9.804847979162286e-06,
+      "loss": 0.1712,
+      "step": 59
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.1423,
+      "step": 60
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 9.789947561577445e-06,
+      "loss": 0.1548,
+      "step": 61
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 9.78229626995238e-06,
+      "loss": 0.143,
+      "step": 62
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 9.774511207220369e-06,
+      "loss": 0.1465,
+      "step": 63
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 9.766592591146353e-06,
+      "loss": 0.1667,
+      "step": 64
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 9.758540643231041e-06,
+      "loss": 0.1508,
+      "step": 65
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 9.750355588704728e-06,
+      "loss": 0.1483,
+      "step": 66
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 9.742037656520984e-06,
+      "loss": 0.1231,
+      "step": 67
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 9.733587079350254e-06,
+      "loss": 0.1446,
+      "step": 68
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 9.725004093573343e-06,
+      "loss": 0.1363,
+      "step": 69
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 9.716288939274818e-06,
+      "loss": 0.1272,
+      "step": 70
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 9.707441860236278e-06,
+      "loss": 0.1652,
+      "step": 71
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.1762,
+      "step": 72
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 9.689352921509725e-06,
+      "loss": 0.1773,
+      "step": 73
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 9.680111567808212e-06,
+      "loss": 0.1138,
+      "step": 74
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 9.670739301325534e-06,
+      "loss": 0.1497,
+      "step": 75
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 9.66123638422413e-06,
+      "loss": 0.1163,
+      "step": 76
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 9.651603082321019e-06,
+      "loss": 0.1411,
+      "step": 77
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.1309,
+      "step": 78
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 9.631946405605932e-06,
+      "loss": 0.1285,
+      "step": 79
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 9.621923580633462e-06,
+      "loss": 0.1472,
+      "step": 80
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 9.611771470522908e-06,
+      "loss": 0.1654,
+      "step": 81
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 9.601490359250616e-06,
+      "loss": 0.1636,
+      "step": 82
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 9.591080534401371e-06,
+      "loss": 0.1448,
+      "step": 83
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 9.580542287160348e-06,
+      "loss": 0.1302,
+      "step": 84
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 9.56987591230498e-06,
+      "loss": 0.1363,
+      "step": 85
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 9.559081708196696e-06,
+      "loss": 0.1276,
+      "step": 86
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.1095,
+      "step": 87
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 9.537111023536973e-06,
+      "loss": 0.1052,
+      "step": 88
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 9.525935157552813e-06,
+      "loss": 0.1609,
+      "step": 89
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 9.514632691433108e-06,
+      "loss": 0.1164,
+      "step": 90
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 9.503203941332128e-06,
+      "loss": 0.1277,
+      "step": 91
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 9.491649226936586e-06,
+      "loss": 0.13,
+      "step": 92
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.47996887145668e-06,
+      "loss": 0.1407,
+      "step": 93
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.1278,
+      "step": 94
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.456232547647695e-06,
+      "loss": 0.1114,
+      "step": 95
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.1358,
+      "step": 96
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.43199762571061e-06,
+      "loss": 0.1187,
+      "step": 97
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.419694035645753e-06,
+      "loss": 0.1518,
+      "step": 98
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 9.40726681723791e-06,
+      "loss": 0.1261,
+      "step": 99
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 9.394716318103098e-06,
+      "loss": 0.1149,
+      "step": 100
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 9.382042889305754e-06,
+      "loss": 0.124,
+      "step": 101
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 9.369246885348926e-06,
+      "loss": 0.1344,
+      "step": 102
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 9.35632866416435e-06,
+      "loss": 0.0972,
+      "step": 103
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 9.343288587102444e-06,
+      "loss": 0.1138,
+      "step": 104
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.1479,
+      "step": 105
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 9.316844327780955e-06,
+      "loss": 0.1345,
+      "step": 106
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 9.303440885224153e-06,
+      "loss": 0.1184,
+      "step": 107
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 9.289917066174887e-06,
+      "loss": 0.0922,
+      "step": 108
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 9.276273248923446e-06,
+      "loss": 0.096,
+      "step": 109
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 9.262509815116732e-06,
+      "loss": 0.1213,
+      "step": 110
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 9.248627149747573e-06,
+      "loss": 0.1172,
+      "step": 111
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 9.234625641143962e-06,
+      "loss": 0.1318,
+      "step": 112
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 9.220505680958194e-06,
+      "loss": 0.1044,
+      "step": 113
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.131,
+      "step": 114
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 9.191911989005038e-06,
+      "loss": 0.1391,
+      "step": 115
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.1194,
+      "step": 116
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 9.162849273173857e-06,
+      "loss": 0.1123,
+      "step": 117
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 9.148143045440181e-06,
+      "loss": 0.1444,
+      "step": 118
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 9.133320785228457e-06,
+      "loss": 0.121,
+      "step": 119
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 9.118382907149164e-06,
+      "loss": 0.1261,
+      "step": 120
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 9.103329829046866e-06,
+      "loss": 0.1203,
+      "step": 121
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 9.088161971988517e-06,
+      "loss": 0.108,
+      "step": 122
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 9.07287976025168e-06,
+      "loss": 0.0694,
+      "step": 123
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 9.057483621312671e-06,
+      "loss": 0.0995,
+      "step": 124
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 9.041973985834595e-06,
+      "loss": 0.1043,
+      "step": 125
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 9.026351287655294e-06,
+      "loss": 0.1378,
+      "step": 126
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 9.01061596377522e-06,
+      "loss": 0.1135,
+      "step": 127
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 8.994768454345207e-06,
+      "loss": 0.1234,
+      "step": 128
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 8.978809202654161e-06,
+      "loss": 0.118,
+      "step": 129
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 8.96273865511666e-06,
+      "loss": 0.1136,
+      "step": 130
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 8.94655726126046e-06,
+      "loss": 0.0982,
+      "step": 131
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.079,
+      "step": 132
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 8.91386374819341e-06,
+      "loss": 0.0974,
+      "step": 133
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 8.897352543490396e-06,
+      "loss": 0.1122,
+      "step": 134
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 8.880732321458785e-06,
+      "loss": 0.0958,
+      "step": 135
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 8.864003547001916e-06,
+      "loss": 0.0825,
+      "step": 136
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 8.847166688059572e-06,
+      "loss": 0.1226,
+      "step": 137
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.1335,
+      "step": 138
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 8.81317060358119e-06,
+      "loss": 0.0939,
+      "step": 139
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 8.796012328988716e-06,
+      "loss": 0.125,
+      "step": 140
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.114,
+      "step": 141
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 8.7613777148529e-06,
+      "loss": 0.0861,
+      "step": 142
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 8.743902344114163e-06,
+      "loss": 0.1096,
+      "step": 143
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 8.726322248378775e-06,
+      "loss": 0.0803,
+      "step": 144
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 8.708637919399798e-06,
+      "loss": 0.0833,
+      "step": 145
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 8.690849851845933e-06,
+      "loss": 0.0882,
+      "step": 146
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 8.672958543287666e-06,
+      "loss": 0.1513,
+      "step": 147
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 8.65496449418336e-06,
+      "loss": 0.0996,
+      "step": 148
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 8.636868207865244e-06,
+      "loss": 0.136,
+      "step": 149
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.0888,
+      "step": 150
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 8.600370951201345e-06,
+      "loss": 0.1044,
+      "step": 151
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 8.581971001762287e-06,
+      "loss": 0.0964,
+      "step": 152
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 8.563470856894316e-06,
+      "loss": 0.0946,
+      "step": 153
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 8.54487103408625e-06,
+      "loss": 0.1264,
+      "step": 154
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 8.526172053615122e-06,
+      "loss": 0.1173,
+      "step": 155
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.507374438531606e-06,
+      "loss": 0.1075,
+      "step": 156
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.48847871464541e-06,
+      "loss": 0.0625,
+      "step": 157
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.469485410510545e-06,
+      "loss": 0.0951,
+      "step": 158
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 8.450395057410561e-06,
+      "loss": 0.1103,
+      "step": 159
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.0794,
+      "step": 160
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 8.411925343007815e-06,
+      "loss": 0.0699,
+      "step": 161
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 8.392547057785662e-06,
+      "loss": 0.1138,
+      "step": 162
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 8.373073875729504e-06,
+      "loss": 0.1351,
+      "step": 163
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 8.353506341546106e-06,
+      "loss": 0.0853,
+      "step": 164
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 8.33384500258146e-06,
+      "loss": 0.1151,
+      "step": 165
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 8.314090408805481e-06,
+      "loss": 0.043,
+      "step": 166
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 8.294243112796625e-06,
+      "loss": 0.0859,
+      "step": 167
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.0728,
+      "step": 168
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 8.254272637343968e-06,
+      "loss": 0.0809,
+      "step": 169
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 8.234150575960288e-06,
+      "loss": 0.0699,
+      "step": 170
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.1079,
+      "step": 171
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 8.193635620149041e-06,
+      "loss": 0.0708,
+      "step": 172
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.173243859011881e-06,
+      "loss": 0.1004,
+      "step": 173
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 8.152763335422612e-06,
+      "loss": 0.0615,
+      "step": 174
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 8.132194622265508e-06,
+      "loss": 0.089,
+      "step": 175
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 8.111538294891684e-06,
+      "loss": 0.1052,
+      "step": 176
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 8.090794931103026e-06,
+      "loss": 0.0871,
+      "step": 177
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 8.06996511113601e-06,
+      "loss": 0.1042,
+      "step": 178
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 8.049049417645478e-06,
+      "loss": 0.0477,
+      "step": 179
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 8.028048435688333e-06,
+      "loss": 0.0636,
+      "step": 180
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 8.006962752707193e-06,
+      "loss": 0.0742,
+      "step": 181
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.0574,
+      "step": 182
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 7.964539645273204e-06,
+      "loss": 0.0414,
+      "step": 183
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 7.943203407485864e-06,
+      "loss": 0.0505,
+      "step": 184
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 7.921784841972355e-06,
+      "loss": 0.0961,
+      "step": 185
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.0781,
+      "step": 186
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 7.87870312654622e-06,
+      "loss": 0.0598,
+      "step": 187
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 7.857041181721788e-06,
+      "loss": 0.0832,
+      "step": 188
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 7.835299319313854e-06,
+      "loss": 0.049,
+      "step": 189
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 7.813478147489052e-06,
+      "loss": 0.056,
+      "step": 190
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 7.791578276632461e-06,
+      "loss": 0.0552,
+      "step": 191
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 7.769600319330553e-06,
+      "loss": 0.0669,
+      "step": 192
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 7.747544890354031e-06,
+      "loss": 0.0643,
+      "step": 193
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 7.725412606640658e-06,
+      "loss": 0.0586,
+      "step": 194
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.0892,
+      "step": 195
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 7.680919953486047e-06,
+      "loss": 0.0879,
+      "step": 196
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 7.65856082859996e-06,
+      "loss": 0.0763,
+      "step": 197
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 7.636127338052513e-06,
+      "loss": 0.072,
+      "step": 198
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 7.613620109356663e-06,
+      "loss": 0.0738,
+      "step": 199
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 7.5910397720879785e-06,
+      "loss": 0.0829,
+      "step": 200
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 7.568386957867033e-06,
+      "loss": 0.0622,
+      "step": 201
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 7.545662300341736e-06,
+      "loss": 0.0665,
+      "step": 202
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 7.522866435169606e-06,
+      "loss": 0.0769,
+      "step": 203
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0253,
+      "step": 204
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 7.477063634456263e-06,
+      "loss": 0.0213,
+      "step": 205
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 7.454057980117842e-06,
+      "loss": 0.0702,
+      "step": 206
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 7.430983680502344e-06,
+      "loss": 0.0982,
+      "step": 207
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 7.407841381047533e-06,
+      "loss": 0.0552,
+      "step": 208
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 7.38463172909327e-06,
+      "loss": 0.0744,
+      "step": 209
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 7.361355373863415e-06,
+      "loss": 0.0463,
+      "step": 210
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 7.3380129664476574e-06,
+      "loss": 0.0986,
+      "step": 211
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 7.314605159783313e-06,
+      "loss": 0.0715,
+      "step": 212
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 7.291132608637053e-06,
+      "loss": 0.05,
+      "step": 213
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 7.2675959695865896e-06,
+      "loss": 0.046,
+      "step": 214
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 7.243995901002312e-06,
+      "loss": 0.051,
+      "step": 215
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 7.2203330630288714e-06,
+      "loss": 0.0274,
+      "step": 216
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 7.196608117566714e-06,
+      "loss": 0.0327,
+      "step": 217
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 7.172821728253563e-06,
+      "loss": 0.0467,
+      "step": 218
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 7.148974560445859e-06,
+      "loss": 0.0435,
+      "step": 219
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 7.1250672812001505e-06,
+      "loss": 0.0794,
+      "step": 220
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 7.1011005592544314e-06,
+      "loss": 0.0072,
+      "step": 221
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.0974,
+      "step": 222
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 7.0529914705098755e-06,
+      "loss": 0.0889,
+      "step": 223
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 7.02885044942567e-06,
+      "loss": 0.0819,
+      "step": 224
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 7.004652677033069e-06,
+      "loss": 0.0599,
+      "step": 225
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.0198,
+      "step": 226
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 6.95608958734605e-06,
+      "loss": 0.0512,
+      "step": 227
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 6.931725628465643e-06,
+      "loss": 0.0743,
+      "step": 228
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 6.90730763506687e-06,
+      "loss": 0.0887,
+      "step": 229
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 6.882836290173493e-06,
+      "loss": 0.0337,
+      "step": 230
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 6.858312278301638e-06,
+      "loss": 0.0356,
+      "step": 231
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 6.833736285440632e-06,
+      "loss": 0.0387,
+      "step": 232
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 6.809108999033825e-06,
+      "loss": 0.027,
+      "step": 233
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 6.78443110795936e-06,
+      "loss": 0.0262,
+      "step": 234
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 6.759703302510898e-06,
+      "loss": 0.0314,
+      "step": 235
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 6.734926274378313e-06,
+      "loss": 0.0691,
+      "step": 236
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.0528,
+      "step": 237
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 6.685227323685209e-06,
+      "loss": 0.0522,
+      "step": 238
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 6.660306791311177e-06,
+      "loss": 0.052,
+      "step": 239
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 0.0454,
+      "step": 240
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 6.610327097892959e-06,
+      "loss": 0.0161,
+      "step": 241
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 6.5852693348882345e-06,
+      "loss": 0.0154,
+      "step": 242
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 6.560167228492436e-06,
+      "loss": 0.0682,
+      "step": 243
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 6.535021480865439e-06,
+      "loss": 0.0074,
+      "step": 244
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 6.5098327953878585e-06,
+      "loss": 0.0618,
+      "step": 245
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 6.484601876641375e-06,
+      "loss": 0.0378,
+      "step": 246
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 6.459329430389023e-06,
+      "loss": 0.0436,
+      "step": 247
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.0143,
+      "step": 248
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.0446,
+      "step": 249
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 6.383270001532636e-06,
+      "loss": 0.0124,
+      "step": 250
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 6.357838525822629e-06,
+      "loss": 0.043,
+      "step": 251
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 6.332369068450175e-06,
+      "loss": 0.0243,
+      "step": 252
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 6.306862341850746e-06,
+      "loss": 0.0199,
+      "step": 253
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 6.2813190595023135e-06,
+      "loss": 0.0609,
+      "step": 254
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 6.255739935905396e-06,
+      "loss": 0.0373,
+      "step": 255
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 6.230125686563068e-06,
+      "loss": 0.0216,
+      "step": 256
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 6.204477027960942e-06,
+      "loss": 0.0485,
+      "step": 257
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 0.0184,
+      "step": 258
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 6.153079353712201e-06,
+      "loss": 0.0303,
+      "step": 259
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 6.127331775769023e-06,
+      "loss": 0.048,
+      "step": 260
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 6.101552663932704e-06,
+      "loss": 0.0352,
+      "step": 261
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 6.07574273930042e-06,
+      "loss": 0.0195,
+      "step": 262
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 6.049902723831243e-06,
+      "loss": 0.0322,
+      "step": 263
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 6.024033340325954e-06,
+      "loss": 0.0385,
+      "step": 264
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 5.998135312406821e-06,
+      "loss": 0.0678,
+      "step": 265
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 5.972209364497355e-06,
+      "loss": 0.0341,
+      "step": 266
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 5.946256221802052e-06,
+      "loss": 0.019,
+      "step": 267
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 5.920276610286102e-06,
+      "loss": 0.0372,
+      "step": 268
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 5.894271256655089e-06,
+      "loss": 0.0278,
+      "step": 269
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0079,
+      "step": 270
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 5.842186233450153e-06,
+      "loss": 0.0168,
+      "step": 271
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 5.816108020806297e-06,
+      "loss": 0.0289,
+      "step": 272
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 5.79000697986675e-06,
+      "loss": 0.0483,
+      "step": 273
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 5.763883840733736e-06,
+      "loss": 0.019,
+      "step": 274
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 5.737739334127611e-06,
+      "loss": 0.0182,
+      "step": 275
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.0339,
+      "step": 276
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 5.685389144345474e-06,
+      "loss": 0.012,
+      "step": 277
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 5.659184925516802e-06,
+      "loss": 0.0154,
+      "step": 278
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 5.632962267868747e-06,
+      "loss": 0.0178,
+      "step": 279
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 5.60672190490541e-06,
+      "loss": 0.0142,
+      "step": 280
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 5.5804645706261515e-06,
+      "loss": 0.0175,
+      "step": 281
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 5.5541909995050554e-06,
+      "loss": 0.0092,
+      "step": 282
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 5.527901926470386e-06,
+      "loss": 0.0102,
+      "step": 283
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 5.5015980868840254e-06,
+      "loss": 0.0027,
+      "step": 284
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 5.475280216520913e-06,
+      "loss": 0.0214,
+      "step": 285
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 5.448949051548459e-06,
+      "loss": 0.0354,
+      "step": 286
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 5.4226053285059466e-06,
+      "loss": 0.0121,
+      "step": 287
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 5.396249784283943e-06,
+      "loss": 0.0403,
+      "step": 288
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 5.369883156103672e-06,
+      "loss": 0.0216,
+      "step": 289
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 5.343506181496405e-06,
+      "loss": 0.019,
+      "step": 290
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 5.317119598282823e-06,
+      "loss": 0.0176,
+      "step": 291
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.0098,
+      "step": 292
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 5.264320558642657e-06,
+      "loss": 0.058,
+      "step": 293
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 0.0116,
+      "step": 294
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 5.2114919447524155e-06,
+      "loss": 0.0131,
+      "step": 295
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 5.185068394501791e-06,
+      "loss": 0.0202,
+      "step": 296
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 5.15863966749034e-06,
+      "loss": 0.007,
+      "step": 297
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 5.132206502986368e-06,
+      "loss": 0.0135,
+      "step": 298
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 5.105769640382309e-06,
+      "loss": 0.0064,
+      "step": 299
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 5.07932981917404e-06,
+      "loss": 0.0091,
+      "step": 300
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 5.052887778940199e-06,
+      "loss": 0.0057,
+      "step": 301
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 5.026444259321489e-06,
+      "loss": 0.0136,
+      "step": 302
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 5e-06,
+      "loss": 0.003,
+      "step": 303
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 4.973555740678512e-06,
+      "loss": 0.0096,
+      "step": 304
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 4.947112221059803e-06,
+      "loss": 0.0278,
+      "step": 305
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 4.9206701808259605e-06,
+      "loss": 0.0257,
+      "step": 306
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 4.894230359617693e-06,
+      "loss": 0.0029,
+      "step": 307
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 4.867793497013634e-06,
+      "loss": 0.0127,
+      "step": 308
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 4.841360332509663e-06,
+      "loss": 0.0065,
+      "step": 309
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 4.81493160549821e-06,
+      "loss": 0.0159,
+      "step": 310
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 4.788508055247585e-06,
+      "loss": 0.0085,
+      "step": 311
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 0.0065,
+      "step": 312
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 4.735679441357343e-06,
+      "loss": 0.0074,
+      "step": 313
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.016,
+      "step": 314
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 4.682880401717178e-06,
+      "loss": 0.0074,
+      "step": 315
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 4.6564938185035954e-06,
+      "loss": 0.005,
+      "step": 316
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 4.630116843896329e-06,
+      "loss": 0.0056,
+      "step": 317
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 4.603750215716057e-06,
+      "loss": 0.0043,
+      "step": 318
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 4.577394671494054e-06,
+      "loss": 0.0063,
+      "step": 319
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 4.551050948451542e-06,
+      "loss": 0.0039,
+      "step": 320
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 4.524719783479088e-06,
+      "loss": 0.004,
+      "step": 321
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 4.498401913115975e-06,
+      "loss": 0.0236,
+      "step": 322
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 4.472098073529617e-06,
+      "loss": 0.004,
+      "step": 323
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 4.445809000494945e-06,
+      "loss": 0.0222,
+      "step": 324
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 4.4195354293738484e-06,
+      "loss": 0.0112,
+      "step": 325
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 4.393278095094591e-06,
+      "loss": 0.0101,
+      "step": 326
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 4.367037732131254e-06,
+      "loss": 0.0115,
+      "step": 327
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 4.340815074483199e-06,
+      "loss": 0.0028,
+      "step": 328
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 4.314610855654528e-06,
+      "loss": 0.0088,
+      "step": 329
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.0025,
+      "step": 330
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 4.26226066587239e-06,
+      "loss": 0.0166,
+      "step": 331
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 4.2361161592662655e-06,
+      "loss": 0.0093,
+      "step": 332
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 4.209993020133251e-06,
+      "loss": 0.0092,
+      "step": 333
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 4.183891979193703e-06,
+      "loss": 0.0097,
+      "step": 334
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 4.1578137665498485e-06,
+      "loss": 0.0044,
+      "step": 335
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.008,
+      "step": 336
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 4.1057287433449145e-06,
+      "loss": 0.0326,
+      "step": 337
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 4.079723389713899e-06,
+      "loss": 0.0037,
+      "step": 338
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 4.053743778197951e-06,
+      "loss": 0.0109,
+      "step": 339
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 4.027790635502646e-06,
+      "loss": 0.0015,
+      "step": 340
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 4.00186468759318e-06,
+      "loss": 0.006,
+      "step": 341
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 3.975966659674048e-06,
+      "loss": 0.0029,
+      "step": 342
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 3.950097276168758e-06,
+      "loss": 0.0036,
+      "step": 343
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 3.924257260699583e-06,
+      "loss": 0.0037,
+      "step": 344
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 3.898447336067297e-06,
+      "loss": 0.0146,
+      "step": 345
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 3.872668224230979e-06,
+      "loss": 0.0193,
+      "step": 346
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 3.8469206462878e-06,
+      "loss": 0.001,
+      "step": 347
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 0.0143,
+      "step": 348
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 3.7955229720390595e-06,
+      "loss": 0.0022,
+      "step": 349
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 3.769874313436933e-06,
+      "loss": 0.0062,
+      "step": 350
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 3.7442600640946045e-06,
+      "loss": 0.0029,
+      "step": 351
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 3.7186809404976877e-06,
+      "loss": 0.0108,
+      "step": 352
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 3.6931376581492564e-06,
+      "loss": 0.0076,
+      "step": 353
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 3.667630931549826e-06,
+      "loss": 0.0025,
+      "step": 354
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 3.6421614741773702e-06,
+      "loss": 0.0051,
+      "step": 355
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 3.6167299984673655e-06,
+      "loss": 0.0073,
+      "step": 356
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.0081,
+      "step": 357
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.0129,
+      "step": 358
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 3.5406705696109777e-06,
+      "loss": 0.0028,
+      "step": 359
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 3.5153981233586277e-06,
+      "loss": 0.0128,
+      "step": 360
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 3.4901672046121428e-06,
+      "loss": 0.0101,
+      "step": 361
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 3.4649785191345613e-06,
+      "loss": 0.0034,
+      "step": 362
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 3.439832771507565e-06,
+      "loss": 0.0035,
+      "step": 363
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 3.4147306651117663e-06,
+      "loss": 0.0021,
+      "step": 364
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 3.389672902107044e-06,
+      "loss": 0.0031,
+      "step": 365
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 0.0047,
+      "step": 366
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 3.3396932086888245e-06,
+      "loss": 0.0213,
+      "step": 367
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 3.3147726763147913e-06,
+      "loss": 0.0136,
+      "step": 368
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.0213,
+      "step": 369
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 3.2650737256216885e-06,
+      "loss": 0.0021,
+      "step": 370
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 3.240296697489104e-06,
+      "loss": 0.0023,
+      "step": 371
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 3.2155688920406415e-06,
+      "loss": 0.0028,
+      "step": 372
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 3.190891000966176e-06,
+      "loss": 0.0023,
+      "step": 373
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 3.16626371455937e-06,
+      "loss": 0.0015,
+      "step": 374
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 3.141687721698363e-06,
+      "loss": 0.0018,
+      "step": 375
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 3.1171637098265063e-06,
+      "loss": 0.0049,
+      "step": 376
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 3.092692364933132e-06,
+      "loss": 0.0054,
+      "step": 377
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 3.0682743715343565e-06,
+      "loss": 0.0018,
+      "step": 378
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 3.043910412653952e-06,
+      "loss": 0.0053,
+      "step": 379
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.0034,
+      "step": 380
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 2.995347322966933e-06,
+      "loss": 0.0076,
+      "step": 381
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 2.9711495505743317e-06,
+      "loss": 0.0068,
+      "step": 382
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 2.9470085294901244e-06,
+      "loss": 0.0035,
+      "step": 383
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.0029,
+      "step": 384
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 2.898899440745569e-06,
+      "loss": 0.0047,
+      "step": 385
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 2.8749327187998516e-06,
+      "loss": 0.0043,
+      "step": 386
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 2.851025439554142e-06,
+      "loss": 0.0026,
+      "step": 387
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 2.8271782717464413e-06,
+      "loss": 0.0075,
+      "step": 388
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 2.803391882433288e-06,
+      "loss": 0.0027,
+      "step": 389
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 2.7796669369711294e-06,
+      "loss": 0.0063,
+      "step": 390
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 2.7560040989976894e-06,
+      "loss": 0.0051,
+      "step": 391
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 2.7324040304134125e-06,
+      "loss": 0.0076,
+      "step": 392
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 2.708867391362948e-06,
+      "loss": 0.0022,
+      "step": 393
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 2.685394840216688e-06,
+      "loss": 0.004,
+      "step": 394
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 2.6619870335523434e-06,
+      "loss": 0.0023,
+      "step": 395
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 2.6386446261365874e-06,
+      "loss": 0.009,
+      "step": 396
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 2.6153682709067317e-06,
+      "loss": 0.0013,
+      "step": 397
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 2.5921586189524694e-06,
+      "loss": 0.0029,
+      "step": 398
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 2.5690163194976576e-06,
+      "loss": 0.0052,
+      "step": 399
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 2.5459420198821604e-06,
+      "loss": 0.0039,
+      "step": 400
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 2.5229363655437394e-06,
+      "loss": 0.0087,
+      "step": 401
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0046,
+      "step": 402
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 2.4771335648303944e-06,
+      "loss": 0.0009,
+      "step": 403
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 2.454337699658267e-06,
+      "loss": 0.0011,
+      "step": 404
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 2.4316130421329696e-06,
+      "loss": 0.0026,
+      "step": 405
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 2.4089602279120224e-06,
+      "loss": 0.003,
+      "step": 406
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 2.3863798906433396e-06,
+      "loss": 0.0022,
+      "step": 407
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 2.363872661947488e-06,
+      "loss": 0.0026,
+      "step": 408
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 2.3414391714000435e-06,
+      "loss": 0.0019,
+      "step": 409
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 2.319080046513954e-06,
+      "loss": 0.0019,
+      "step": 410
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.014,
+      "step": 411
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 2.274587393359342e-06,
+      "loss": 0.0018,
+      "step": 412
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 2.2524551096459703e-06,
+      "loss": 0.0011,
+      "step": 413
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 2.230399680669449e-06,
+      "loss": 0.0029,
+      "step": 414
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 2.2084217233675386e-06,
+      "loss": 0.0032,
+      "step": 415
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 2.1865218525109496e-06,
+      "loss": 0.0023,
+      "step": 416
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 2.1647006806861472e-06,
+      "loss": 0.0046,
+      "step": 417
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 2.1429588182782147e-06,
+      "loss": 0.0014,
+      "step": 418
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 2.1212968734537813e-06,
+      "loss": 0.0042,
+      "step": 419
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 0.0056,
+      "step": 420
+    },
+    {
+      "epoch": 7.02,
+      "learning_rate": 2.0782151580276467e-06,
+      "loss": 0.0116,
+      "step": 421
+    },
+    {
+      "epoch": 7.03,
+      "learning_rate": 2.0567965925141366e-06,
+      "loss": 0.0018,
+      "step": 422
+    },
+    {
+      "epoch": 7.05,
+      "learning_rate": 2.0354603547267985e-06,
+      "loss": 0.0055,
+      "step": 423
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.0021,
+      "step": 424
+    },
+    {
+      "epoch": 7.08,
+      "learning_rate": 1.9930372472928095e-06,
+      "loss": 0.0012,
+      "step": 425
+    },
+    {
+      "epoch": 7.1,
+      "learning_rate": 1.971951564311668e-06,
+      "loss": 0.0031,
+      "step": 426
+    },
+    {
+      "epoch": 7.12,
+      "learning_rate": 1.9509505823545232e-06,
+      "loss": 0.0063,
+      "step": 427
+    },
+    {
+      "epoch": 7.13,
+      "learning_rate": 1.9300348888639915e-06,
+      "loss": 0.0022,
+      "step": 428
+    },
+    {
+      "epoch": 7.15,
+      "learning_rate": 1.9092050688969736e-06,
+      "loss": 0.0017,
+      "step": 429
+    },
+    {
+      "epoch": 7.17,
+      "learning_rate": 1.8884617051083183e-06,
+      "loss": 0.0007,
+      "step": 430
+    },
+    {
+      "epoch": 7.18,
+      "learning_rate": 1.867805377734494e-06,
+      "loss": 0.0084,
+      "step": 431
+    },
+    {
+      "epoch": 7.2,
+      "learning_rate": 1.8472366645773892e-06,
+      "loss": 0.0025,
+      "step": 432
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.826756140988119e-06,
+      "loss": 0.0062,
+      "step": 433
+    },
+    {
+      "epoch": 7.23,
+      "learning_rate": 1.8063643798509594e-06,
+      "loss": 0.0026,
+      "step": 434
+    },
+    {
+      "epoch": 7.25,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.0016,
+      "step": 435
+    },
+    {
+      "epoch": 7.27,
+      "learning_rate": 1.7658494240397127e-06,
+      "loss": 0.0135,
+      "step": 436
+    },
+    {
+      "epoch": 7.28,
+      "learning_rate": 1.7457273626560328e-06,
+      "loss": 0.0012,
+      "step": 437
+    },
+    {
+      "epoch": 7.3,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.0031,
+      "step": 438
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 1.7057568872033758e-06,
+      "loss": 0.0018,
+      "step": 439
+    },
+    {
+      "epoch": 7.33,
+      "learning_rate": 1.68590959119452e-06,
+      "loss": 0.0012,
+      "step": 440
+    },
+    {
+      "epoch": 7.35,
+      "learning_rate": 1.6661549974185426e-06,
+      "loss": 0.0042,
+      "step": 441
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.646493658453896e-06,
+      "loss": 0.0011,
+      "step": 442
+    },
+    {
+      "epoch": 7.38,
+      "learning_rate": 1.626926124270497e-06,
+      "loss": 0.0023,
+      "step": 443
+    },
+    {
+      "epoch": 7.4,
+      "learning_rate": 1.6074529422143398e-06,
+      "loss": 0.002,
+      "step": 444
+    },
+    {
+      "epoch": 7.42,
+      "learning_rate": 1.5880746569921867e-06,
+      "loss": 0.0015,
+      "step": 445
+    },
+    {
+      "epoch": 7.43,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.0009,
+      "step": 446
+    },
+    {
+      "epoch": 7.45,
+      "learning_rate": 1.549604942589441e-06,
+      "loss": 0.0013,
+      "step": 447
+    },
+    {
+      "epoch": 7.47,
+      "learning_rate": 1.5305145894894547e-06,
+      "loss": 0.0026,
+      "step": 448
+    },
+    {
+      "epoch": 7.48,
+      "learning_rate": 1.5115212853545924e-06,
+      "loss": 0.0016,
+      "step": 449
+    },
+    {
+      "epoch": 7.5,
+      "learning_rate": 1.4926255614683931e-06,
+      "loss": 0.0027,
+      "step": 450
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.4738279463848803e-06,
+      "loss": 0.0023,
+      "step": 451
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 1.4551289659137497e-06,
+      "loss": 0.0018,
+      "step": 452
+    },
+    {
+      "epoch": 7.55,
+      "learning_rate": 1.4365291431056871e-06,
+      "loss": 0.0018,
+      "step": 453
+    },
+    {
+      "epoch": 7.57,
+      "learning_rate": 1.4180289982377138e-06,
+      "loss": 0.0175,
+      "step": 454
+    },
+    {
+      "epoch": 7.58,
+      "learning_rate": 1.3996290487986568e-06,
+      "loss": 0.0112,
+      "step": 455
+    },
+    {
+      "epoch": 7.6,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 0.0031,
+      "step": 456
+    },
+    {
+      "epoch": 7.62,
+      "learning_rate": 1.3631317921347564e-06,
+      "loss": 0.0051,
+      "step": 457
+    },
+    {
+      "epoch": 7.63,
+      "learning_rate": 1.345035505816642e-06,
+      "loss": 0.0023,
+      "step": 458
+    },
+    {
+      "epoch": 7.65,
+      "learning_rate": 1.3270414567123342e-06,
+      "loss": 0.008,
+      "step": 459
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.3091501481540676e-06,
+      "loss": 0.0047,
+      "step": 460
+    },
+    {
+      "epoch": 7.68,
+      "learning_rate": 1.2913620806002026e-06,
+      "loss": 0.0023,
+      "step": 461
+    },
+    {
+      "epoch": 7.7,
+      "learning_rate": 1.2736777516212267e-06,
+      "loss": 0.0035,
+      "step": 462
+    },
+    {
+      "epoch": 7.72,
+      "learning_rate": 1.2560976558858372e-06,
+      "loss": 0.0031,
+      "step": 463
+    },
+    {
+      "epoch": 7.73,
+      "learning_rate": 1.238622285147103e-06,
+      "loss": 0.0022,
+      "step": 464
+    },
+    {
+      "epoch": 7.75,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.0028,
+      "step": 465
+    },
+    {
+      "epoch": 7.77,
+      "learning_rate": 1.2039876710112847e-06,
+      "loss": 0.0034,
+      "step": 466
+    },
+    {
+      "epoch": 7.78,
+      "learning_rate": 1.1868293964188099e-06,
+      "loss": 0.002,
+      "step": 467
+    },
+    {
+      "epoch": 7.8,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.0039,
+      "step": 468
+    },
+    {
+      "epoch": 7.82,
+      "learning_rate": 1.152833311940429e-06,
+      "loss": 0.0078,
+      "step": 469
+    },
+    {
+      "epoch": 7.83,
+      "learning_rate": 1.135996452998085e-06,
+      "loss": 0.0021,
+      "step": 470
+    },
+    {
+      "epoch": 7.85,
+      "learning_rate": 1.1192676785412154e-06,
+      "loss": 0.0038,
+      "step": 471
+    },
+    {
+      "epoch": 7.87,
+      "learning_rate": 1.1026474565096068e-06,
+      "loss": 0.003,
+      "step": 472
+    },
+    {
+      "epoch": 7.88,
+      "learning_rate": 1.0861362518065915e-06,
+      "loss": 0.0021,
+      "step": 473
+    },
+    {
+      "epoch": 7.9,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 0.0012,
+      "step": 474
+    },
+    {
+      "epoch": 7.92,
+      "learning_rate": 1.0534427387395391e-06,
+      "loss": 0.0024,
+      "step": 475
+    },
+    {
+      "epoch": 7.93,
+      "learning_rate": 1.0372613448833429e-06,
+      "loss": 0.0033,
+      "step": 476
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 1.0211907973458391e-06,
+      "loss": 0.0013,
+      "step": 477
+    },
+    {
+      "epoch": 7.97,
+      "learning_rate": 1.0052315456547934e-06,
+      "loss": 0.0025,
+      "step": 478
+    },
+    {
+      "epoch": 7.98,
+      "learning_rate": 9.893840362247809e-07,
+      "loss": 0.0033,
+      "step": 479
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 9.73648712344707e-07,
+      "loss": 0.0018,
+      "step": 480
+    },
+    {
+      "epoch": 8.02,
+      "learning_rate": 9.580260141654057e-07,
+      "loss": 0.0072,
+      "step": 481
+    },
+    {
+      "epoch": 8.03,
+      "learning_rate": 9.425163786873292e-07,
+      "loss": 0.0014,
+      "step": 482
+    },
+    {
+      "epoch": 8.05,
+      "learning_rate": 9.271202397483214e-07,
+      "loss": 0.0019,
+      "step": 483
+    },
+    {
+      "epoch": 8.07,
+      "learning_rate": 9.118380280114858e-07,
+      "loss": 0.0021,
+      "step": 484
+    },
+    {
+      "epoch": 8.08,
+      "learning_rate": 8.966701709531344e-07,
+      "loss": 0.0031,
+      "step": 485
+    },
+    {
+      "epoch": 8.1,
+      "learning_rate": 8.816170928508367e-07,
+      "loss": 0.0041,
+      "step": 486
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 8.666792147715447e-07,
+      "loss": 0.0013,
+      "step": 487
+    },
+    {
+      "epoch": 8.13,
+      "learning_rate": 8.518569545598198e-07,
+      "loss": 0.002,
+      "step": 488
+    },
+    {
+      "epoch": 8.15,
+      "learning_rate": 8.371507268261436e-07,
+      "loss": 0.004,
+      "step": 489
+    },
+    {
+      "epoch": 8.17,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.0023,
+      "step": 490
+    },
+    {
+      "epoch": 8.18,
+      "learning_rate": 8.08088010994963e-07,
+      "loss": 0.001,
+      "step": 491
+    },
+    {
+      "epoch": 8.2,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.0025,
+      "step": 492
+    },
+    {
+      "epoch": 8.22,
+      "learning_rate": 7.79494319041808e-07,
+      "loss": 0.0025,
+      "step": 493
+    },
+    {
+      "epoch": 8.23,
+      "learning_rate": 7.653743588560387e-07,
+      "loss": 0.0041,
+      "step": 494
+    },
+    {
+      "epoch": 8.25,
+      "learning_rate": 7.513728502524286e-07,
+      "loss": 0.0055,
+      "step": 495
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 7.374901848832683e-07,
+      "loss": 0.0079,
+      "step": 496
+    },
+    {
+      "epoch": 8.28,
+      "learning_rate": 7.237267510765549e-07,
+      "loss": 0.0023,
+      "step": 497
+    },
+    {
+      "epoch": 8.3,
+      "learning_rate": 7.100829338251147e-07,
+      "loss": 0.0064,
+      "step": 498
+    },
+    {
+      "epoch": 8.32,
+      "learning_rate": 6.965591147758482e-07,
+      "loss": 0.0025,
+      "step": 499
+    },
+    {
+      "epoch": 8.33,
+      "learning_rate": 6.831556722190453e-07,
+      "loss": 0.0017,
+      "step": 500
+    },
+    {
+      "epoch": 8.35,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.0019,
+      "step": 501
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 6.567114128975571e-07,
+      "loss": 0.0015,
+      "step": 502
+    },
+    {
+      "epoch": 8.38,
+      "learning_rate": 6.436713358356506e-07,
+      "loss": 0.0029,
+      "step": 503
+    },
+    {
+      "epoch": 8.4,
+      "learning_rate": 6.307531146510754e-07,
+      "loss": 0.0025,
+      "step": 504
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 6.179571106942466e-07,
+      "loss": 0.0141,
+      "step": 505
+    },
+    {
+      "epoch": 8.43,
+      "learning_rate": 6.052836818969027e-07,
+      "loss": 0.0025,
+      "step": 506
+    },
+    {
+      "epoch": 8.45,
+      "learning_rate": 5.927331827620902e-07,
+      "loss": 0.0033,
+      "step": 507
+    },
+    {
+      "epoch": 8.47,
+      "learning_rate": 5.803059643542491e-07,
+      "loss": 0.0034,
+      "step": 508
+    },
+    {
+      "epoch": 8.48,
+      "learning_rate": 5.680023742893926e-07,
+      "loss": 0.0023,
+      "step": 509
+    },
+    {
+      "epoch": 8.5,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 0.0009,
+      "step": 510
+    },
+    {
+      "epoch": 8.52,
+      "learning_rate": 5.437674523523056e-07,
+      "loss": 0.0025,
+      "step": 511
+    },
+    {
+      "epoch": 8.53,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.0021,
+      "step": 512
+    },
+    {
+      "epoch": 8.55,
+      "learning_rate": 5.200311285433213e-07,
+      "loss": 0.0028,
+      "step": 513
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 5.083507730634152e-07,
+      "loss": 0.0025,
+      "step": 514
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 4.967960586678722e-07,
+      "loss": 0.0024,
+      "step": 515
+    },
+    {
+      "epoch": 8.6,
+      "learning_rate": 4.853673085668947e-07,
+      "loss": 0.0022,
+      "step": 516
+    },
+    {
+      "epoch": 8.62,
+      "learning_rate": 4.740648424471872e-07,
+      "loss": 0.0024,
+      "step": 517
+    },
+    {
+      "epoch": 8.63,
+      "learning_rate": 4.628889764630279e-07,
+      "loss": 0.0021,
+      "step": 518
+    },
+    {
+      "epoch": 8.65,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.0037,
+      "step": 519
+    },
+    {
+      "epoch": 8.67,
+      "learning_rate": 4.4091829180330503e-07,
+      "loss": 0.0019,
+      "step": 520
+    },
+    {
+      "epoch": 8.68,
+      "learning_rate": 4.3012408769502134e-07,
+      "loss": 0.0109,
+      "step": 521
+    },
+    {
+      "epoch": 8.7,
+      "learning_rate": 4.194577128396521e-07,
+      "loss": 0.0014,
+      "step": 522
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 4.089194655986306e-07,
+      "loss": 0.0013,
+      "step": 523
+    },
+    {
+      "epoch": 8.73,
+      "learning_rate": 3.985096407493838e-07,
+      "loss": 0.0041,
+      "step": 524
+    },
+    {
+      "epoch": 8.75,
+      "learning_rate": 3.882285294770938e-07,
+      "loss": 0.0097,
+      "step": 525
+    },
+    {
+      "epoch": 8.77,
+      "learning_rate": 3.7807641936653984e-07,
+      "loss": 0.0021,
+      "step": 526
+    },
+    {
+      "epoch": 8.78,
+      "learning_rate": 3.680535943940688e-07,
+      "loss": 0.0036,
+      "step": 527
+    },
+    {
+      "epoch": 8.8,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 0.0017,
+      "step": 528
+    },
+    {
+      "epoch": 8.82,
+      "learning_rate": 3.48396917678982e-07,
+      "loss": 0.0015,
+      "step": 529
+    },
+    {
+      "epoch": 8.83,
+      "learning_rate": 3.3876361577587115e-07,
+      "loss": 0.0033,
+      "step": 530
+    },
+    {
+      "epoch": 8.85,
+      "learning_rate": 3.2926069867446673e-07,
+      "loss": 0.0055,
+      "step": 531
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 3.1988843219178776e-07,
+      "loss": 0.0011,
+      "step": 532
+    },
+    {
+      "epoch": 8.88,
+      "learning_rate": 3.106470784902754e-07,
+      "loss": 0.0036,
+      "step": 533
+    },
+    {
+      "epoch": 8.9,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0015,
+      "step": 534
+    },
+    {
+      "epoch": 8.92,
+      "learning_rate": 2.9255813976372227e-07,
+      "loss": 0.0007,
+      "step": 535
+    },
+    {
+      "epoch": 8.93,
+      "learning_rate": 2.8371106072518194e-07,
+      "loss": 0.0014,
+      "step": 536
+    },
+    {
+      "epoch": 8.95,
+      "learning_rate": 2.7499590642665773e-07,
+      "loss": 0.0025,
+      "step": 537
+    },
+    {
+      "epoch": 8.97,
+      "learning_rate": 2.664129206497479e-07,
+      "loss": 0.0042,
+      "step": 538
+    },
+    {
+      "epoch": 8.98,
+      "learning_rate": 2.579623434790174e-07,
+      "loss": 0.0012,
+      "step": 539
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2.4964441129527337e-07,
+      "loss": 0.0014,
+      "step": 540
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 2.4145935676896106e-07,
+      "loss": 0.0021,
+      "step": 541
+    },
+    {
+      "epoch": 9.03,
+      "learning_rate": 2.3340740885364922e-07,
+      "loss": 0.0039,
+      "step": 542
+    },
+    {
+      "epoch": 9.05,
+      "learning_rate": 2.2548879277963065e-07,
+      "loss": 0.0018,
+      "step": 543
+    },
+    {
+      "epoch": 9.07,
+      "learning_rate": 2.1770373004762035e-07,
+      "loss": 0.0026,
+      "step": 544
+    },
+    {
+      "epoch": 9.08,
+      "learning_rate": 2.1005243842255552e-07,
+      "loss": 0.001,
+      "step": 545
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.0052,
+      "step": 546
+    },
+    {
+      "epoch": 9.12,
+      "learning_rate": 1.9515202083771388e-07,
+      "loss": 0.0019,
+      "step": 547
+    },
+    {
+      "epoch": 9.13,
+      "learning_rate": 1.8790331167464758e-07,
+      "loss": 0.0067,
+      "step": 548
+    },
+    {
+      "epoch": 9.15,
+      "learning_rate": 1.807892072002898e-07,
+      "loss": 0.0031,
+      "step": 549
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 1.738099064114368e-07,
+      "loss": 0.0025,
+      "step": 550
+    },
+    {
+      "epoch": 9.18,
+      "learning_rate": 1.6696560453413446e-07,
+      "loss": 0.0015,
+      "step": 551
+    },
+    {
+      "epoch": 9.2,
+      "learning_rate": 1.6025649301821877e-07,
+      "loss": 0.0021,
+      "step": 552
+    },
+    {
+      "epoch": 9.22,
+      "learning_rate": 1.5368275953196177e-07,
+      "loss": 0.0019,
+      "step": 553
+    },
+    {
+      "epoch": 9.23,
+      "learning_rate": 1.4724458795681962e-07,
+      "loss": 0.0007,
+      "step": 554
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 1.4094215838229176e-07,
+      "loss": 0.0033,
+      "step": 555
+    },
+    {
+      "epoch": 9.27,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.0027,
+      "step": 556
+    },
+    {
+      "epoch": 9.28,
+      "learning_rate": 1.2874522660316412e-07,
+      "loss": 0.0018,
+      "step": 557
+    },
+    {
+      "epoch": 9.3,
+      "learning_rate": 1.2285106557296479e-07,
+      "loss": 0.0016,
+      "step": 558
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 1.1709332888263869e-07,
+      "loss": 0.0033,
+      "step": 559
+    },
+    {
+      "epoch": 9.33,
+      "learning_rate": 1.1147217758845752e-07,
+      "loss": 0.0014,
+      "step": 560
+    },
+    {
+      "epoch": 9.35,
+      "learning_rate": 1.0598776892610685e-07,
+      "loss": 0.0022,
+      "step": 561
+    },
+    {
+      "epoch": 9.37,
+      "learning_rate": 1.0064025630628583e-07,
+      "loss": 0.0026,
+      "step": 562
+    },
+    {
+      "epoch": 9.38,
+      "learning_rate": 9.542978931041835e-08,
+      "loss": 0.0022,
+      "step": 563
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 0.0014,
+      "step": 564
+    },
+    {
+      "epoch": 9.42,
+      "learning_rate": 8.542057134485638e-08,
+      "loss": 0.0006,
+      "step": 565
+    },
+    {
+      "epoch": 9.43,
+      "learning_rate": 8.06221003545038e-08,
+      "loss": 0.0028,
+      "step": 566
+    },
+    {
+      "epoch": 9.45,
+      "learning_rate": 7.59612349389599e-08,
+      "loss": 0.0019,
+      "step": 567
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 7.143810547264762e-08,
+      "loss": 0.0008,
+      "step": 568
+    },
+    {
+      "epoch": 9.48,
+      "learning_rate": 6.705283847722288e-08,
+      "loss": 0.0031,
+      "step": 569
+    },
+    {
+      "epoch": 9.5,
+      "learning_rate": 6.280555661802857e-08,
+      "loss": 0.0052,
+      "step": 570
+    },
+    {
+      "epoch": 9.52,
+      "learning_rate": 5.869637870067002e-08,
+      "loss": 0.0024,
+      "step": 571
+    },
+    {
+      "epoch": 9.53,
+      "learning_rate": 5.472541966768552e-08,
+      "loss": 0.0023,
+      "step": 572
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.0031,
+      "step": 573
+    },
+    {
+      "epoch": 9.57,
+      "learning_rate": 4.719859869049659e-08,
+      "loss": 0.0046,
+      "step": 574
+    },
+    {
+      "epoch": 9.58,
+      "learning_rate": 4.3642947287654284e-08,
+      "loss": 0.0019,
+      "step": 575
+    },
+    {
+      "epoch": 9.6,
+      "learning_rate": 4.02259358460233e-08,
+      "loss": 0.001,
+      "step": 576
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 3.694765994675886e-08,
+      "loss": 0.0022,
+      "step": 577
+    },
+    {
+      "epoch": 9.63,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.0043,
+      "step": 578
+    },
+    {
+      "epoch": 9.65,
+      "learning_rate": 3.080767769372939e-08,
+      "loss": 0.0023,
+      "step": 579
+    },
+    {
+      "epoch": 9.67,
+      "learning_rate": 2.7946143088466437e-08,
+      "loss": 0.0062,
+      "step": 580
+    },
+    {
+      "epoch": 9.68,
+      "learning_rate": 2.5223687517770823e-08,
+      "loss": 0.0013,
+      "step": 581
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 0.0124,
+      "step": 582
+    },
+    {
+      "epoch": 9.72,
+      "learning_rate": 2.0196314199349977e-08,
+      "loss": 0.0076,
+      "step": 583
+    },
+    {
+      "epoch": 9.73,
+      "learning_rate": 1.789153707806357e-08,
+      "loss": 0.0029,
+      "step": 584
+    },
+    {
+      "epoch": 9.75,
+      "learning_rate": 1.5726120240288632e-08,
+      "loss": 0.0023,
+      "step": 585
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 1.3700124257388092e-08,
+      "loss": 0.0011,
+      "step": 586
+    },
+    {
+      "epoch": 9.78,
+      "learning_rate": 1.1813605800825578e-08,
+      "loss": 0.003,
+      "step": 587
+    },
+    {
+      "epoch": 9.8,
+      "learning_rate": 1.006661764057837e-08,
+      "loss": 0.0048,
+      "step": 588
+    },
+    {
+      "epoch": 9.82,
+      "learning_rate": 8.459208643659122e-09,
+      "loss": 0.0019,
+      "step": 589
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 6.991423772753636e-09,
+      "loss": 0.0013,
+      "step": 590
+    },
+    {
+      "epoch": 9.85,
+      "learning_rate": 5.6633040849601865e-09,
+      "loss": 0.002,
+      "step": 591
+    },
+    {
+      "epoch": 9.87,
+      "learning_rate": 4.474886730641004e-09,
+      "loss": 0.0048,
+      "step": 592
+    },
+    {
+      "epoch": 9.88,
+      "learning_rate": 3.4262049523847707e-09,
+      "loss": 0.0015,
+      "step": 593
+    },
+    {
+      "epoch": 9.9,
+      "learning_rate": 2.5172880840745873e-09,
+      "loss": 0.014,
+      "step": 594
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 1.7481615500691829e-09,
+      "loss": 0.0058,
+      "step": 595
+    },
+    {
+      "epoch": 9.93,
+      "learning_rate": 1.118846864490708e-09,
+      "loss": 0.0012,
+      "step": 596
+    },
+    {
+      "epoch": 9.95,
+      "learning_rate": 6.293616306246586e-10,
+      "loss": 0.0031,
+      "step": 597
+    },
+    {
+      "epoch": 9.97,
+      "learning_rate": 2.797195404247166e-10,
+      "loss": 0.0031,
+      "step": 598
+    },
+    {
+      "epoch": 9.98,
+      "learning_rate": 6.993037413194348e-11,
+      "loss": 0.0009,
+      "step": 599
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 0.0,
+      "loss": 0.0016,
+      "step": 600
+    },
+    {
+      "epoch": 10.0,
+      "step": 600,
+      "total_flos": 3.37187743727616e+16,
+      "train_loss": 0.051574386974486214,
+      "train_runtime": 1007.7691,
+      "train_samples_per_second": 1.191,
+      "train_steps_per_second": 0.595
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 1000,
+  "total_flos": 3.37187743727616e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}

training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e732eb00beed04591f31644de882a12cc289c18475cc0879a46b18e03481ff36
+size 4728