Add application file

LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

app.py

@@ -0,0 +1,155 @@
+import streamlit as st
+
+import io
+import os
+import yaml
+import pyarrow
+import tokenizers
+
+
+os.environ["TOKENIZERS_PARALLELISM"] = "true"
+
+# SETTING PAGE CONFIG TO WIDE MODE
+st.set_page_config(layout="wide")
+
+@st.cache
+def from_library():
+    from retro_reader import RetroReader
+    from retro_reader import constants as C
+    return C, RetroReader
+
+C, RetroReader = from_library()
+
+# https://stackoverflow.com/questions/70274841/streamlit-unhashable-typeerror-when-i-use-st-cache
+my_hash_func = {
+    io.TextIOWrapper: lambda _: None,
+    pyarrow.lib.Buffer: lambda _: 0,
+    tokenizers.Tokenizer: lambda _: None,
+    tokenizers.AddedToken: lambda _: None
+}
+
+# @st.cache(hash_funcs=my_hash_func, allow_output_mutation=True)
+# def load_ko_roberta_large_model():
+#     config_file = "configs/inference_ko_roberta_large.yaml"
+#     return RetroReader.load(config_file=config_file)
+
+
+# @st.cache(hash_funcs=my_hash_func, allow_output_mutation=True)
+# def load_ko_electra_small_model():
+#     config_file = "configs/inference_ko_electra_small.yaml"
+#     return RetroReader.load(config_file=config_file)
+
+
+# @st.cache(hash_funcs=my_hash_func, allow_output_mutation=True)
+# def load_en_electra_large_model():
+#     config_file = "configs/inference_en_electra_large.yaml"
+#     return RetroReader.load(config_file=config_file)
+
+@st.cache(hash_funcs=my_hash_func, allow_output_mutation=True)
+def load_vi_electra_base_model():
+    config_file = "configs/inference_vi_electra_base.yaml"
+    return RetroReader.load(config_file=config_file)
+
+RETRO_READER_HOST = {
+    # "klue/roberta-large": load_ko_roberta_large_model(),
+    # "monologg/koelectra-small-v3-discriminator": load_ko_electra_small_model(),
+    "google/electra-large-discriminator": load_vi_electra_base_model(),
+}
+
+
+def main():
+    st.title("Retrospective Reader Demo")
+    
+    # st.markdown("## Model name")
+    # option = st.selectbox(
+    #     label="Choose the model used in retro reader",
+    #     options=(
+    #         # "[ko_KR] klue/roberta-large",
+    #         # "[ko_KR] monologg/koelectra-small-v3-discriminator",
+    #         "[vi_XX] google/electra-large-discriminator",
+    #     ),
+    #     index=0,
+    # )
+    # lang_code, model_name = option.split(" ")
+    
+    retro_reader = load_vi_electra_base_model()
+    
+    # retro_reader = load_model()
+    lang_prefix = "EN"
+    height = 300
+    
+    # retro_reader.null_score_diff_threshold = st.sidebar.slider(
+    #     label="null_score_diff_threshold",
+    #     min_value=-10.0, max_value=10.0, value=0.0, step=1.0,
+    #     help="ma!",
+    # )
+    # retro_reader.rear_threshold = st.sidebar.slider(
+    #     label="rear_threshold",
+    #     min_value=-10.0, max_value=10.0, value=0.0, step=1.0,
+    #     help="ma!",
+    # )
+    # retro_reader.n_best_size = st.sidebar.slider(
+    #     label="n_best_size",
+    #     min_value=1, max_value=50, value=20, step=1,
+    #     help="ma!",
+    # )
+    # retro_reader.beta1 = st.sidebar.slider(
+    #     label="beta1",
+    #     min_value=-10.0, max_value=10.0, value=1.0, step=1.0,
+    #     help="ma!",
+    # )
+    # retro_reader.beta2 = st.sidebar.slider(
+    #     label="beta2",
+    #     min_value=-10.0, max_value=10.0, value=1.0, step=1.0,
+    #     help="ma!",
+    # )
+    # retro_reader.best_cof = st.sidebar.slider(
+    #     label="best_cof",
+    #     min_value=-10.0, max_value=10.0, value=1.0, step=1.0,
+    #     help="ma!",
+    # )
+    # return_submodule_outputs = st.sidebar.checkbox('return_submodule_outputs', value=False)
+    return_submodule_outputs = False
+    st.markdown("## Demonstration")
+    with st.form(key="my_form"):
+        query = st.text_input(
+            label="Type your query",
+            value=getattr(C, f"{lang_prefix}_EXAMPLE_QUERY"),
+            max_chars=None,
+            help=getattr(C, f"{lang_prefix}_QUERY_HELP_TEXT"),
+        )
+        context = st.text_area(
+            label="Type your context",
+            value=getattr(C, f"{lang_prefix}_EXAMPLE_CONTEXTS"),
+            height=height,
+            max_chars=None,
+            help=getattr(C, f"{lang_prefix}_CONTEXT_HELP_TEXT"),
+        )
+        submit_button = st.form_submit_button(label="Submit")
+        
+    if submit_button:
+        with st.spinner("Please wait.."):
+            outputs = retro_reader(
+                query=query,
+                context=context,
+                return_submodule_outputs=return_submodule_outputs,
+            )
+        answer, score = outputs[0]["id-01"], outputs[1]
+        if not answer:
+            answer = "No answer"
+        st.markdown("## Results")
+        st.write(answer)
+        st.markdown("### Rear Verification Score")
+        st.json(score)
+        # if return_submodule_outputs:
+        #     score_ext, nbest_preds, score_diff = outputs[2:]
+        #     st.markdown("### Sketch Reader Score (score_ext)")
+        #     st.json(score_ext)
+        #     st.markdown("### Intensive Reader Score (score_diff)")
+        #     st.json(score_diff)
+        #     st.markdown("### N Best Predictions (from intensive reader)")
+        #     st.json(nbest_preds)
+    
+
+if __name__ == "__main__":
+    main()

configs/inference_vi_electra_base.yaml

@@ -0,0 +1,41 @@
+RetroDataModelArguments:
+
+    # DataArguments
+    max_seq_length: 512
+    max_answer_length: 30
+    doc_stride: 128
+    return_token_type_ids: True
+    pad_to_max_length: True
+    preprocessing_num_workers: 5
+    overwrite_cache: False
+    version_2_with_negative: True
+    null_score_diff_threshold: 0.0
+    rear_threshold: 0.0
+    n_best_size: 20
+    use_choice_logits: False
+    start_n_top: -1
+    end_n_top: -1
+    beta1: 1
+    beta2: 1
+    best_cof: 1
+    
+    # ModelArguments
+    use_auth_token: False
+    
+    # SketchModelArguments
+    sketch_revision: en-electra-large-sketch
+    sketch_model_name: ./outputs/sketch/checkpoint-23900/
+    sketch_architectures: ElectraForSequenceClassification
+    
+    # IntensiveModelArguments
+    intensive_revision: en-electra-largs-intensive
+    intensive_model_name: ./outputs/intensive/checkpoint-23900/
+    intensive_architectures: ElectraForQuestionAnsweringAVPool
+    
+
+TrainingArguments:
+    output_dir: outputs
+    no_cuda: True # If you want to use cuda,
+                  # change `no_cuda` to False and `fp16` to True
+    per_device_train_batch_size: 1
+    per_device_eval_batch_size: 1

configs/train_vi_electra_base.yaml

@@ -0,0 +1,49 @@
+RetroDataModelArguments:
+
+    # DataArguments
+    max_seq_length: 512
+    max_answer_length: 30
+    doc_stride: 128
+    return_token_type_ids: True
+    pad_to_max_length: True
+    preprocessing_num_workers: 5
+    overwrite_cache: False
+    version_2_with_negative: True
+    null_score_diff_threshold: 0.0
+    rear_threshold: 0.0
+    n_best_size: 20
+    use_choice_logits: False
+    start_n_top: -1
+    end_n_top: -1
+    beta1: 1
+    beta2: 1
+    best_cof: 1
+    
+    # SketchModelArguments
+    sketch_model_name: NlpHUST/electra-base-vn
+    sketch_architectures: ElectraForSequenceClassification
+    
+    # IntensiveModelArguments
+    intensive_model_name: NlpHUST/electra-base-vn
+    intensive_architectures: ElectraForQuestionAnsweringAVPool
+    
+
+TrainingArguments:
+    report_to: wandb
+    run_name: squadv2-electra-large-sketch,squadv2-electra-large-intensive
+    output_dir: outputs
+    overwrite_output_dir: False
+    learning_rate: 2e-5
+    evaluation_strategy: epoch
+    save_strategy: epoch
+    per_device_train_batch_size: 12
+    per_device_eval_batch_size: 12
+    num_train_epochs: 10.0
+    # save_steps: 5000
+    save_total_limit: 2
+    fp16: True
+    warmup_ratio: 0.1
+    weight_decay: 0.01
+    load_best_model_at_end: True
+    metric_for_best_model: f1,exact
+    logging_dir: logs

outputs/intensive/checkpoint-23900/config.json

@@ -0,0 +1,30 @@
+{
+  "_name_or_path": "NlpHUST/electra-base-vn",
+  "architectures": [
+    "ElectraForQuestionAnsweringAVPool"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "embedding_size": 768,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.21.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 62000
+}

outputs/intensive/checkpoint-23900/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e31e5f953e409735eb91c55c9c865412dd75d45a99a6a4a0b156b7ae7ac2be2e
+size 532350957

outputs/intensive/checkpoint-23900/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3f746b0186296f2263ff69d55705b37e5ffd3f32456e0e399cf3c6b97a0f7d4f
+size 14503

outputs/intensive/checkpoint-23900/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5538239ac364de835557402682c0e056dcc1e97f5ee87f4a2958b61773eb625d
+size 559

outputs/intensive/checkpoint-23900/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d3d346564087658131fb56d711b982ef1a8a44b9b23d2ae1417d098510e4849
+size 623

outputs/intensive/checkpoint-23900/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

outputs/intensive/checkpoint-23900/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "name_or_path": "NlpHUST/electra-base-vn",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "ElectraTokenizer",
+  "unk_token": "[UNK]"
+}

outputs/intensive/checkpoint-23900/trainer_state.json

@@ -0,0 +1,468 @@
+{
+  "best_metric": 60.008646779074795,
+  "best_model_checkpoint": "outputs/intensive/checkpoint-23900",
+  "epoch": 10.0,
+  "global_step": 23900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.150627615062761e-06,
+      "loss": 10.7765,
+      "step": 500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 8.334728033472804e-06,
+      "loss": 6.9262,
+      "step": 1000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2518828451882846e-05,
+      "loss": 5.081,
+      "step": 1500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.670292887029289e-05,
+      "loss": 4.5427,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_HasAns_exact": 34.392563769995675,
+      "eval_HasAns_f1": 49.07185024538246,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 37.602680501513184,
+      "eval_best_exact_thresh": 6.36328125,
+      "eval_best_f1": 52.28196697689995,
+      "eval_best_f1_thresh": 6.36328125,
+      "eval_exact": 34.392563769995675,
+      "eval_f1": 49.07185024538246,
+      "eval_runtime": 157.3533,
+      "eval_samples_per_second": 59.293,
+      "eval_steps_per_second": 4.944,
+      "eval_total": 9252,
+      "step": 2390
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.99023709902371e-05,
+      "loss": 4.2046,
+      "step": 2500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9437470943747097e-05,
+      "loss": 3.8663,
+      "step": 3000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.897257089725709e-05,
+      "loss": 3.7485,
+      "step": 3500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.850767085076709e-05,
+      "loss": 3.6324,
+      "step": 4000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.804370060437006e-05,
+      "loss": 3.5147,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_HasAns_exact": 44.32555123216602,
+      "eval_HasAns_f1": 58.396422080934855,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 50.72416774751405,
+      "eval_best_exact_thresh": 6.75,
+      "eval_best_f1": 64.79503859628278,
+      "eval_best_f1_thresh": 6.75,
+      "eval_exact": 44.32555123216602,
+      "eval_f1": 58.396422080934855,
+      "eval_runtime": 158.1216,
+      "eval_samples_per_second": 59.005,
+      "eval_steps_per_second": 4.92,
+      "eval_total": 9252,
+      "step": 4780
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.7579730357973036e-05,
+      "loss": 3.2078,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.711576011157601e-05,
+      "loss": 2.7938,
+      "step": 5500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.665086006508601e-05,
+      "loss": 2.781,
+      "step": 6000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6185960018596003e-05,
+      "loss": 2.7884,
+      "step": 6500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.5721059972105997e-05,
+      "loss": 2.6837,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0,
+      "eval_HasAns_exact": 49.48119325551232,
+      "eval_HasAns_f1": 61.58352384463478,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 66.6126242974492,
+      "eval_best_exact_thresh": 13.5234375,
+      "eval_best_f1": 78.71495488657156,
+      "eval_best_f1_thresh": 13.5234375,
+      "eval_exact": 49.48119325551232,
+      "eval_f1": 61.58352384463478,
+      "eval_runtime": 161.8926,
+      "eval_samples_per_second": 57.631,
+      "eval_steps_per_second": 4.806,
+      "eval_total": 9252,
+      "step": 7170
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.5256159925615994e-05,
+      "loss": 2.2675,
+      "step": 7500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.479125987912599e-05,
+      "loss": 2.0381,
+      "step": 8000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4326359832635986e-05,
+      "loss": 2.0343,
+      "step": 8500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.3861459786145978e-05,
+      "loss": 2.0476,
+      "step": 9000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.3396559739655974e-05,
+      "loss": 2.0245,
+      "step": 9500
+    },
+    {
+      "epoch": 4.0,
+      "eval_HasAns_exact": 51.14569822741029,
+      "eval_HasAns_f1": 61.58064803440931,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 77.45352356247298,
+      "eval_best_exact_thresh": 17.703125,
+      "eval_best_f1": 87.88847336947204,
+      "eval_best_f1_thresh": 17.703125,
+      "eval_exact": 51.14569822741029,
+      "eval_f1": 61.58064803440931,
+      "eval_runtime": 159.5263,
+      "eval_samples_per_second": 58.486,
+      "eval_steps_per_second": 4.877,
+      "eval_total": 9252,
+      "step": 9560
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.293165969316597e-05,
+      "loss": 1.4757,
+      "step": 10000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.2466759646675965e-05,
+      "loss": 1.4193,
+      "step": 10500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.200278940027894e-05,
+      "loss": 1.4397,
+      "step": 11000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.1537889353788937e-05,
+      "loss": 1.4482,
+      "step": 11500
+    },
+    {
+      "epoch": 5.0,
+      "eval_HasAns_exact": 54.31258106355383,
+      "eval_HasAns_f1": 62.505983151723,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 84.15477734543883,
+      "eval_best_exact_thresh": 19.4375,
+      "eval_best_f1": 92.34817943360805,
+      "eval_best_f1_thresh": 19.4375,
+      "eval_exact": 54.31258106355383,
+      "eval_f1": 62.505983151723,
+      "eval_runtime": 110.395,
+      "eval_samples_per_second": 84.515,
+      "eval_steps_per_second": 7.047,
+      "eval_total": 9252,
+      "step": 11950
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.107298930729893e-05,
+      "loss": 1.4318,
+      "step": 12000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1.0608089260808926e-05,
+      "loss": 1.0225,
+      "step": 12500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.0143189214318922e-05,
+      "loss": 1.0562,
+      "step": 13000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 9.679218967921897e-06,
+      "loss": 1.0755,
+      "step": 13500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 9.214318921431893e-06,
+      "loss": 1.0455,
+      "step": 14000
+    },
+    {
+      "epoch": 6.0,
+      "eval_HasAns_exact": 56.91742325983571,
+      "eval_HasAns_f1": 63.24948077893548,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 87.2568093385214,
+      "eval_best_exact_thresh": 25.234375,
+      "eval_best_f1": 93.58886685762127,
+      "eval_best_f1_thresh": 25.234375,
+      "eval_exact": 56.91742325983571,
+      "eval_f1": 63.24948077893548,
+      "eval_runtime": 109.5824,
+      "eval_samples_per_second": 85.141,
+      "eval_steps_per_second": 7.1,
+      "eval_total": 9252,
+      "step": 14340
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 8.749418874941889e-06,
+      "loss": 0.9775,
+      "step": 14500
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 8.284518828451885e-06,
+      "loss": 0.7959,
+      "step": 15000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 7.820548582054858e-06,
+      "loss": 0.7763,
+      "step": 15500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 7.3565783356578345e-06,
+      "loss": 0.841,
+      "step": 16000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.891678289167829e-06,
+      "loss": 0.8104,
+      "step": 16500
+    },
+    {
+      "epoch": 7.0,
+      "eval_HasAns_exact": 58.36575875486381,
+      "eval_HasAns_f1": 63.35313006280799,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 89.95892779939473,
+      "eval_best_exact_thresh": 28.4375,
+      "eval_best_f1": 94.9462991073391,
+      "eval_best_f1_thresh": 28.4375,
+      "eval_exact": 58.36575875486381,
+      "eval_f1": 63.35313006280799,
+      "eval_runtime": 111.4095,
+      "eval_samples_per_second": 83.745,
+      "eval_steps_per_second": 6.983,
+      "eval_total": 9252,
+      "step": 16730
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 6.426778242677825e-06,
+      "loss": 0.6864,
+      "step": 17000
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5.96187819618782e-06,
+      "loss": 0.6432,
+      "step": 17500
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 5.496978149697816e-06,
+      "loss": 0.6308,
+      "step": 18000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 5.032078103207811e-06,
+      "loss": 0.6595,
+      "step": 18500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 4.567178056717806e-06,
+      "loss": 0.6227,
+      "step": 19000
+    },
+    {
+      "epoch": 8.0,
+      "eval_HasAns_exact": 59.262862083873756,
+      "eval_HasAns_f1": 63.181652613846246,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 91.74232598357112,
+      "eval_best_exact_thresh": 29.84375,
+      "eval_best_f1": 95.6611165135438,
+      "eval_best_f1_thresh": 29.84375,
+      "eval_exact": 59.262862083873756,
+      "eval_f1": 63.181652613846246,
+      "eval_runtime": 111.1644,
+      "eval_samples_per_second": 83.93,
+      "eval_steps_per_second": 6.999,
+      "eval_total": 9252,
+      "step": 19120
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 4.102278010227801e-06,
+      "loss": 0.5235,
+      "step": 19500
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 3.638307763830777e-06,
+      "loss": 0.4937,
+      "step": 20000
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 3.173407717340772e-06,
+      "loss": 0.5096,
+      "step": 20500
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 2.7085076708507676e-06,
+      "loss": 0.5335,
+      "step": 21000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2.2436076243607625e-06,
+      "loss": 0.5032,
+      "step": 21500
+    },
+    {
+      "epoch": 9.0,
+      "eval_HasAns_exact": 59.814094249891916,
+      "eval_HasAns_f1": 63.04596094446208,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 92.19628188499784,
+      "eval_best_exact_thresh": 30.34375,
+      "eval_best_f1": 95.42814857956823,
+      "eval_best_f1_thresh": 30.34375,
+      "eval_exact": 59.814094249891916,
+      "eval_f1": 63.04596094446208,
+      "eval_runtime": 110.9375,
+      "eval_samples_per_second": 84.101,
+      "eval_steps_per_second": 7.013,
+      "eval_total": 9252,
+      "step": 21510
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 1.778707577870758e-06,
+      "loss": 0.4126,
+      "step": 22000
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.3138075313807533e-06,
+      "loss": 0.4188,
+      "step": 22500
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 8.498372849837285e-07,
+      "loss": 0.4326,
+      "step": 23000
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 3.858670385867039e-07,
+      "loss": 0.4378,
+      "step": 23500
+    },
+    {
+      "epoch": 10.0,
+      "eval_HasAns_exact": 60.008646779074795,
+      "eval_HasAns_f1": 62.92824770736451,
+      "eval_HasAns_total": 9252,
+      "eval_best_exact": 92.7583225248595,
+      "eval_best_exact_thresh": 31.09375,
+      "eval_best_f1": 95.67792345314943,
+      "eval_best_f1_thresh": 31.09375,
+      "eval_exact": 60.008646779074795,
+      "eval_f1": 62.92824770736451,
+      "eval_runtime": 107.3386,
+      "eval_samples_per_second": 86.921,
+      "eval_steps_per_second": 7.248,
+      "eval_total": 9252,
+      "step": 23900
+    }
+  ],
+  "max_steps": 23900,
+  "num_train_epochs": 10,
+  "total_flos": 7.493865187135488e+16,
+  "trial_name": null,
+  "trial_params": null
+}

outputs/intensive/checkpoint-23900/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2a2ea25156a9af14389b6fb124369129c6a6bd78f82c00c33c31d87753dbaa92
+size 3311

outputs/intensive/nbest_predictions.json

@@ -0,0 +1,124 @@
+{
+    "id-01": [
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": 5.49807071685791,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng",
+            "probability": 0.9822916984558105
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": 1.412649154663086,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean)",
+            "probability": 0.01651826500892639
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -1.957545280456543,
+            "text": "Nguy\u1ec5n",
+            "probability": 0.0005679467576555908
+        },
+        {
+            "start_logit": -2.662353992462158,
+            "end_logit": 5.49807071685791,
+            "text": "C\u01b0\u1eddng",
+            "probability": 0.0001623508578632027
+        },
+        {
+            "start_logit": -2.665097713470459,
+            "end_logit": 5.49807071685791,
+            "text": "\u0110\u1ee9c C\u01b0\u1eddng",
+            "probability": 0.00016190586029551923
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -3.320735454559326,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989",
+            "probability": 0.0001453054283047095
+        },
+        {
+            "start_logit": 1.1949851512908936,
+            "end_logit": 0.35255151987075806,
+            "text": "l\u00e0 m\u1ed9t nam rapper v\u00e0 nh\u1ea1c s\u0129 ng\u01b0\u1eddi Vi\u1ec7t Nam",
+            "probability": 4.477184484130703e-05
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -4.638460636138916,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (",
+            "probability": 3.890457082889043e-05
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -5.202346324920654,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c",
+            "probability": 2.2136480765766464e-05
+        },
+        {
+            "start_logit": 0.23849983513355255,
+            "end_logit": 0.35255151987075806,
+            "text": "m\u1ed9t nam rapper v\u00e0 nh\u1ea1c s\u0129 ng\u01b0\u1eddi Vi\u1ec7t Nam",
+            "probability": 1.720317050057929e-05
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -6.049441814422607,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi",
+            "probability": 9.488983778283e-06
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -6.32877254486084,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean",
+            "probability": 7.176417511800537e-06
+        },
+        {
+            "start_logit": 6.045530319213867,
+            "end_logit": -6.484272003173828,
+            "text": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean),",
+            "probability": 6.142924121377291e-06
+        },
+        {
+            "start_logit": -2.662353992462158,
+            "end_logit": 1.412649154663086,
+            "text": "C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean)",
+            "probability": 2.7300973215460544e-06
+        },
+        {
+            "start_logit": -2.665097713470459,
+            "end_logit": 1.412649154663086,
+            "text": "\u0110\u1ee9c C\u01b0\u1eddng (sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean)",
+            "probability": 2.722619228734402e-06
+        },
+        {
+            "start_logit": -1.3381463289260864,
+            "end_logit": -1.3950865268707275,
+            "text": "",
+            "probability": 6.1928426475788e-07
+        },
+        {
+            "start_logit": -4.9331583976745605,
+            "end_logit": 1.412649154663086,
+            "text": "sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean)",
+            "probability": 2.8182577693769417e-07
+        },
+        {
+            "start_logit": -5.086843013763428,
+            "end_logit": 1.412649154663086,
+            "text": ")",
+            "probability": 2.416775828351092e-07
+        },
+        {
+            "start_logit": 1.1949851512908936,
+            "end_logit": -5.277959823608398,
+            "text": "l\u00e0 m\u1ed9t nam rapper",
+            "probability": 1.6058500307281065e-07
+        },
+        {
+            "start_logit": -5.693793296813965,
+            "end_logit": 1.412649154663086,
+            "text": "(sinh ng\u00e0y 13 th\u00e1ng 5 n\u0103m 1989 t\u1ea1i Qu\u1ea3ng Ninh, nh\u01b0ng qu\u00ea g\u1ed1c \u1edf \u00c2n Thi, H\u01b0ng Y\u00ean)",
+            "probability": 1.3171673174383614e-07
+        }
+    ]
+}

outputs/intensive/null_odds.json

@@ -0,0 +1,3 @@
+{
+    "id-01": -14.276833534240723
+}

outputs/intensive/predictions.json

@@ -0,0 +1,3 @@
+{
+    "id-01": "Nguy\u1ec5n \u0110\u1ee9c C\u01b0\u1eddng"
+}

outputs/sketch/checkpoint-23900/config.json

@@ -0,0 +1,31 @@
+{
+  "_name_or_path": "NlpHUST/electra-base-vn",
+  "architectures": [
+    "ElectraForSequenceClassification"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "embedding_size": 768,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 768,
+  "initializer_range": 0.02,
+  "intermediate_size": 3072,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "electra",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 12,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "problem_type": "single_label_classification",
+  "summary_activation": "gelu",
+  "summary_last_dropout": 0.1,
+  "summary_type": "first",
+  "summary_use_proj": true,
+  "torch_dtype": "float32",
+  "transformers_version": "4.21.1",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 62000
+}

outputs/sketch/checkpoint-23900/pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ee5ef0330b58fc9299f1f07ac1f057596d02f2350f66ed80eb8f1479aef56bec
+size 534707117

outputs/sketch/checkpoint-23900/rng_state.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37085a94bb8702116f42aba75288847f204d26e3445d8ef1e587654032735258
+size 14503

outputs/sketch/checkpoint-23900/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0d3fcb2fd6a93bcec231636ca78e55ddc7e0cf35aa3208c3cb78468921200b8d
+size 559

outputs/sketch/checkpoint-23900/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4037b40aa9d735b8382af812ddac621c3546c12b2b539da6400f157fd74b61ae
+size 623

outputs/sketch/checkpoint-23900/special_tokens_map.json

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}

outputs/sketch/checkpoint-23900/tokenizer_config.json

@@ -0,0 +1,15 @@
+{
+  "cls_token": "[CLS]",
+  "do_basic_tokenize": true,
+  "do_lower_case": false,
+  "mask_token": "[MASK]",
+  "name_or_path": "NlpHUST/electra-base-vn",
+  "never_split": null,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "special_tokens_map_file": null,
+  "strip_accents": null,
+  "tokenize_chinese_chars": true,
+  "tokenizer_class": "ElectraTokenizer",
+  "unk_token": "[UNK]"
+}

outputs/sketch/checkpoint-23900/trainer_state.json

@@ -0,0 +1,408 @@
+{
+  "best_metric": 0.9979028875625101,
+  "best_model_checkpoint": "outputs/sketch/checkpoint-23900",
+  "epoch": 10.0,
+  "global_step": 23900,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.184100418410042e-06,
+      "loss": 0.6451,
+      "step": 500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 8.368200836820084e-06,
+      "loss": 0.6334,
+      "step": 1000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.2552301255230125e-05,
+      "loss": 0.6162,
+      "step": 1500
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6719665271966527e-05,
+      "loss": 0.5225,
+      "step": 2000
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8397642015005359,
+      "eval_f1": 0.7340330901974738,
+      "eval_precision": 0.8176773682124455,
+      "eval_recall": 0.6659134925758554,
+      "eval_runtime": 59.1615,
+      "eval_samples_per_second": 157.704,
+      "eval_steps_per_second": 13.15,
+      "step": 2390
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.9899581589958163e-05,
+      "loss": 0.4783,
+      "step": 2500
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.9434681543468157e-05,
+      "loss": 0.4164,
+      "step": 3000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.896978149697815e-05,
+      "loss": 0.4054,
+      "step": 3500
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.8505811250581126e-05,
+      "loss": 0.3954,
+      "step": 4000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.8040911204091124e-05,
+      "loss": 0.4002,
+      "step": 4500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.9234726688102894,
+      "eval_f1": 0.8800403225806451,
+      "eval_precision": 0.9176594253679047,
+      "eval_recall": 0.8453841187863137,
+      "eval_runtime": 106.4194,
+      "eval_samples_per_second": 87.672,
+      "eval_steps_per_second": 7.311,
+      "step": 4780
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.75769409576941e-05,
+      "loss": 0.3342,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.7112040911204093e-05,
+      "loss": 0.2597,
+      "step": 5500
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.6647140864714087e-05,
+      "loss": 0.2773,
+      "step": 6000
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.6182240818224085e-05,
+      "loss": 0.2942,
+      "step": 6500
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.571827057182706e-05,
+      "loss": 0.2799,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.9638799571275456,
+      "eval_f1": 0.9449076344613373,
+      "eval_precision": 0.9572706194104008,
+      "eval_recall": 0.9328599096191091,
+      "eval_runtime": 105.0123,
+      "eval_samples_per_second": 88.847,
+      "eval_steps_per_second": 7.409,
+      "step": 7170
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.5253370525337054e-05,
+      "loss": 0.2081,
+      "step": 7500
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.478847047884705e-05,
+      "loss": 0.202,
+      "step": 8000
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 1.4323570432357044e-05,
+      "loss": 0.1983,
+      "step": 8500
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 1.385960018596002e-05,
+      "loss": 0.2074,
+      "step": 9000
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.3394700139470016e-05,
+      "loss": 0.2099,
+      "step": 9500
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.9785637727759914,
+      "eval_f1": 0.9672988881621974,
+      "eval_precision": 0.9801192842942346,
+      "eval_recall": 0.9548095545513234,
+      "eval_runtime": 142.4236,
+      "eval_samples_per_second": 65.509,
+      "eval_steps_per_second": 5.463,
+      "step": 9560
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.2929800092980009e-05,
+      "loss": 0.1391,
+      "step": 10000
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.2464900046490005e-05,
+      "loss": 0.1301,
+      "step": 10500
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.2000929800092982e-05,
+      "loss": 0.1409,
+      "step": 11000
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.1536029753602976e-05,
+      "loss": 0.1544,
+      "step": 11500
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.9836012861736334,
+      "eval_f1": 0.9755552005112638,
+      "eval_precision": 0.9658335969629864,
+      "eval_recall": 0.9854744996772111,
+      "eval_runtime": 107.7483,
+      "eval_samples_per_second": 86.591,
+      "eval_steps_per_second": 7.221,
+      "step": 11950
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.1071129707112971e-05,
+      "loss": 0.1549,
+      "step": 12000
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1.0606229660622967e-05,
+      "loss": 0.092,
+      "step": 12500
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.014225941422594e-05,
+      "loss": 0.0973,
+      "step": 13000
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 9.678289167828918e-06,
+      "loss": 0.122,
+      "step": 13500
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 9.213389121338912e-06,
+      "loss": 0.1221,
+      "step": 14000
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.9926045016077171,
+      "eval_f1": 0.9888799355358582,
+      "eval_precision": 0.9874476987447699,
+      "eval_recall": 0.9903163331181407,
+      "eval_runtime": 135.6226,
+      "eval_samples_per_second": 68.794,
+      "eval_steps_per_second": 5.737,
+      "step": 14340
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 8.748489074848908e-06,
+      "loss": 0.1079,
+      "step": 14500
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 8.283589028358903e-06,
+      "loss": 0.0886,
+      "step": 15000
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 7.818688981868899e-06,
+      "loss": 0.0816,
+      "step": 15500
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 7.354718735471874e-06,
+      "loss": 0.0723,
+      "step": 16000
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.889818688981869e-06,
+      "loss": 0.0941,
+      "step": 16500
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.9945337620578778,
+      "eval_f1": 0.9917966865047451,
+      "eval_precision": 0.9884578390509778,
+      "eval_recall": 0.9951581665590704,
+      "eval_runtime": 105.7234,
+      "eval_samples_per_second": 88.249,
+      "eval_steps_per_second": 7.359,
+      "step": 16730
+    },
+    {
+      "epoch": 7.11,
+      "learning_rate": 6.424918642491865e-06,
+      "loss": 0.066,
+      "step": 17000
+    },
+    {
+      "epoch": 7.32,
+      "learning_rate": 5.96001859600186e-06,
+      "loss": 0.0537,
+      "step": 17500
+    },
+    {
+      "epoch": 7.53,
+      "learning_rate": 5.496048349604835e-06,
+      "loss": 0.0484,
+      "step": 18000
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 5.031148303114831e-06,
+      "loss": 0.0544,
+      "step": 18500
+    },
+    {
+      "epoch": 7.95,
+      "learning_rate": 4.566248256624826e-06,
+      "loss": 0.0673,
+      "step": 19000
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.9965702036441586,
+      "eval_f1": 0.9948403740728798,
+      "eval_precision": 0.9938788659793815,
+      "eval_recall": 0.9958037443511943,
+      "eval_runtime": 107.9506,
+      "eval_samples_per_second": 86.428,
+      "eval_steps_per_second": 7.207,
+      "step": 19120
+    },
+    {
+      "epoch": 8.16,
+      "learning_rate": 4.101348210134822e-06,
+      "loss": 0.035,
+      "step": 19500
+    },
+    {
+      "epoch": 8.37,
+      "learning_rate": 3.636448163644817e-06,
+      "loss": 0.0337,
+      "step": 20000
+    },
+    {
+      "epoch": 8.58,
+      "learning_rate": 3.1724779172477922e-06,
+      "loss": 0.0397,
+      "step": 20500
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 2.707577870757787e-06,
+      "loss": 0.0398,
+      "step": 21000
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 2.242677824267783e-06,
+      "loss": 0.0513,
+      "step": 21500
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.9980707395498393,
+      "eval_f1": 0.9970939618986115,
+      "eval_precision": 0.9974160206718347,
+      "eval_recall": 0.9967721110393802,
+      "eval_runtime": 106.5837,
+      "eval_samples_per_second": 87.537,
+      "eval_steps_per_second": 7.299,
+      "step": 21510
+    },
+    {
+      "epoch": 9.21,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 0.0219,
+      "step": 22000
+    },
+    {
+      "epoch": 9.41,
+      "learning_rate": 1.3128777312877733e-06,
+      "loss": 0.029,
+      "step": 22500
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 8.489074848907486e-07,
+      "loss": 0.0274,
+      "step": 23000
+    },
+    {
+      "epoch": 9.83,
+      "learning_rate": 3.8400743840074387e-07,
+      "loss": 0.0308,
+      "step": 23500
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.9986066452304394,
+      "eval_f1": 0.9979028875625101,
+      "eval_precision": 0.9974201870364399,
+      "eval_recall": 0.9983860555196902,
+      "eval_runtime": 106.9592,
+      "eval_samples_per_second": 87.23,
+      "eval_steps_per_second": 7.274,
+      "step": 23900
+    }
+  ],
+  "max_steps": 23900,
+  "num_train_epochs": 10,
+  "total_flos": 7.54576195666944e+16,
+  "trial_name": null,
+  "trial_params": null
+}

outputs/sketch/checkpoint-23900/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c0f222ab026cf85beb8d17adfee575acbc981e16c71caead2903f720c9273d20
+size 3311

outputs/sketch/cls_score.json

@@ -0,0 +1,3 @@
+{
+    "id-01": 10.994331359863281
+}

requirements.txt

@@ -0,0 +1,6 @@
+transformers==4.21.1
+wandb==0.13.3
+torch==1.12.1
+datasets
+sklearn
+streamlit==1.2.0

retro_reader/__init__.py

@@ -0,0 +1,3 @@
+from .retro_reader import RetroReader
+
+__all__ = ["constants", "retro_reader", "args"]